Changes in government health spending and short-term mortality in four European countries in the 21st century

Changes in government health spending and short-term mortality in four European countries in the 21st century | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Changes in government health spending and short-term mortality in four European countries in the 21st century Politi Julieta, Donat Marta, Guerras Juan Miguel, Herrero Lidia, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7772794/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 14 You are reading this latest preprint version Abstract Background The recent slowdown in mortality decline in high-income countries is often attributed to decreasing in health spending. Objective To characterize the association between government health spending and short-term mortality risk in UK, Italy, Spain, and Greece. Methods Cross-sectional time-series design using annual data from all residents in those countries during 2002–2019. The exposure was the government health spending in thousands of constant dollars per capita and the main outcome was the age-standardized mortality rate. We examined time trends in exposure and outcome. Then, we applied linear regression models (country-fixed-effects -FE- and first-differences -FD-), adjusting for year, country, heatwave, coldwave, influenza, income per capita, and government social spending to estimate the percent change (PC) in mortality rate per unit increase in health spending. Results Mortality rates declined and health spending increased sharply during 2002–2010 in all four countries. Subsequently, mortality slowed its decline, while health spending either slowed its rise (UK) or decreased (other countries). Regression models show a significant association (p < 0.05) between health spending and mortality in the Italy-Spain-Greece group: inverse in 2002–2010 (PC-FE: -7.4%, PC-FD: -9.5%) and direct during 2010–2019 (PC-FE: 7.4%, PC-FD: 6.0%), while in the UK, the association probably was inverse throughout the period 2002–2019. Main results remain introducing cumulative lagged values ​​of health spending. Conclusion The health spending-mortality association went from inverse to direct between 2002–2010 and 2011–2019 in the Italy-Spain-Greece group, so the decline in health spending could not adequately explain the slowdown in mortality decline during 2011–2019 in these countries. Health sciences/Diseases Health sciences/Health care Health sciences/Medical research Health spending. Mortality. Time series analysis. Europe Figures Figure 1 1. INTRODUCTION Mortality reflects the main determinants of population health. After 2010, the downward trend in mortality slowed down in many high-income countries(Global Burden of Disease Greece, 2018; Hiam et al., 2018; PHW-NHS, 2020 ; Raleigh, 2019 ; Ramsay et al., 2020 ; Walsh, Wyper, et al., 2022), something attributed to increased influenza activity or coldwaves(Murphy, 2021 ), obesity increase(Walsh, Tod, et al., 2022 ), and mainly austerity policies (measures to improve the government's fiscal balance) implemented after the Great Recession. Numerous studies suggests a direct association between increased mortality and general austerity(Broadbent et al., 2024 ; McCartney et al., 2022; Rajmil & Fernandez de Sanmamed, 2019 ; Walsh, Wyper, et al., 2022). Among other mechanisms, austerity could be associated with mortality through cuts in government health and social spending, stress associated with unemployment and job insecurity, and changes in lifestyles or risk exposures due to a decrease in economic activity and disposable income. Specifically, cuts in government health spending could worsen population health by reducing the quantity or quality of health care and preventive interventions, for example by increasing waiting times for diagnosis and treatment or facilitating co-payments, which would especially affect vulnerable people(Doetsch et al., 2023 ; Jacques & Noel, 2022). There is evidence quite consistent that health spending is inversely associated with suicide, or with child or maternal mortality(Antonakakis & Collins, 2015; Budhdeo et al., 2015). Regarding general or adult mortality, the evidence is much less consistent, finding an inverse association in some studies or for some causes of death(Budhdeo et al., 2015; Golinelli et al., 2018 ; Hiam et al., 2024 ; Leider et al., 2018 ; Martin et al., 2021 ), but not in other studies or for other causes(Borra et al., 2020; Franklin et al., 2017 ; Lippi et al., 2016 ; Tanaka et al., 2022 ). Research on this topic has considerable methodological limitations, derived mainly from the difficulty of controlling potential confounders, delayed effects, and reverse causality(Murphy & Grundy, 2022 ). The internal validity of studies on the health spending-mortality relationship can be compromised by multiple factors that should be controlled. Thus, studies in high-income countries have shown a direct association of mortality with increased heatwaves(Faurie et al., 2022 ; Xu et al., 2016 ), coldwaves or influenza-activity(Murphy, 2021 ) and economic growth (Parmar et al., 2016 ; Regidor et al., 2016 ; Tapia Granados & Ionides, 2017 ), and with decreased government social spending(Koltai et al., 2021; Loopstra et al., 2016 ; Martin et al., 2021 ; Richardson et al., 2020 ; Seaman et al., 2023 ; Watkins et al., 2017 ). On the other hand, evidence shows that economic growth and government social spending are related to government health spending, so they would be clear confounders. In contrast, influenza activity, heatwaves, and coldwaves are phenomena whose occurrence would apparently not be associated with health spending. However, it is difficult to believe that they are not indirectly related through other omitted covariates. Thus, greater investment or spending in certain health or social infrastructures, or in preventive interventions (e.g., influenza vaccination) can mitigate he adverse effects of these covariates on mortality. Furthermore, these phenomena cause significant short-term variations in mortality, and adjusting for them allows us to reduce the residual variance of the model and obtain more precise association estimates. eFigure 1 represents the main relationships between health spending, mortality, and the adjustment covariates. Previous studies on the health expenditure-mortality relationship have controlled for these covariates to varying degrees, and not all of them have isolated within-country variations in annual health spending (which could be attributed to austerity policies) from between-country variations in mean health spending (probably due to structural factors other than austerity). Such studies have also often failed to assess the possible heterogeneity of the association between health spending and mortality by period, season, country, sex, or age. Consequently, adopting strategies to reduce these weaknesses can improve the validity of these studies' conclusions. We want to test the hypotheses that there is an inverse relationship between government health spending and mortality, and that cuts in government health spending may have contributed to the slowdown in declining mortality during 2011–2019 in the United Kingdom (UK), Italy, Spain, and Greece. The selection of these four countries is justified for several reasons: a) They made important fiscal adjustments from 2010 to reduce government deficits and public debt. The most severe and prolonged measures were imposed on Greece by international creditors in order to receive financial bailouts. In addition to health spending, the measures included other public spending cuts (social protection and welfare, pensions, investments, other public services and even public salaries), tax increases, and structural reforms. These measures were accompanied by a decrease in GDP per capita, and caused intense social and political rejection(Darvas, 2014; Franklin et al., 2017 ; OECD, 2011 ; Paulus, 2017 ). b) There is considerable debate about the health effects of austerity in these countries. For example, some authors point out that in Greece austerity was accompanied by a public health tragedy, with an increase in unmet medical needs, suicides or HIV infections among drug injectors("The Greek health crisis," 2015; Kentikelenis et al., 2014 ), while others conclude that such a tragedy has not existed(Tapia Granados & Rodriguez, 2015). c) Much scientific papers, mainly from the UK, attributes the slowdown in mortality decline after 2010 to austerity policies(Alexiou et al., 2021; Arca et al., 2020; Hiam et al., 2021; Hiam et al., 2024 ; Koltai et al., 2021; Martin et al., 2021 ; McCartney et al., 2022; Richardson et al., 2020 ; Seaman et al., 2023 ; Walsh, Wyper, et al., 2022), including the reduction in government health spending(Arca et al., 2020; Hiam et al., 2021; Martin et al., 2021 ). d) It was quite difficult to obtain data for more countries stratified simultaneously by year, age group, sex, and month of the year, as well as annual data on some adjustment covariates such as heatwaves or coldwaves. The main objective was to characterize the association between government health spending and short-term mortality risk by period, season, sex, age, and country in UK, Italy, Spain, and Greece during 2002–2019. 2. METHODS 2.1. Study design, population and variables A balanced cross-sectional time-series design was applied. Data on all-cause deaths and population by country, year, month, sex, and five-year age group, and data on Gross Domestic Product (GDP), government's expenditures on health and social protection, influenza activity, heatwaves, and coldwaves by country and year were obtained from various sources. GDP was measured in 2015 US dollars per capita (real GDP per capita). Government’s expenditures were measured in thousands 2015 US dollars per capita. Finally, influenza activity, heatwaves and coldwaves were categorized, respectively, into four, three and two intensity levels. Details on variable definitions and data sources are shown in eTables 1 and eTable 2 2.2. Data analysis Initially, annual age-standardized mortality rates per 100,000 person-years (py) and per capita healthcare expenditure were calculated by country, year, and season, specifically spring-summer (April to September) and autumn-winter (the remaining months). Subsequently, time trends in mortality and government health spending were estimated for 2002–2009 and 2010–2019, assuming a linear time trend within each period. Finally, two linear regression models of log-transformed age-standardized mortality rates on health spending, with covariate adjustment, were used in order to estimate the independent association within each country between the variation in health spending and mortality(Wooldridge, 2001 ), specifically the country-level fixed-effects (FE) model and the first-differences (FD) model. Both models used robust (sandwich) standard errors to account for intra-country correlation of rates and these errors were carried over into 95% confidence intervals. The country-level FE model included the level values ​​of variables during the current year and adjusted for country, time trend (as a restricted cubic spline), coldwave, influenza activity, heatwave, non-health social spending and real GDP per capita. However, models for spring-summer mortality were not adjusted for coldwave or influenza activity, and those for autumn-winter mortality were not adjusted for heatwave. Modeling time using cubic splines is a more realistic approximation than using the linear time trend because it better captures and controls the influence of unmeasured confounders that vary slowly over time, allowing for a better model fit and better estimates of the health spending regression coefficient (β). The country indicators removed the effect of measured and unmeasured time-invariant characteristics linked to the country. In this model, the β of health spending gives the multiplicative change in the annual mortality rate per one thousand US dollars increase in annual health spending. This implies that as spending increases by one thousand dollars, the mortality rate is multiplied by e β . The FD model used the first difference between the value of variables during the current year and the previous year, including all the adjustment covariates of FE model except country. In this model, the β of health spending gives the multiplicative change in the ratio of interannual mortality rates per one thousand dollars increase in interannual difference of health spending. It provides an estimate of the association between interannual variations in health spending and mortality when unobserved time-constant variables and other covariates included in the model are kept constant. To facilitate interpretation, all βs were transformed into percentage changes (PCs) as 100[exp( β ) – 1], estimating 95% confidence intervals (CIs) as 100[exp( β ± 1.96 SE ) − 1], where SE is the robust standard error of β . The results were stratified by country, sex, age (< 75 and ≥ 75 years) and period (2002–2010 and 2011–2019). The stratification for 2002–2010 and 2011–2019 was decided because we suspected that there might be period heterogeneity in the results due to unmeasured time-varying factors (e.g., changes in health system productivity). Some regression analyses were also implemented using lagged health and social expenditure data, specifically the mean of the current and previous year (0–1 years lag mean). Incorporating lags ​​is justified primarily because the impact of health expenditures (e.g., hospital infrastructure, health prevention, professional training) can take some time to be reflected in mortality rates. Furthermore, such mean lags reduce the number of predictors (by avoiding including two annual lags), and the risk of multicollinearity due to correlation between both lags. Sensitivity analyses were performed to assess the robustness of the results, using other cumulative lag structures (e.g. 0-1-2 years lag mean) and measuring health expenditure in GDP percent. Models with sequential adjustment for the different covariates (step-up models) were also implemented to see the influence of each adjustment on the results. Some details on data analysis are given in eTable 3. All calculations were performed using Stata and R. 3. RESULTS 3.1. Mortality and health spending by country The evolution of mortality rates and health spending per capita are shown in Fig. 1 . In general, there was a downward trend in annual, spring-summer and autumn-winter mortality in the four countries, although its slope softened from 2011 to 2019. Thus, during 2002–2010 the mortality APCs were − 1.5% (Spain), -1.3% (UK), -1.2% (Greece), and − 1.1% (Italy), while during 2011–2019, they were − 1.1% (Spain), -0.9% (UK), -0.8% (Italy), and − 0.5% (Greece). Meanwhile, health spending increased sharply during 2002–2010 in the four countries, with APCs of 4.5% (Spain), 4.4% (UK), 4.4% (Greece) and 1.8% (Italy), and during 2011–2019, its increase slowed considerably in the UK (1.0%) and Spain (0.5%) and decreased in Greece (-3.9%) and Italy (-0.9%). There were clear annual spikes in spring-summer mortality in 2003 in all countries, 2007 and 2012 in Greece and 2015 in Italy, and also in autumn-winter mortality in 2005 in Italy and Spain, 2015 in all countries, and 2012 and 2017 in Italy, Spain and Greece. 3.2. Association between health spending and mortality risk Figure 1 shows that coinciding with the strong growth in health spending in 2002–2010, mortality fell considerably, while subsequently spending moderated its growth or fell and the downward trend in mortality slowed in all four countries. Focusing on the APCs, it is observed that during 2011–2019 the largest cuts in health spending and the greatest slowdown in mortality occurred in Greece. In the four-country pooled series, lagged and non-lagged FE and FD regression models showed an inverse association between health spending and mortality during 2002–2010 (i.e. mortality decreases as health spending increases), stronger for autumn-winter than spring-summer mortality. However, during 2011–2019, they showed a direct association stronger for spring-summer mortality (Table 1 ). The association was inverse during 2002–2010 and direct during 2011–2019, both in men and women and people aged < 75 and ≥ 75, being stronger in people aged ≥ 75 than < 75 years. Furthermore, during 2002–2010, in FE models, the association was stronger in men than in women, while in FD models, the opposite occurred (Table 2 ). Table 1 Association between age-standardized mortality rates and annual government health spending from regression models by model type, period and season, 2002–2019 N 5 Annual mortality Spring-summer mortality 6 Autumn-winter mortality 7 PC 95%CI PC 95%CI PC 95%CI Non lagged regression model 1 Country-level fixed-effect (FE) model 2 2002–2010 36 -12.0 -15.7, -8.0 -10.3 -15.2, -5.2 -15.4 -21.5, -9.0 2011–2019 36 8.8 4.4, 13.3 13.2 6.7, 20.2 7.0 4.3, 9.8 Total 72 -5.0 -8.0, -2.0 -3.0 -5.7, -0.2 -7.5 -10.6, -4.2 First-differences (FD) model 3 2002–2010 36 -12.5 -17.6, -7.1 -7.6 -20.2, 7.0 -15.2 -22.8, -6.8 2011–2019 36 5.5 1.9, 9.2 11.4 4.9, 18.3 4.2 -1.1, 9.8 Total 72 -5.6 -9.3, -1.8 -1.8 -8.8, 5.7 -7.7 -13.5, -1.6 Lagged regression model (0–1 year lag mean) 4 FE model 2 2002–2010 36 -12,5 -16.3, -8.5 -11.4 -13.2, -9.6 -15.9 -21.1, -10.3 2011–2019 36 7.4 2.5, 12.5 9.5 4.8, 14.3 5.9 -1.0, 13.4 Total 72 -4,4 -6.3, -2.4 -2.6 -4.4, -0.8 -6.8 -8.5, -5.1 FD model 3 2002–2010 36 -9.0 -20.1, 3.7 -5.9 -16.7, 6.4 -10.2 -25.5, 8.2 2011–2019 36 6.8 -2.7, 17.2 10.5 -2.8, 25.6 -3.8 -16.6, 11.0 Total 72 -5.0 -7.3, -2.6 -0.1 -2.0, 1.8 -10.1 -13.1, -7.0 PC : Percent change in annual mortality rate (FE model) or ratio of interannual mortality rates (FD model) per one thousand US dollars increase in annual health spending (FE model) or interanual difference in health spending (FD model). 95%CI : 95% confidence interval of PC. 1 Health spending and social spending in the regression model correspond to the same year as mortality and the rest of the independent variables. 2 Multiple linear regression of logarithm of age-standardized mortality rates on health spending in thousands of 2015 US dollars ($) per capita, adjusting for country, year as a restricted cubic spline and other covariates, as heatwave, coldwave, influenza activity, real GDP per capita and social spending per capita. 3 Multiple linear regression of first differences between the logarithm of age-standardized mortality rate in a given year and country, and its logarithm in the previous year on first differences of health spending, adjusting for first-differences in year as a restricted cubic spline and first-differences in the covariates heatwave, coldwave, influenza activity, real GDP per capita, and social spending in thousands of 2015 US $ per capita. 4 Health spending and social spending in the regression model correspond to the mean of the current and previous year's values. 5 It refers to the number of points in the regression models. 6 It corresponds to mortality between April and September, both included. For this outcome, results of the regression models were not adjusted for influenza activity or coldwave. 7 It corresponds to mortality during the remaining months of the year. For this outcome, results of the regression models were not adjusted for heatwave. Table 2 Association between annual age-standardized mortality rates and annual government health spending from regression models by model type and period, in men and women and in two age group, 2002–2019. N 5 Men Women < 75 years ≥ 75 years PC 95%CI PC 95%CI PC 95%CI PC 95%CI Non lagged regression model 1 Country-level fixed-effect (FE) model 2 2002–2010 36 -10.9 -15.0, -6.6 -13.8 -17.8, -9.6 -7.8 -9.1, -6.4 -13.7 -18.9, -8.1 2011–2019 36 8.1 2.1, 14.4 9.8 6.4, 13.3 2.5 0.0, 5.0 11.9 5.7, 18.4 Total 72 -5.2 -9.5, -0.6 -5.3 -7.6, -2.8 -5.5 -8.3, -2.6 -4.6 -9.8, 0.9 First-differences (FD) model 3 2002–2010 36 -14.2 -19.7, -8.3 -11.4 -16.5, -6.0 -8.6 -11.9, -5.1 -14.2 -21.0, -6.8 2011–2019 36 6.4 1.3, 11.7 5.6 3.0, 8.2 2.7 0.2, 5.4 6.9 1.2, 12.9 Total 72 -5.7 -11.6, 0.7 -5.6 -7.5, -3.6 -5.0 -6.2, -3.7 -5.9 -11.7, 0.3 Lagged regression model (0–1 year lag mean) 4 FE model 2 2002–2010 36 -10.5 -15.5, -5.2 -15.4 -18.5, -12.1 -8.0 -9.4, -6.6 -14.3 -19.4, -8.9 2011–2019 36 7.0 1.7, 12.5 7.0 3.0, 11.2 0.9 -0.8, 2.7 10.5 3.6, 17.9 Total 72 -4.6 -7.9, -1.1 -4.5 -6.0, -3.0 -5.1 -8.4, -1.7 -3.8 -7.9, 0.5 FD model 3 2002–2010 36 -10.0 -20.6, 1.9 -8.6 -21.2, 6.2 -6.2 -9.4, -2.9 -10.4 -25.0, 7.1 2011–2019 36 7.9 0.5, 15.9 6.4 -5.2, 19.3 3.7 -3.3, 11.1 8.4 -2.2, 20.1 Total 72 -4.3 -7.9, -0.6 -5.7 -10.0, -1.2 -4.5 -9.3, 0.5 -5.2 -7.9, -2.4 PC : Percent change in annual mortality rate (FE model) or ratio of interannual mortality rates (FD model) per one thousand US dollars increase in annual health spending (FE model) or interanual difference in health spending (FD model). 95%CI : 95% confidence interval of PC. 1 Health spending and social spending in the regression model correspond to the same year as mortality and the rest of the independent variables. 2 Multiple linear regression of logarithm of age-standardized mortality rates on government health spending in thousands of 2015 US dollars ($) per capita, adjusting for country, year as a restricted cubic spline and other covariates, as heatwave, coldwave, influenza activity, real GDP per capita and social spending per capita. 3 Multiple linear regression of first differences between the logarithm of age-standardized mortality rate in a given year and country, and its logarithm in the previous year on first differences of health spending, adjusting for first-differences in year as a restricted cubic spline and first-differences in the covariates heatwave, coldwave, influenza activity, real GDP per capita and social spending in thousands of 2015 US $ per capita. 4 Health spending and social spending in the regression model correspond to the mean of the current and previous year's values. 5 It refers to the number of points in the regression models. Table 3 shows the association between health spending and mortality separately for the UK and the other three countries from FE models adjusted by different sets of covariates. Although the analysis is not very robust for the UK due to the limited number of degrees of freedom, there was an inverse association in both country groups during 2002–2010, which appears to be much stronger in the UK. Meanwhile, during 2011–2019, the PC sign changed in the three-country group, becoming positive, something that was also observed in the fully adjusted FD model (not tabulated), where the PC changed from − 9.5% (95%CI:-13.7, -5.0%) during 2002–2010 to 6.1% (95%CI:3.9–8.3%) during 2011–2019. In the UK, however, according to Table 3 the association appears to remain inverse. Table 3 Association between annual age-standardized mortality rates and annual government health spending from country-level fixed-effects models adjusted for various set of covariates by period and country group, 2002–2019. 2002–2010 2011–2019 Total N PC 95%CI N PC 95%CI N PC 95%CI UK 1 Model 1 2 9 -18.7 -40.8, 11.7 9 -24.6 -53.1, 21.1 18 -18.9 -37.0, 4.6 Model 2 3 9 -28.3 -48.1, -1.0 9 -1.1 -47.8, 87.6 18 -18.8 -38.3, 7.0 Model 3 4 9 -23.4 -41.5, 0.3 9 -27.5 -60.8, 33.9 18 -19.8 -37.9, 3.5 Model 4 5 9 -13.0 -55.1, 68.6 9 -16.3 -60.3, 76.1 18 -18.6 -37.5, 6.0 Model 5 6 9 -11.7 -39.8, 29.5 9 -32.6 -57.5, 6.9 18 -18.0 -36.5, 5.9 Model 6 7 9 -26.9 -76.9, 131.2 9 -26.8 -57.0, 24.5 18 -8.8 -30.9, 20.3 Other three countries 8 Model 1 2 27 -0.1 -9.3, 0.1 27 3.2 -0.6, 7.1 54 -5.5 -9.6, -1.3 Model 2 3 27 -6.0 -9.8, -1.9 27 2.6 -0.7,6.0 54 -5.2 -7.7, -2.7 Model 3 4 27 0.8 -7.7, 9.9 27 3.4 -1.0, 7.9 54 -5.7 -9.5, -1.7 Model 4 5 27 -0.4 -11.5, 12.2 27 3.3 0.8, 5.8 54 -5.7 -10.4, -0.7 Model 5 6 27 1.2 -8.0, 11.4 27 3.5 0.1, 7.1 54 -3.1 -6.0, -0.1 Model 6 7 27 -3.3 -11.3, 5.5 27 9.8 0.5, 20.0 54 -9.5 -15.5, -3.0 Model 7 9 27 -7.4 -9.8, -5.0 27 7.4 4.2, 10.8 54 -7.0 -10.7, -3.2 PC : Percent change in annual mortality rate per one thousand US dollars increase in annual health spending. 95%CI : 95% confidence interval of PC. 1 Includes England & Wales and Scotland for mortality and influenza activity and the entire United Kingdom for the rest of the variables. 2 Model 1: Fixed-effects multiple linear regression models of logarithm of age-standardized mortality rates on health spending in thousands of 2015 US dollars per capita, adjusting for country and year as a restricted cubic spline. 3 Model 2: Same as model 1, but additionally adjusting for social spending in thousands of 2015 US dollars per capita (SS). 4 Model 3: Same as model 1, but adjusting for heatwaves instead of SS. Model 4: Same as model 1, but adjusting for coldwaves instead of SS. 6 Model 5: Same as model 1, but adjusting for influenza activity instead of SS. 7 Model 6: Same as model 1, but adjusting for Gross Domestic Product per capita instead SS. 8 Includes Italy, Spain and Greece analysed together and applying country fixed-effects. 9 It refers to the number of points in regression models. 9 Model 7: Full model adjusted for all variables considered. This model was not calculated for the UK due to a lack of degrees of freedom. 3.3. Sensitivity analyses The association between health spending and annual mortality was quite similar when spending was measured in GDP percent (eTable 4) and when various accumulative lag structures of health and social spending were used (eTable 5). The forward entry models showed a considerable increase in the strength of the inverse association between health spending and mortality during 2002–2010 when introducing social spending as an adjustment covariate (eTable 6), reflecting a direct association of this covariate with mortality during that period. 4. DISCUSSION 4.1. Main findings The trend analysis shows that the strong upward trend in government health spending was interrupted around 2010 (with a slowdown in the rise or even a decline, depending on the country), while mortality continued to decline, albeit at a slower pace. The regression analysis, after adjusting for heatwaves, coldwaves, influenza activity, economic growth and government social spending, does not show an inverse association between health spending and short-term mortality in Italy, Spain and Greece during 2011–2019. However, a fairly strong inverse association was identified in 2002–2010 in all four countries, which appears to have remained in the UK during 2011–2019. Some heterogeneity in such associations was detected according to age and season. Main results remain introducing cumulative lagged values ​​of health and social spending (up to 3 years). 4.2. On the explanatory factors of the slowdown in mortality decline We had hypothesized that cuts in government health expenditure may have contributed to the slowdown in declining mortality in the countries analysed during 2011–2019. The results would support the hypothesis for the UK, but not for the other three countries, where such cuts do not seem to be a relevant explanatory factor, requiring other time-varying explanatory factors to be considered(Jasilionis et al., 2023 ; Raleigh, 2019 ). As noted, we controled some of the time-varying confounders, such as influenza activity, coldwaves, heatwaves, economic growth or government social expenditure. Although we identified a positive association between influenza activity and mortality, as in previous studies(Murphy, 2021 ), no increase in influenza was detected in 2011–2019, that could explain the slowdown in declining mortality, something previously ruled out for the UK and USA(Hiam et al., 2024 ; Murphy & Grundy, 2022 ). The same is true for coldwaves. Regarding government social spending, although previous research in high-income countries shows an inverse association with mortality (Koltai et al., 2021; Loopstra et al., 2016 ; Martin et al., 2021 ; Richardson et al., 2020 ; Seaman et al., 2023 ; Watkins et al., 2017 ), we did not find such association in Italy, Spain and Greece during 2011–2019. Therefore, it does not seem to be a relevant factor in explaining the slowdown in mortality decline. On the contrary, global warming and heatwaves(Global Burden of Disease Greece, 2018; Lee et al., 2019) could help partially explain such slowdown. Thus, in our work, a clear positive association between heatwaves and mortality was found in Italy, Spain and Greece, stronger during 2011–2019 than previously. Moreover, from international data(WBG, 2024 ) it is estimated that between 2002–2010 and 2011–2019, the average maximum air surface temperature across the quarter June-September increased 0.5–0.6 degrees Celsius in Italy, Spain and Greece. Other time-varying unmeasured factors may have contributed to confounding the association between health spending and mortality and explain the slowdown in declining mortality during 2011–2019, such as obesity, alcohol and drug use, general austerity, changes in the productivity of health systems or even changes in the frequency or lethality of diseases and injuries. Although obesity has been proposed as a possible confounder(Walsh, Tod, et al., 2022 ) , (Murphy & Grundy, 2022 ), is unlikely to have been a relevant factor because a clear increase in its prevalence during 2008–2020 in the countries studied has not been identified(WHO, 2024 ). However, some doubt remains, because obesity could have a very lagged impact on mortality and reports on obesity become less reliable as the response rate decreases(Walsh, Tod, et al., 2022 ). Per capita alcohol consumption may have contributed to the slowdown in mortality decline because a downward trend was observed during 2002–2010, which, except in Greece, slowed or reversed starting in 2011.Thus, the annual decline in UK, Italy and Spain went from − 1.3%, -3.7% and − 2.4%, respectively, in 2002–2010 to -0.2%, 1.2% and 1.6% in 2011–2019(WHO, 2024 ). There are also reports from the UK and Spain that during the second decade of the 21st century the downward trend in alcohol-related mortality stabilised or reversed(Augarde et al., 2022; Donat et al., 2023). Regarding drug use, some UK research finds evidence of an inverse association between government spending on welfare and drug-related mortality(Friebel et al., 2022; Koltai et al., 2021), which can be explained by the increase in drug use due to coping with the stress attributable to austerity or the decrease in investments in prevention and harm reduction programs. However, the proportion of drug-related deaths in the analysed countries is low, so it is unlikely that drug use is a relevant explanatory factor for the slowdown in declining overall mortality. Previous research in high-income countries has found a direct association between mortality and general austerity (Alexiou et al., 2021; Golinelli et al., 2017 ; McCartney et al., 2022; Rajmil & Fernandez de Sanmamed, 2019 ; Walsh, Wyper, et al., 2022). However, we did not find this association in Italy, Spain and Greece during 2011–2019, using the cyclically adjusted fiscal balance as an indicator (data not shown). Finally, another candidate explanatory factor to explain the slowdown in declining mortality is the increase in productivity of health systems due to changes in spending distribution, shifts in care models, technological innovations, changes in health services management or greater workforce's dedication(Garcia-Escribano, 2022 ; Golinelli et al., 2017 ; Hernández, 2016 ; Jacques & Noel, 2022; ONS, 2022 ). It is not easy to find evidence to clarify this issue. However, a recent study concludes that during 2012–2015 the health systems productivity increased in most European countries, including Spain, Italy and Greece, but not in the UK, where it decreased. In addition, this increased productivity was positively associated with increased efficiency in health systems management, but not in health technology level(Gomez-Gallego, 2019 ). However, OECD data on diagnostic examinations with computerized axial tomography and magnetic resonance, show that mainly since 2016 the technological level increased considerably in all analysed countries, especially in Spain and the UK. Another study draws similar conclusions for Spain(Regidor, 2021 ). 4.3. On the association between health spending and mortality The inverse association between government health spending and short-term mortality during 2002–2010 in all four countries and in UK during 2011–2019 is consistent with the results of numerous previous studies (often from the UK)(Budhdeo et al., 2015; Hiam et al., 2024 ; Leider et al., 2018 ; Mackenbach et al., 2011 ; Martin et al., 2021 ; Mays & Smith, 2011; Watkins et al., 2017 ). This apparently supports the conclusion of some authors that the decrease in health spending worsen population health and vice versa. However, this does not seem too consistent with our finding that during 2011–2019 mortality continued to decline at a good pace (between − 0.5% and − 1.1% annually) in all four countries, despite a radical change in health spending trends after 2010 (from a strong upward trend to a decline or slowdown in the rise, depending on the country). And it is also not consistent with the unexpected change in the direction of the health spending-mortality association, from inverse to direct in Italy, Spain and Greece during 2011–2019, a period that includes years of strong austerity (2011–2015) and declining government health spending. Taken together, these findings suggest that at the levels of government health spending analysed, the hypothesized inverse association between health spending and mortality in certain contexts could not exist or be weak and easily offset or reversed by unmeasured or poorly measured time-varying factors. These confounders may include increased productivity of health systems or other factors mentioned above. It should be noted that previous studies in high-income countries have also failed to find a clear inverse association(Borra et al., 2020; Franklin et al., 2017 ; Lippi et al., 2016 ; Tanaka et al., 2022 ). 4.4. Heterogeneity in health spending-mortality association by age, sex and season of the year The study results show a stronger association between health expenditure and mortality among older people (≥ 75 years), whether the association is inverse (2002–2010) or direct (2011–2019 in Italy, Spain and Greece). Some previous studies(Hiam et al., 2018; Loopstra et al., 2016 ; Watkins et al., 2017 ), although not all(Toffolutti & Suhrcke, 2019 ), have suggested an inverse association between general government austerity and mortality in older people, where health or social spending and mortality are concentrated(Kinge et al., 2023 ). Our results do not show clear differences by sex in the association studied. There are hardly any studies on this topic, but a British study found some evidence that among more deprived populations, female mortality rates have worsened to a greater degree during the years of general austerity(Walsh, Dundas, et al., 2022). Finally, our results suggest that when the association health spending-mortality was inverse (2002–2010) it was stronger in autumn-winter, while when it was positive (2011–2019 in Italy, Spain and Greece) the association was stronger in spring-summer. It is difficult to explain these differences and, to our knowledge, there are no previous studies focusing on this topic, although results suggest that health spending would preferentially affect treatable or preventable conditions that are more prevalent or that tend to worsen in autumn-winter. 4.5. Strengths and limitations The main strengths are the absorption through regression models of between-country differences in mean health spending to isolate the association between variations in health spending and mortality within each country. A strategy that would make it possible to attribute changes in mortality to changes in health spending with greater validity. Another strength is a more complete control strategy of possible time-varying confounders than in previous studies. Main limitations include insufficient quality of data on extreme temperatures and influenza and a short time series, which limits the statistical power for some stratified analysis, for example, by country and period. Although a few potential confounders were controlled, residual confounding possibly persists. Further control was not possible due to sample size limitations, difficulties in obtaining data, and the need to avoid overfitting. For example, adjusting for the unemployment rate would have meant overfitting because its changes are strongly associated with increase in GDP per capita. Private health expenditure is very unlikely to bias the results because it represents a modest proportion (20–37%) of total health expenditure in the analysed countries with a very small temporal variation within countries (although during the financial crisis, its proportion decreased slightly). Finally, it is not convenient to adjust for possible mediators between health spending and mortality (e.g., waiting time for diagnosis or treatment, or practicing physicians per capita), because it could eliminate inadequately the health spending effect that passes through the mediator. There could also be numerator-denominator biases related to migration and tourism (probably small), as well as harvesting phenomena or mortality displacements, given that during economic crises (generally associated with decreased mortality) some deaths are delayed, while the opposite usually happens during periods of extreme temperatures or increased influenza activity. Our design does not allow inferring causal relationships. However, it seems unlikely that the slowdown in declining mortality caused the observed changes in health spending (e.g. stagnation or decline), because the latter generally began before the change in mortality trend. Furthermore, the associations remained after introducing cumulative lagged values ​​of health spending. 4.6. Policy and research implications The results during 2011–2019 in three of four high-income countries analysed, suggest that the nature of the association health spending-mortality is very dependent on time-varying omitted covariates linked to the period and country. Therefore, as long as those omitted covariates that mediate or confound the association are not clearly identified, the generalisability of the results on such association is limited. To obtain more valid results, and avoid the influence of structural factors, it is advisable to focus only on within-country temporal variations in health spending and mortality. It is also very useful to have large time series with many geographical units and years, which allows controlling multiple time-varying confounders, including their possible delayed effects. The observation that stagnation or decrease in health spending is not always accompanied by an increase in mortality suggests that health system decision-makers, in addition to ensuring sufficient funding, should focus on improving other aspects (e.g., organizational or management) that increase the health systems' productivity. Declarations DECLARATION OF INTEREST STATEMENT: The authors have had no financial or personal relationships with other people or organizations that could inappropriately influence or bias their work. ACKNOWLEDGEMENTS: The authors would like to thank the statistical offices of the countries included in the study, specifically the Office for National Statistics (England & Wales), National Records of Scotland/Scotlands People (Scotland), Istituto Nazionale di Statistica -ISTAT- (Italy), Instituto Nacional de Estadística -INE- (Spain) and the Hellenic Statistical Authority (ELSTAT) for sending mortality data or for facilitating access to data. FUNDING : This work was supported by a grant from the Carlos III Health Institute and the European Regional Development Fund (PI16/00455). The funding organization did not intervene or interfere in the design and conduction of the study; collection, management, analysis, and interpretation of the data; nor in the preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. AUTHOR CONTRIBUTION Conceptualization and study design: Julieta Politi, Marta Donat, Gregorio Barrio, Enrique Regidor, María José Belza. Data acquisition and management: Julieta Politi, Lidia Herrero, Enrique Regidor, Juan Miguel Guerras. Data analysis and interpretation: Julieta Politi, Marta Donat, Gregorio Barrio, Roberto Pastor-Barriuso, Juan Miguel Guerras. Drafting the manuscript: Julieta Politi, Marta Donat, Juan Miguel Guerras. Critical revision for important intellectual content: All authors. Supervision and senior authorship: María José Belza, Enrique Regidor. All authors read and approved the final version of the manuscript. ETHICAL APPROVAL The Review Committee of the Carlos III Health Institute approved the study proposal. Ethics Committee approval was not deemed necessary because the mortality microdata used, although comprised of individual records, do not contain variables that could directly or indirectly facilitate personal identification. The national statistical offices adopts the necessary logical, physical, and administrative measures to ensure the effective protection of confidential data, from collection to publication. In accordance with the Public Statistical Service and Regulation (EC) No 223/2009, it does not disseminate personal data from any source. Microdata files are distributed anonymously, taking precautions to avoid any indirect identification. For example, the deceased's name, surname, Identification Document, address, death registration information, date of birth, or codes for municipalities with municipalities with a small population size are not provided. This study was based on aggregated, anonymized population-level data provided by national statistical offices (ONS, National Records of Scotland, ISTAT, INE, and ELSTAT). As the analyses did not involve identifiable individual-level data, approval from a Research Ethics Committee was not required according to national regulations. DATA AVAILABILITY The datasets generated and/or analysed during the current study are available from the corresponding author upon reasonable request, for academic and non-commercial purposes. Mortality and population data were obtained from the national statistical offices listed above in aggregated and anonymized form, under data use agreements that do not permit public sharing. References Alexiou, A., Fahy, K., Mason, K., Bennett, D., Brown, H., Bambra, C., Taylor-Robinson, D., & Barr, B. (2021, Sep). Local government funding and life expectancy in England: a longitudinal ecological study. Lancet Public Health, 6 (9), e641-e647. https://doi.org/10.1016/S2468-2667(21)00110-9 Antonakakis, N., & Collins, A. (2015, Nov). The impact of fiscal austerity on suicide mortality: Evidence across the 'Eurozone periphery'. Soc Sci Med, 145 , 63-78. https://doi.org/10.1016/j.socscimed.2015.09.033 Arca, E., Principe, F., & Van Doorslaer, E. (2020, Dec). Death by austerity? The impact of cost containment on avoidable mortality in Italy. Health Econ, 29 (12), 1500-1516. https://doi.org/10.1002/hec.4147 Augarde, E., Gunnell, D., Mars, B., & Hickman, M. (2022, Jun). An ecological study of temporal trends in 'deaths of despair' in England and Wales. Soc Psychiatry Psychiatr Epidemiol, 57 (6), 1135-1144. https://doi.org/10.1007/s00127-022-02251-9 Borra, C., Pons-Pons, J., & Vilar-Rodriguez, M. (2020, Apr). Austerity, healthcare provision, and health outcomes in Spain. Eur J Health Econ, 21 (3), 409-423. https://doi.org/10.1007/s10198-019-01141-3 Broadbent, P., Walsh, D., Katikireddi, S. V., Gallagher, C., Dundas, R., & McCartney, G. (2024, Oct). Is Austerity Responsible for the Stalled Mortality Trends Across Many High-Income Countries? A Systematic Review. Int J Soc Determinants Health Health Serv, 54 (4), 362-379. https://doi.org/10.1177/27551938241255041 Budhdeo, S., Watkins, J., Atun, R., Williams, C., Zeltner, T., & Maruthappu, M. (2015, Dec). Changes in government spending on healthcare and population mortality in the European union, 1995-2010: a cross-sectional ecological study. J R Soc Med, 108 (12), 490-498. https://doi.org/10.1177/0141076815600907 Darvas, Z. H., P.; De Sousa, C.; Terzi, A.; Tschekassin, O. . (2014). Austerity and Poverty in the European Union . E. P. D. G. f. I. Policies. https://www.europarl.europa.eu/RegData/etudes/STUD/2014/518758/IPOL_STU(2014)518758_EN.pdf Doetsch, J. N., Schlosser, C., Barros, H., Shaw, D., Krafft, T., & Pilot, E. (2023, Jan 5). A scoping review on the impact of austerity on healthcare access in the European Union: rethinking austerity for the most vulnerable. Int J Equity Health, 22 (1), 3. https://doi.org/10.1186/s12939-022-01806-1 Donat, M., Regidor, E., Barrio, G., Ambrosio, E., Sordo, L., Guerras, J. M., Politi, J., & Belza, M. J. (2023, Oct). Increase in educational inequalities in alcohol-related mortality in Spain during a period of economic growth. Addiction, 118 (10), 1920-1931. https://doi.org/10.1111/add.16264 Faurie, C., Varghese, B. M., Liu, J., & Bi, P. (2022, Dec 15). Association between high temperature and heatwaves with heat-related illnesses: A systematic review and meta analysis. Sci Total Environ, 852 , 158332. https://doi.org/10.1016/j.scitotenv.2022.158332 Franklin, B., Hochlaf, D., & Holley-Moore, G. (2017). Europe during the austerity years a research report from ILC-UK . Friebel, R., Yoo, K. J., & Maynou, L. (2022, Apr). Opioid abuse and austerity: Evidence on health service use and mortality in England. Soc Sci Med, 298 , 114511. https://doi.org/10.1016/j.socscimed.2021.114511 Garcia-Escribano, M. J., P. ; Mogues, T. (2022). Patterns and Drivers of Health Spending Efficiency . I. M. Found. Global Burden of Disease Greece Collaborators. (2018, Aug). The burden of disease in Greece, health loss, risk factors, and health financing, 2000-16: an analysis of the Global Burden of Disease Study 2016. Lancet Public Health, 3 (8), e395-e406. https://doi.org/10.1016/S2468-2667(18)30130-0 Golinelli, D., Bucci, A., Toscano, F., Filicori, F., & Fantini, M. P. (2018, Aug 29). Real and predicted mortality under health spending constraints in Italy: a time trend analysis through artificial neural networks. BMC Health Serv Res, 18 (1), 671. https://doi.org/10.1186/s12913-018-3473-3 Golinelli, D., Toscano, F., Bucci, A., Lenzi, J., Fantini, M. P., Nante, N., & Messina, G. (2017, Dec). Health Expenditure and All-Cause Mortality in the 'Galaxy' of Italian Regional Healthcare Systems: A 15-Year Panel Data Analysis. Appl Health Econ Health Policy, 15 (6), 773-783. https://doi.org/10.1007/s40258-017-0342-x Gomez-Gallego, J. C. G., J.F.; Gómez-Gallego, M. (2019). Efiuciencia y productividad de los sistemas de salud de los países de la Unión Europea [Efficiency and productivity of the health systems of the countries of the European Union]. Estudios de Economía Aplicada, 37 (2), 1-19. https://repositorio.ucam.edu/handle/10952/5365 The Greek health crisis. (2015, Jul 11). Lancet, 386 (9989), 104. https://doi.org/10.1016/S0140-6736(15)61226-0 Hernández, J. (2016). Productividad y gasto sanitario público durante la crisis: una aproximación al cálculo de la productividad agregada del SNS en el periodo 2009-2013. Presupuesto y Gasto Público, 85 , 165-189. Hiam, L., Dorling, D., & McKee, M. (2021, Mar). Austerity, not influenza, caused the UK's health to deteriorate. Let's not make the same mistake again. J Epidemiol Community Health, 75 (3), 312. https://doi.org/10.1136/jech-2020-215556 Hiam, L., Harrison, D., McKee, M., & Dorling, D. (2018, May). Why is life expectancy in England and Wales 'stalling'? J Epidemiol Community Health, 72 (5), 404-408. https://doi.org/10.1136/jech-2017-210401 Hiam, L., McKee, M., & Dorling, D. (2024, Mar 13). Influenza: cause or excuse? An analysis of flu's influence on worsening mortality trends in England and Wales, 2010-19. Br Med Bull, 149 (1), 72-89. https://doi.org/10.1093/bmb/ldad028 Jacques, O., & Noel, A. (2022, Sep). The politics of public health investments. Soc Sci Med, 309 , 115272. https://doi.org/10.1016/j.socscimed.2022.115272 Jasilionis, D., van Raalte, A. A., Klusener, S., & Grigoriev, P. (2023, Aug). The underwhelming German life expectancy. Eur J Epidemiol, 38 (8), 839-850. https://doi.org/10.1007/s10654-023-00995-5 Kentikelenis, A., Karanikolos, M., Reeves, A., McKee, M., & Stuckler, D. (2014, Feb 22). Greece's health crisis: from austerity to denialism. Lancet, 383 (9918), 748-753. https://doi.org/10.1016/S0140-6736(13)62291-6 Kinge, J. M., Dieleman, J. L., Karlstad, O., Knudsen, A. K., Klitkou, S. T., Hay, S. I., Vos, T., Murray, C. J. L., & Vollset, S. E. (2023, Jun 6). Disease-specific health spending by age, sex, and type of care in Norway: a national health registry study. BMC Med, 21 (1), 201. https://doi.org/10.1186/s12916-023-02896-6 Koltai, J., McKee, M., & Stuckler, D. (2021, Sep). Association between disability-related budget reductions and increasing drug-related mortality across local authorities in Great Britain. Soc Sci Med, 284 , 114225. https://doi.org/10.1016/j.socscimed.2021.114225 Lee, J. Y., Kim, H., Gasparrini, A., Armstrong, B., Bell, M. L., Sera, F., Lavigne, E., Abrutzky, R., Tong, S., Coelho, M., Saldiva, P. H. N., Correa, P. M., Ortega, N. V., Kan, H., Garcia, S. O., Kysely, J., Urban, A., Orru, H., Indermitte, E., Jaakkola, J. J. K., Ryti, N. R. I., Pascal, M., Goodman, P. G., Zeka, A., Michelozzi, P., Scortichini, M., Hashizume, M., Honda, Y., Hurtado, M., Cruz, J., Seposo, X., Nunes, B., Teixeira, J. P., Tobias, A., Iniguez, C., Forsberg, B., Astrom, C., Vicedo-Cabrera, A. M., Ragettli, M. S., Guo, Y. L., Chen, B. Y., Zanobetti, A., Schwartz, J., Dang, T. N., Do Van, D., Mayvaneh, F., Overcenco, A., Li, S., & Guo, Y. (2019, Oct). Predicted temperature-increase-induced global health burden and its regional variability. Environ Int, 131 , 105027. https://doi.org/10.1016/j.envint.2019.105027 Leider, J. P., Alfonso, N., Resnick, B., Brady, E., McCullough, J. M., & Bishai, D. (2018, Apr). Assessing The Value Of 40 Years Of Local Public Expenditures On Health. Health Aff (Millwood), 37 (4), 560-569. https://doi.org/10.1377/hlthaff.2017.1171 Lippi, G., Mattiuzzi, C., & Cervellin, G. (2016, Jul). No correlation between health care expenditure and mortality in the European Union. Eur J Intern Med, 32 , e13-14. https://doi.org/10.1016/j.ejim.2016.02.025 Loopstra, R., McKee, M., Katikireddi, S. V., Taylor-Robinson, D., Barr, B., & Stuckler, D. (2016, Mar). Austerity and old-age mortality in England: a longitudinal cross-local area analysis, 2007-2013. J R Soc Med, 109 (3), 109-116. https://doi.org/10.1177/0141076816632215 Mackenbach, J. P., Slobbe, L., Looman, C. W., van der Heide, A., Polder, J., & Garssen, J. (2011, Dec). Sharp upturn of life expectancy in the Netherlands: effect of more health care for the elderly? Eur J Epidemiol, 26 (12), 903-914. https://doi.org/10.1007/s10654-011-9633-y Martin, S., Longo, F., Lomas, J., & Claxton, K. (2021, Oct 15). Causal impact of social care, public health and healthcare expenditure on mortality in England: cross-sectional evidence for 2013/2014. BMJ Open, 11 (10), e046417. https://doi.org/10.1136/bmjopen-2020-046417 Mays, G. P., & Smith, S. A. (2011, Aug). Evidence links increases in public health spending to declines in preventable deaths. Health Aff (Millwood), 30 (8), 1585-1593. https://doi.org/10.1377/hlthaff.2011.0196 McCartney, G., McMaster, R., Popham, F., Dundas, R., & Walsh, D. (2022, Nov). Is austerity a cause of slower improvements in mortality in high-income countries? A panel analysis. Soc Sci Med, 313 , 115397. https://doi.org/10.1016/j.socscimed.2022.115397 Murphy, M. (2021, May 5). Recent mortality in Britain: a review of trends and explanations. Age Ageing, 50 (3), 676-683. https://doi.org/10.1093/ageing/afab016 Murphy, M. J., & Grundy, E. M. D. (2022, May 27). Slowdown in Mortality Improvement in the Past Decade: A US/UK Comparison. J Gerontol B Psychol Sci Soc Sci, 77 (Suppl_2), S138-S147. https://doi.org/10.1093/geronb/gbab220 OECD. (2011). OECD Economic Outlook. https://doi.org/http://dx.doi.org/10.1787/eco_outlook-v2011-2-e ONS. (2022). Public service productivity: total, UK, 2019. https://www.ons.gov.uk/economy/economicoutputandproductivity/publicservicesproductivity/articles/publicservicesproductivityestimatestotalpublicservices/2019 Parmar, D., Stavropoulou, C., & Ioannidis, J. P. (2016, Sep 6). Health outcomes during the 2008 financial crisis in Europe: systematic literature review. BMJ, 354 , i4588. https://doi.org/10.1136/bmj.i4588 Paulus, A. F., F.; Sutherland, H. (2017). The design of fiscal consolidation measures in the European Union: distributional effects and implications for macro-economic recovery. Oxford Economic Papers, 69 (3), 632–654. https://doi.org/10.1093/oep/gpw054 PHW-NHS. (2020). Life expectancy and mortality in Wales . PHW-NHS-Trust. https://phw.nhs.wales/services-and-teams/observatory/data-and-analysis/life-expectancy-and-mortality-in-wales-2020/ Rajmil, L., & Fernandez de Sanmamed, M. J. (2019, May). Austerity Policies and Mortality Rates in European Countries, 2011-2015. Am J Public Health, 109 (5), 768-770. https://doi.org/10.2105/AJPH.2019.304997 Raleigh, V. S. (2019). Trends in life expectancy in eu and other OECD countries: Why are improvements slowing? (DELSA/HEA/WD/HWP(2019)1). (OECD Health Working Paper No. 108. https://www.oecd-ilibrary.org/docserver/223159ab-en.pdf?expires=1715622443&id=id&accname=guest&checksum=86F3E26040BFF8FD311BDDDC4AFFE8AC Ramsay, J., Minton, J., Fischbacher, C., Fenton, L., Kaye-Bardgett, M., Wyper, G. M. A., Richardson, E., & McCartney, G. (2020, Oct 7). How have changes in death by cause and age group contributed to the recent stalling of life expectancy gains in Scotland? Comparative decomposition analysis of mortality data, 2000-2002 to 2015-2017. BMJ Open, 10 (10), e036529. https://doi.org/10.1136/bmjopen-2019-036529 Regidor, E., Vallejo, F., Granados, J. A. T., Viciana-Fernandez, F. J., de la Fuente, L., & Barrio, G. (2016, Nov 26). Mortality decrease according to socioeconomic groups during the economic crisis in Spain: a cohort study of 36 million people. Lancet, 388 (10060), 2642-2652. https://doi.org/10.1016/S0140-6736(16)30446-9 Regidor, E. R., A. (2021). Crisis económica- 08, Covid-19 y salud: palabras cruzadas sobre la práctica científica . Grupo Editorial Círculo Rojo, S.L. https://editorialcirculorojo.com/autores/enrique-regidor/ Richardson, E., Taulbut, M., Robinson, M., Pulford, A., & McCartney, G. (2020, Oct 20). The contribution of changes to tax and social security to stalled life expectancy trends in Scotland: a modelling study. J Epidemiol Community Health, 75 (4), 365-370. https://doi.org/10.1136/jech-2020-214770 Seaman, R., Walsh, D., Beatty, C., McCartney, G., & Dundas, R. (2023, Nov 7). Social security cuts and life expectancy: a longitudinal analysis of local authorities in England, Scotland and Wales. J Epidemiol Community Health, 78 (2), 82-87. https://doi.org/10.1136/jech-2023-220328 Tanaka, T., Okamoto, S., & Canning, D. (2022, Feb 19). National Health Spending, Health-Care Resources, Service Utilization, and Health Outcomes. Am J Epidemiol, 191 (3), 386-396. https://doi.org/10.1093/aje/kwab179 Tapia Granados, J. A., & Ionides, E. L. (2017, Dec). Population health and the economy: Mortality and the Great Recession in Europe. Health Econ, 26 (12), e219-e235. https://doi.org/10.1002/hec.3495 Tapia Granados, J. A., & Rodriguez, J. M. (2015, Jul). Health, economic crisis, and austerity: A comparison of Greece, Finland and Iceland. Health Policy, 119 (7), 941-953. https://doi.org/10.1016/j.healthpol.2015.04.009 Toffolutti, V., & Suhrcke, M. (2019, May). Does austerity really kill? Econ Hum Biol, 33 , 211-223. https://doi.org/10.1016/j.ehb.2019.03.002 Walsh, D., Dundas, R., McCartney, G., Gibson, M., & Seaman, R. (2022, Dec). Bearing the burden of austerity: how do changing mortality rates in the UK compare between men and women? J Epidemiol Community Health, 76 (12), 1027-1033. https://doi.org/10.1136/jech-2022-219645 Walsh, D., Tod, E., McCartney, G., & Levin, K. A. (2022, Dec 14). How much of the stalled mortality trends in Scotland and England can be attributed to obesity? BMJ Open, 12 (12), e067310. https://doi.org/10.1136/bmjopen-2022-067310 Walsh, D., Wyper, G. M. A., & McCartney, G. (2022, Jun 6). Trends in healthy life expectancy in the age of austerity. J Epidemiol Community Health, 76 (8), 743-745. https://doi.org/10.1136/jech-2022-219011 Watkins, J., Wulaningsih, W., Da Zhou, C., Marshall, D. C., Sylianteng, G. D. C., Dela Rosa, P. G., Miguel, V. A., Raine, R., King, L. P., & Maruthappu, M. (2017, Nov 15). Effects of health and social care spending constraints on mortality in England: a time trend analysis. BMJ Open, 7 (11), e017722. https://doi.org/10.1136/bmjopen-2017-017722 WBG. (2024). The Climate Change Knowledge Portal. Country. Current climate. Trends within variability. Variability and trends of average mean surface air temperature across seasonal cycle, 1951-2020 https://climateknowledgeportal.worldbank.org/ WHO. (2024). The Global Health Observatory (GHO). Data. GHO. Indicators. Indicators index https://www.who.int/data/gho/data/indicators/indicators-index Wooldridge, J. (2001). Econometric Analysis of Cross Section and Panel Data . MIT Press. Xu, Z., FitzGerald, G., Guo, Y., Jalaludin, B., & Tong, S. (2016, Apr-May). Impact of heatwave on mortality under different heatwave definitions: A systematic review and meta-analysis. Environ Int, 89-90 , 193-203. https://doi.org/10.1016/j.envint.2016.02.007 Additional Declarations No competing interests reported. Supplementary Files SupplMaterial.docx Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 13 Mar, 2026 Reviews received at journal 09 Mar, 2026 Reviewers agreed at journal 09 Mar, 2026 Reviewers agreed at journal 09 Mar, 2026 Reviews received at journal 07 Mar, 2026 Reviewers agreed at journal 05 Feb, 2026 Reviewers agreed at journal 05 Feb, 2026 Reviews received at journal 30 Jan, 2026 Reviewers agreed at journal 22 Dec, 2025 Reviewers invited by journal 05 Dec, 2025 Editor assigned by journal 05 Dec, 2025 Editor invited by journal 05 Dec, 2025 Submission checks completed at journal 14 Nov, 2025 First submitted to journal 14 Nov, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7772794","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":557001065,"identity":"ddf51872-9660-4b86-ad7a-0626650940f0","order_by":0,"name":"Politi Julieta","email":"","orcid":"","institution":"Instituto de Salud Carlos III (ISCIII)","correspondingAuthor":false,"prefix":"","firstName":"Politi","middleName":"","lastName":"Julieta","suffix":""},{"id":557001066,"identity":"7622cd90-0797-46a9-9991-30418c40657c","order_by":1,"name":"Donat Marta","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAv0lEQVRIiWNgGAWjYJACCRDBzwwkEgwkiNciIdlMshaDA8Q6Srf9AOONjzvs6oyPsz/88KDAgoGfn4BmszMJzJYzzyRLmB3mMZYAOUxyRgIBLQcS2KR525hBWtjAfjG4QcBhZucfgLTUSxg3sz8Da7E/T8hhN8C2HJYwYGYwg9jCQMhhNx42W85sOy45A+oXHokbhLScTz5442NbNT9///GHH3/8qZPj7yfgMAYGxgYULg8h9aNgFIyCUTAKiAAAwPU5s8WZdosAAAAASUVORK5CYII=","orcid":"","institution":"Instituto de Salud Carlos III (ISCIII)","correspondingAuthor":true,"prefix":"","firstName":"Donat","middleName":"","lastName":"Marta","suffix":""},{"id":557001067,"identity":"2ccf6d32-9c3b-4fc8-a71c-92113386ad27","order_by":2,"name":"Guerras Juan Miguel","email":"","orcid":"","institution":"Instituto de Salud Carlos III (ISCIII)","correspondingAuthor":false,"prefix":"","firstName":"Guerras","middleName":"Juan","lastName":"Miguel","suffix":""},{"id":557001068,"identity":"132d84db-1867-4ebc-95ef-34ebfafcb29d","order_by":3,"name":"Herrero Lidia","email":"","orcid":"","institution":"Statistics National Institute","correspondingAuthor":false,"prefix":"","firstName":"Herrero","middleName":"","lastName":"Lidia","suffix":""},{"id":557001069,"identity":"57b27093-3af5-4818-9099-53364900088c","order_by":4,"name":"Barrio Gregorio","email":"","orcid":"","institution":"Instituto de Salud Carlos III (ISCIII)","correspondingAuthor":false,"prefix":"","firstName":"Barrio","middleName":"","lastName":"Gregorio","suffix":""},{"id":557001070,"identity":"f007ac2d-ef2d-4af5-910c-4e7810e5c246","order_by":5,"name":"Roberto Pastor-Barriuso","email":"","orcid":"","institution":"Instituto de Salud Carlos III (ISCIII)","correspondingAuthor":false,"prefix":"","firstName":"Roberto","middleName":"","lastName":"Pastor-Barriuso","suffix":""},{"id":557001071,"identity":"b1f31be1-b5b0-4901-84a1-9913cb7ddbcc","order_by":6,"name":"Belza María José","email":"","orcid":"","institution":"Instituto de Salud Carlos III (ISCIII)","correspondingAuthor":false,"prefix":"","firstName":"Belza","middleName":"María","lastName":"José","suffix":""},{"id":557001073,"identity":"f42713bf-ea35-44e4-8027-dc90b115aaae","order_by":7,"name":"Regidor Enrique","email":"","orcid":"","institution":"Department of Public Health \u0026 Maternal and Child Health","correspondingAuthor":false,"prefix":"","firstName":"Regidor","middleName":"","lastName":"Enrique","suffix":""}],"badges":[],"createdAt":"2025-10-03 10:53:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7772794/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7772794/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97818899,"identity":"5b96e8ee-2a70-459a-ab1d-9c0e560ec2f6","added_by":"auto","created_at":"2025-12-09 17:33:38","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":444147,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscriptrevisedv2.docx","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/4d11262ec1b66bfeca876627.docx"},{"id":97818888,"identity":"14c344ec-e189-4e7d-81ed-9c5ad38d3c29","added_by":"auto","created_at":"2025-12-09 17:33:38","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":10162,"visible":true,"origin":"","legend":"","description":"","filename":"31079ee2f00f45b98c1d8ffa20b8da1c.json","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/994c907fbdf37172322afb39.json"},{"id":97898618,"identity":"f829d2b7-ac9d-4c2c-93e2-6da1afec5c3e","added_by":"auto","created_at":"2025-12-10 15:39:22","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":68801,"visible":true,"origin":"","legend":"","description":"","filename":"SupplMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/040fc5d18d99dc3f21eb2d8a.docx"},{"id":97818895,"identity":"cb4b4f17-d344-4041-b822-f47f6c93fb91","added_by":"auto","created_at":"2025-12-09 17:33:38","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":194584,"visible":true,"origin":"","legend":"","description":"","filename":"31079ee2f00f45b98c1d8ffa20b8da1c1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/878f3ec77c4430b0f795430d.xml"},{"id":97898079,"identity":"7b18c2f7-4709-432b-b1dc-5ce9a41ab7f0","added_by":"auto","created_at":"2025-12-10 15:38:39","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":238871,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/78b5a8a4c0ccfe9509cabd19.jpeg"},{"id":97896233,"identity":"cf2a3cfc-ab5a-47ce-bf9a-ac98f34a49ca","added_by":"auto","created_at":"2025-12-10 15:36:11","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":43365,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/1c9abe7e6522e90ae84473ba.jpeg"},{"id":97898568,"identity":"348e5165-ce9a-446e-9d91-e6c37cfd1b44","added_by":"auto","created_at":"2025-12-10 15:39:18","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":448978,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/4c6acfb5df4b762e9d6dc6ab.jpeg"},{"id":97899010,"identity":"a9cc58d0-a3bb-40e5-a783-fcb5a55d7d94","added_by":"auto","created_at":"2025-12-10 15:40:41","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":54365,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/46cfd17120068e133e85bca4.png"},{"id":97818892,"identity":"d532e11c-a09b-4575-9ba3-4111b4179bcb","added_by":"auto","created_at":"2025-12-09 17:33:38","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9034,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/77c804bc04052b73435ea3b2.png"},{"id":97818894,"identity":"a5ec91c5-f107-46cf-8b3b-cc265735e719","added_by":"auto","created_at":"2025-12-09 17:33:38","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":104217,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/35748c9bc9ede5b4b1328db3.png"},{"id":97898127,"identity":"b167511b-d567-4752-b057-a03f4ab474fe","added_by":"auto","created_at":"2025-12-10 15:38:42","extension":"xml","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":188490,"visible":true,"origin":"","legend":"","description":"","filename":"31079ee2f00f45b98c1d8ffa20b8da1c1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/7c975acce394a68c9b654541.xml"},{"id":97818901,"identity":"8d7bcf0d-3cfc-4495-912b-f2b10d18666b","added_by":"auto","created_at":"2025-12-09 17:33:38","extension":"html","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":207638,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/a79b9f390f3b1ec4d3b81913.html"},{"id":97818889,"identity":"a6b48910-2d87-48bf-a590-a27b2a235f63","added_by":"auto","created_at":"2025-12-09 17:33:38","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":648244,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/263b3025720f60e53cb1c004.png"},{"id":97903247,"identity":"ede4ea57-2925-4ad6-9b8c-7cfdca67b893","added_by":"auto","created_at":"2025-12-10 15:54:47","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2192795,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/c9466f7a-48a6-409a-a954-5efaba06e92b.pdf"},{"id":97898835,"identity":"5e41ec3e-1272-45af-b3b9-1f0cc8c0c405","added_by":"auto","created_at":"2025-12-10 15:39:49","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":68801,"visible":true,"origin":"","legend":"","description":"","filename":"SupplMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-7772794/v1/900fe7858cbbe005513352a6.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Changes in government health spending and short-term mortality in four European countries in the 21st century","fulltext":[{"header":"1. INTRODUCTION","content":"\u003cp\u003eMortality reflects the main determinants of population health. After 2010, the downward trend in mortality slowed down in many high-income countries(Global Burden of Disease Greece, 2018; Hiam et al., 2018; PHW-NHS, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Raleigh, \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Ramsay et al., \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Walsh, Wyper, et al., 2022), something attributed to increased influenza activity or coldwaves(Murphy, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), obesity increase(Walsh, Tod, et al., \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), and mainly austerity policies (measures to improve the government's fiscal balance) implemented after the Great Recession. Numerous studies suggests a direct association between increased mortality and general austerity(Broadbent et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; McCartney et al., 2022; Rajmil \u0026amp; Fernandez de Sanmamed, \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Walsh, Wyper, et al., 2022). Among other mechanisms, austerity could be associated with mortality through cuts in government health and social spending, stress associated with unemployment and job insecurity, and changes in lifestyles or risk exposures due to a decrease in economic activity and disposable income.\u003c/p\u003e\u003cp\u003eSpecifically, cuts in government health spending could worsen population health by reducing the quantity or quality of health care and preventive interventions, for example by increasing waiting times for diagnosis and treatment or facilitating co-payments, which would especially affect vulnerable people(Doetsch et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Jacques \u0026amp; Noel, 2022). There is evidence quite consistent that health spending is inversely associated with suicide, or with child or maternal mortality(Antonakakis \u0026amp; Collins, 2015; Budhdeo et al., 2015). Regarding general or adult mortality, the evidence is much less consistent, finding an inverse association in some studies or for some causes of death(Budhdeo et al., 2015; Golinelli et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Hiam et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Leider et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Martin et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), but not in other studies or for other causes(Borra et al., 2020; Franklin et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Lippi et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Tanaka et al., \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eResearch on this topic has considerable methodological limitations, derived mainly from the difficulty of controlling potential confounders, delayed effects, and reverse causality(Murphy \u0026amp; Grundy, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The internal validity of studies on the health spending-mortality relationship can be compromised by multiple factors that should be controlled. Thus, studies in high-income countries have shown a direct association of mortality with increased heatwaves(Faurie et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Xu et al., \u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), coldwaves or influenza-activity(Murphy, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) and economic growth (Parmar et al., \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Regidor et al., \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Tapia Granados \u0026amp; Ionides, \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), and with decreased government social spending(Koltai et al., 2021; Loopstra et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Martin et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Richardson et al., \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Seaman et al., \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Watkins et al., \u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eOn the other hand, evidence shows that economic growth and government social spending are related to government health spending, so they would be clear confounders. In contrast, influenza activity, heatwaves, and coldwaves are phenomena whose occurrence would apparently not be associated with health spending. However, it is difficult to believe that they are not indirectly related through other omitted covariates. Thus, greater investment or spending in certain health or social infrastructures, or in preventive interventions (e.g., influenza vaccination) can mitigate he adverse effects of these covariates on mortality. Furthermore, these phenomena cause significant short-term variations in mortality, and adjusting for them allows us to reduce the residual variance of the model and obtain more precise association estimates. eFigure 1 represents the main relationships between health spending, mortality, and the adjustment covariates.\u003c/p\u003e\u003cp\u003ePrevious studies on the health expenditure-mortality relationship have controlled for these covariates to varying degrees, and not all of them have isolated within-country variations in annual health spending (which could be attributed to austerity policies) from between-country variations in mean health spending (probably due to structural factors other than austerity). Such studies have also often failed to assess the possible heterogeneity of the association between health spending and mortality by period, season, country, sex, or age. Consequently, adopting strategies to reduce these weaknesses can improve the validity of these studies' conclusions. We want to test the hypotheses that there is an inverse relationship between government health spending and mortality, and that cuts in government health spending may have contributed to the slowdown in declining mortality during 2011\u0026ndash;2019 in the United Kingdom (UK), Italy, Spain, and Greece. The selection of these four countries is justified for several reasons: a) They made important fiscal adjustments from 2010 to reduce government deficits and public debt. The most severe and prolonged measures were imposed on Greece by international creditors in order to receive financial bailouts. In addition to health spending, the measures included other public spending cuts (social protection and welfare, pensions, investments, other public services and even public salaries), tax increases, and structural reforms. These measures were accompanied by a decrease in GDP per capita, and caused intense social and political rejection(Darvas, 2014; Franklin et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; OECD, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Paulus, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). b) There is considerable debate about the health effects of austerity in these countries. For example, some authors point out that in Greece austerity was accompanied by a public health tragedy, with an increase in unmet medical needs, suicides or HIV infections among drug injectors(\"The Greek health crisis,\" 2015; Kentikelenis et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), while others conclude that such a tragedy has not existed(Tapia Granados \u0026amp; Rodriguez, 2015). c) Much scientific papers, mainly from the UK, attributes the slowdown in mortality decline after 2010 to austerity policies(Alexiou et al., 2021; Arca et al., 2020; Hiam et al., 2021; Hiam et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Koltai et al., 2021; Martin et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; McCartney et al., 2022; Richardson et al., \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Seaman et al., \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Walsh, Wyper, et al., 2022), including the reduction in government health spending(Arca et al., 2020; Hiam et al., 2021; Martin et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). d) It was quite difficult to obtain data for more countries stratified simultaneously by year, age group, sex, and month of the year, as well as annual data on some adjustment covariates such as heatwaves or coldwaves. The main objective was to characterize the association between government health spending and short-term mortality risk by period, season, sex, age, and country in UK, Italy, Spain, and Greece during 2002\u0026ndash;2019.\u003c/p\u003e"},{"header":"2. METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1. Study design, population and variables\u003c/h2\u003e\u003cp\u003eA balanced cross-sectional time-series design was applied. Data on all-cause deaths and population by country, year, month, sex, and five-year age group, and data on Gross Domestic Product (GDP), government's expenditures on health and social protection, influenza activity, heatwaves, and coldwaves by country and year were obtained from various sources. GDP was measured in 2015 US dollars per capita (real GDP per capita). Government\u0026rsquo;s expenditures were measured in thousands 2015 US dollars per capita. Finally, influenza activity, heatwaves and coldwaves were categorized, respectively, into four, three and two intensity levels. Details on variable definitions and data sources are shown in eTables 1 and eTable 2\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2. Data analysis\u003c/h2\u003e\u003cp\u003eInitially, annual age-standardized mortality rates per 100,000 person-years (py) and per capita healthcare expenditure were calculated by country, year, and season, specifically spring-summer (April to September) and autumn-winter (the remaining months). Subsequently, time trends in mortality and government health spending were estimated for 2002\u0026ndash;2009 and 2010\u0026ndash;2019, assuming a linear time trend within each period. Finally, two linear regression models of log-transformed age-standardized mortality rates on health spending, with covariate adjustment, were used in order to estimate the independent association within each country between the variation in health spending and mortality(Wooldridge, \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e2001\u003c/span\u003e), specifically the country-level fixed-effects (FE) model and the first-differences (FD) model. Both models used robust (sandwich) standard errors to account for intra-country correlation of rates and these errors were carried over into 95% confidence intervals. The country-level FE model included the level values ​​of variables during the current year and adjusted for country, time trend (as a restricted cubic spline), coldwave, influenza activity, heatwave, non-health social spending and real GDP per capita. However, models for spring-summer mortality were not adjusted for coldwave or influenza activity, and those for autumn-winter mortality were not adjusted for heatwave. Modeling time using cubic splines is a more realistic approximation than using the linear time trend because it better captures and controls the influence of unmeasured confounders that vary slowly over time, allowing for a better model fit and better estimates of the health spending regression coefficient (β). The country indicators removed the effect of measured and unmeasured time-invariant characteristics linked to the country. In this model, the β of health spending gives the multiplicative change in the annual mortality rate per one thousand US dollars increase in annual health spending. This implies that as spending increases by one thousand dollars, the mortality rate is multiplied by e\u003csup\u003eβ\u003c/sup\u003e. The FD model used the first difference between the value of variables during the current year and the previous year, including all the adjustment covariates of FE model except country. In this model, the β of health spending gives the multiplicative change in the ratio of interannual mortality rates per one thousand dollars increase in interannual difference of health spending. It provides an estimate of the association between interannual variations in health spending and mortality when unobserved time-constant variables and other covariates included in the model are kept constant. To facilitate interpretation, all \u003cem\u003eβs\u003c/em\u003e were transformed into percentage changes (PCs) as 100[exp(\u003cem\u003eβ\u003c/em\u003e) \u0026ndash; 1], estimating 95% confidence intervals (CIs) as 100[exp(\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;\u0026plusmn;\u0026thinsp;1.96\u003cem\u003eSE\u003c/em\u003e)\u0026thinsp;\u0026minus;\u0026thinsp;1], where \u003cem\u003eSE\u003c/em\u003e is the robust standard error of \u003cem\u003eβ\u003c/em\u003e. The results were stratified by country, sex, age (\u0026lt;\u0026thinsp;75 and \u0026ge;\u0026thinsp;75 years) and period (2002\u0026ndash;2010 and 2011\u0026ndash;2019). The stratification for 2002\u0026ndash;2010 and 2011\u0026ndash;2019 was decided because we suspected that there might be period heterogeneity in the results due to unmeasured time-varying factors (e.g., changes in health system productivity).\u003c/p\u003e\u003cp\u003eSome regression analyses were also implemented using lagged health and social expenditure data, specifically the mean of the current and previous year (0\u0026ndash;1 years lag mean). Incorporating lags ​​is justified primarily because the impact of health expenditures (e.g., hospital infrastructure, health prevention, professional training) can take some time to be reflected in mortality rates. Furthermore, such mean lags reduce the number of predictors (by avoiding including two annual lags), and the risk of multicollinearity due to correlation between both lags. Sensitivity analyses were performed to assess the robustness of the results, using other cumulative lag structures (e.g. 0-1-2 years lag mean) and measuring health expenditure in GDP percent. Models with sequential adjustment for the different covariates (step-up models) were also implemented to see the influence of each adjustment on the results. Some details on data analysis are given in eTable 3. All calculations were performed using Stata and R.\u003c/p\u003e\u003c/div\u003e"},{"header":"3. RESULTS","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\n\u003ch2\u003e3.1. Mortality and health spending by country\u003c/h2\u003e\n\u003cp\u003eThe evolution of mortality rates and health spending per capita are shown in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e. In general, there was a downward trend in annual, spring-summer and autumn-winter mortality in the four countries, although its slope softened from 2011 to 2019. Thus, during 2002\u0026ndash;2010 the mortality APCs were \u0026minus;\u0026thinsp;1.5% (Spain), -1.3% (UK), -1.2% (Greece), and \u0026minus;\u0026thinsp;1.1% (Italy), while during 2011\u0026ndash;2019, they were \u0026minus;\u0026thinsp;1.1% (Spain), -0.9% (UK), -0.8% (Italy), and \u0026minus;\u0026thinsp;0.5% (Greece). Meanwhile, health spending increased sharply during 2002\u0026ndash;2010 in the four countries, with APCs of 4.5% (Spain), 4.4% (UK), 4.4% (Greece) and 1.8% (Italy), and during 2011\u0026ndash;2019, its increase slowed considerably in the UK (1.0%) and Spain (0.5%) and decreased in Greece (-3.9%) and Italy (-0.9%). There were clear annual spikes in spring-summer mortality in 2003 in all countries, 2007 and 2012 in Greece and 2015 in Italy, and also in autumn-winter mortality in 2005 in Italy and Spain, 2015 in all countries, and 2012 and 2017 in Italy, Spain and Greece.\u0026nbsp;\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\n\u003ch2\u003e3.2. Association between health spending and mortality risk\u003c/h2\u003e\n\u003cp\u003eFigure \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e shows that coinciding with the strong growth in health spending in 2002\u0026ndash;2010, mortality fell considerably, while subsequently spending moderated its growth or fell and the downward trend in mortality slowed in all four countries. Focusing on the APCs, it is observed that during 2011\u0026ndash;2019 the largest cuts in health spending and the greatest slowdown in mortality occurred in Greece.\u003c/p\u003e\n\u003cp\u003eIn the four-country pooled series, lagged and non-lagged FE and FD regression models showed an inverse association between health spending and mortality during 2002\u0026ndash;2010 (i.e. mortality decreases as health spending increases), stronger for autumn-winter than spring-summer mortality. However, during 2011\u0026ndash;2019, they showed a direct association stronger for spring-summer mortality (Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). The association was inverse during 2002\u0026ndash;2010 and direct during 2011\u0026ndash;2019, both in men and women and people aged\u0026thinsp;\u0026lt;\u0026thinsp;75 and \u0026ge;\u0026thinsp;75, being stronger in people aged\u0026thinsp;\u0026ge;\u0026thinsp;75 than \u0026lt;\u0026thinsp;75 years. Furthermore, during 2002\u0026ndash;2010, in FE models, the association was stronger in men than in women, while in FD models, the opposite occurred (Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n\u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n\u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n\u003ctable id=\"Tab1\" border=\"1\"\u003e\u003ccaption\u003e\n\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n\u003cdiv class=\"CaptionContent\"\u003e\n\u003cp\u003eAssociation between age-standardized mortality rates and annual government health spending from regression models by model type, period and season, 2002\u0026ndash;2019\u003c/p\u003e\n\u003c/div\u003e\n\u003c/caption\u003e\n\u003cthead\u003e\n\u003ctr\u003e\n\u003cth rowspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eN\u003csup\u003e5\u003c/sup\u003e\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003eAnnual mortality\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003eSpring-summer mortality\u003csup\u003e6\u003c/sup\u003e\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003eAutumn-winter mortality\u003csup\u003e7\u003c/sup\u003e\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ePC\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e95%CI\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ePC\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e95%CI\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ePC\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e95%CI\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eNon lagged regression model\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003c/tr\u003e\n\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eCountry-level fixed-effect (FE) model\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2002\u0026ndash;2010\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-12.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-15.7, -8.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-10.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-15.2, -5.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-15.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-21.5, -9.0\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2011\u0026ndash;2019\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e8.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e4.4, 13.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e13.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.7, 20.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e7.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e4.3, 9.8\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eTotal\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e72\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-8.0, -2.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-3.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.7, -0.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-10.6, -4.2\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eFirst-differences (FD) model\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2002\u0026ndash;2010\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-12.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-17.6, -7.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-20.2, 7.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-15.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-22.8, -6.8\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2011\u0026ndash;2019\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e5.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.9, 9.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e11.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e4.9, 18.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e4.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-1.1, 9.8\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eTotal\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e72\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-9.3, -1.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-1.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-8.8, 5.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-13.5, -1.6\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eLagged regression model (0\u0026ndash;1 year lag mean)\u003c/strong\u003e \u003csup\u003e\u003cstrong\u003e4\u003c/strong\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eFE model\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2002\u0026ndash;2010\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-12,5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-16.3, -8.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-11.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-13.2, -9.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-15.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-21.1, -10.3\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2011\u0026ndash;2019\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e7.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2.5, 12.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e4.8, 14.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e5.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-1.0, 13.4\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eTotal\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e72\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-4,4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-6.3, -2.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-2.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-4.4, -0.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-6.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-8.5, -5.1\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eFD model\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2002\u0026ndash;2010\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-9.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-20.1, 3.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-16.7, 6.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-10.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-25.5, 8.2\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2011\u0026ndash;2019\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-2.7, 17.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e10.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-2.8, 25.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-3.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-16.6, 11.0\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eTotal\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e72\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.3, -2.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-0.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-2.0, 1.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-10.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-13.1, -7.0\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003ctfoot\u003e\n\u003ctr\u003e\n\u003ctd colspan=\"11\"\u003e\u003cstrong\u003ePC\u003c/strong\u003e: Percent change in annual mortality rate (FE model) or ratio of interannual mortality rates (FD model) per one thousand US dollars increase in annual health spending (FE model) or interanual difference in health spending (FD model). \u003cstrong\u003e95%CI\u003c/strong\u003e: 95% confidence interval of PC. \u003csup\u003e\u003cstrong\u003e1\u003c/strong\u003e\u003c/sup\u003e Health spending and social spending in the regression model correspond to the same year as mortality and the rest of the independent variables. \u003csup\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sup\u003e Multiple linear regression of logarithm of age-standardized mortality rates on health spending in thousands of 2015 US dollars ($) per capita, adjusting for country, year as a restricted cubic spline and other covariates, as heatwave, coldwave, influenza activity, real GDP per capita and social spending per capita. \u003csup\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sup\u003e Multiple linear regression of first differences between the logarithm of age-standardized mortality rate in a given year and country, and its logarithm in the previous year on first differences of health spending, adjusting for first-differences in year as a restricted cubic spline and first-differences in the covariates heatwave, coldwave, influenza activity, real GDP per capita, and social spending in thousands of 2015 US $ per capita. \u003csup\u003e\u003cstrong\u003e4\u003c/strong\u003e\u003c/sup\u003e Health spending and social spending in the regression model correspond to the mean of the current and previous year's values. \u003csup\u003e\u003cstrong\u003e5\u003c/strong\u003e\u003c/sup\u003e It refers to the number of points in the regression models. \u003csup\u003e\u003cstrong\u003e6\u003c/strong\u003e\u003c/sup\u003e It corresponds to mortality between April and September, both included. For this outcome, results of the regression models were not adjusted for influenza activity or coldwave. \u003csup\u003e\u003cstrong\u003e7\u003c/strong\u003e\u003c/sup\u003e It corresponds to mortality during the remaining months of the year. For this outcome, results of the regression models were not adjusted for heatwave.\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tfoot\u003e\n\u003c/table\u003e\n\u003c/div\u003e\n\u003cdiv class=\"gridtable\"\u003e\n\u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n\u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n\u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n\u003ctable id=\"Tab2\" border=\"1\"\u003e\u003ccaption\u003e\n\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n\u003cdiv class=\"CaptionContent\"\u003e\n\u003cp\u003e\u003cstrong\u003eAssociation between annual age-standardized mortality rates and annual government health spending from regression models by model type and period, in men and women and in two age group, 2002\u0026ndash;2019.\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003c/div\u003e\n\u003c/caption\u003e\n\u003cthead\u003e\n\u003ctr\u003e\n\u003cth rowspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eN\u003csup\u003e5\u003c/sup\u003e\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003eMen\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003eWomen\u003c/p\u003e\n\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"3\" align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;75 years\u003c/p\u003e\n\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"3\" align=\"left\"\u003e\n\u003cp\u003e\u0026ge;\u0026thinsp;75 years\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ePC\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e95%CI\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ePC\u003c/p\u003e\n\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e95%CI\u003c/p\u003e\n\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ePC\u003c/p\u003e\n\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e95%CI\u003c/p\u003e\n\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ePC\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e95%CI\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eNon lagged regression model\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003c/tr\u003e\n\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eCountry-level fixed-effect (FE) model\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2002\u0026ndash;2010\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-10.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-15.0, -6.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-13.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-17.8, -9.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-9.1, -6.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-13.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-18.9, -8.1\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2011\u0026ndash;2019\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e8.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2.1, 14.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e6.4, 13.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e0.0, 5.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e11.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e5.7, 18.4\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eTotal\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e72\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-9.5, -0.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-7.6, -2.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-8.3, -2.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-4.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-9.8, 0.9\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eFirst-differences (FD) model\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2002\u0026ndash;2010\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-14.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-19.7, -8.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-11.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-16.5, -6.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-8.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-11.9, -5.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-14.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-21.0, -6.8\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2011\u0026ndash;2019\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.3, 11.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e5.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e3.0, 8.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e0.2, 5.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.2, 12.9\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eTotal\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e72\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-11.6, 0.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-7.5, -3.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-6.2, -3.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-11.7, 0.3\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eLagged regression model (0\u0026ndash;1 year lag mean)\u003c/strong\u003e \u003csup\u003e\u003cstrong\u003e4\u003c/strong\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"3\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"3\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eFE model\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2002\u0026ndash;2010\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-10.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-15.5, -5.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-15.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-18.5, -12.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-8.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-9.4, -6.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-14.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-19.4, -8.9\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2011\u0026ndash;2019\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e7.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.7, 12.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e7.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e3.0, 11.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-0.8, 2.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e10.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.6, 17.9\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eTotal\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e72\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-4.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.9, -1.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-4.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-6.0, -3.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-8.4, -1.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-3.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.9, 0.5\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eFD model\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2002\u0026ndash;2010\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-10.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-20.6, 1.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-8.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-21.2, 6.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-6.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-9.4, -2.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-10.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-25.0, 7.1\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2011\u0026ndash;2019\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e36\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e7.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.5, 15.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-5.2, 19.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-3.3, 11.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e8.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-2.2, 20.1\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eTotal\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e72\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-4.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.9, -0.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-10.0, -1.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-4.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e-9.3, 0.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd colspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.9, -2.4\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cstrong\u003ePC\u003c/strong\u003e: Percent change in annual mortality rate (FE model) or ratio of interannual mortality rates (FD model) per one thousand US dollars increase in annual health spending (FE model) or interanual difference in health spending (FD model). \u003cstrong\u003e95%CI\u003c/strong\u003e: 95% confidence interval of PC. \u003csup\u003e\u003cstrong\u003e1\u003c/strong\u003e\u003c/sup\u003e Health spending and social spending in the regression model correspond to the same year as mortality and the rest of the independent variables. \u003csup\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sup\u003e Multiple linear regression of logarithm of age-standardized mortality rates on government health spending in thousands of 2015 US dollars ($) per capita, adjusting for country, year as a restricted cubic spline and other covariates, as heatwave, coldwave, influenza activity, real GDP per capita and social spending per capita. \u003csup\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sup\u003e Multiple linear regression of first differences between the logarithm of age-standardized mortality rate in a given year and country, and its logarithm in the previous year on first differences of health spending, adjusting for first-differences in year as a restricted cubic spline and first-differences in the covariates heatwave, coldwave, influenza activity, real GDP per capita and social spending in thousands of 2015 US $ per capita. \u003csup\u003e\u003cstrong\u003e4\u003c/strong\u003e\u003c/sup\u003e Health spending and social spending in the regression model correspond to the mean of the current and previous year's values. \u003csup\u003e\u003cstrong\u003e5\u003c/strong\u003e\u003c/sup\u003e It refers to the number of points in the regression models.\u003c/p\u003e\n\u003cp\u003eTable\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e shows the association between health spending and mortality separately for the UK and the other three countries from FE models adjusted by different sets of covariates. Although the analysis is not very robust for the UK due to the limited number of degrees of freedom, there was an inverse association in both country groups during 2002\u0026ndash;2010, which appears to be much stronger in the UK. Meanwhile, during 2011\u0026ndash;2019, the PC sign changed in the three-country group, becoming positive, something that was also observed in the fully adjusted FD model (not tabulated), where the PC changed from \u0026minus;\u0026thinsp;9.5% (95%CI:-13.7, -5.0%) during 2002\u0026ndash;2010 to 6.1% (95%CI:3.9\u0026ndash;8.3%) during 2011\u0026ndash;2019. In the UK, however, according to Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e the association appears to remain inverse.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n\u003ctable id=\"Tab3\" border=\"1\"\u003e\u003ccaption\u003e\n\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\n\u003cdiv class=\"CaptionContent\"\u003e\n\u003cp\u003eAssociation between annual age-standardized mortality rates and annual government health spending from country-level fixed-effects models adjusted for various set of covariates by period and country group, 2002\u0026ndash;2019.\u003c/p\u003e\n\u003c/div\u003e\n\u003c/caption\u003e\n\u003cthead\u003e\n\u003ctr\u003e\n\u003cth rowspan=\"2\" align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"3\" align=\"left\"\u003e\n\u003cp\u003e2002\u0026ndash;2010\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"3\" align=\"left\"\u003e\n\u003cp\u003e2011\u0026ndash;2019\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth colspan=\"3\" align=\"left\"\u003e\n\u003cp\u003eTotal\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eN\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ePC\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e95%CI\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eN\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ePC\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e95%CI\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eN\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ePC\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e95%CI\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eUK\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003c/tr\u003e\n\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 1\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-18.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-40.8, 11.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-24.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-53.1, 21.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e18\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-18.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-37.0, 4.6\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 2\u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-28.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-48.1, -1.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-1.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-47.8, 87.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e18\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-18.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-38.3, 7.0\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 3\u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-23.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-41.5, 0.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-27.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-60.8, 33.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e18\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-19.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-37.9, 3.5\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 4\u003csup\u003e5\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-13.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-55.1, 68.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-16.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-60.3, 76.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e18\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-18.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-37.5, 6.0\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 5\u003csup\u003e6\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-11.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-39.8, 29.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-32.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-57.5, 6.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e18\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-18.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-36.5, 5.9\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 6\u003csup\u003e7\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-26.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-76.9, 131.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-26.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-57.0, 24.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e18\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-8.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-30.9, 20.3\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eOther three countries\u003c/strong\u003e\u003csup\u003e8\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 1\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-0.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-9.3, 0.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-0.6, 7.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e54\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-9.6, -1.3\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 2\u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-6.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-9.8, -1.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-0.7,6.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e54\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.7, -2.7\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 3\u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.7, 9.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-1.0, 7.9\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e54\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-9.5, -1.7\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 4\u003csup\u003e5\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-0.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-11.5, 12.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.8, 5.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e54\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-5.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-10.4, -0.7\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 5\u003csup\u003e6\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-8.0, 11.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.1, 7.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e54\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-3.1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-6.0, -0.1\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 6\u003csup\u003e7\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-3.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-11.3, 5.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e9.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.5, 20.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e54\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-9.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-15.5, -3.0\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eModel 7\u003csup\u003e9\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-9.8, -5.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e7.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e4.2, 10.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e54\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-7.0\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-10.7, -3.2\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003ctfoot\u003e\n\u003ctr\u003e\n\u003ctd colspan=\"12\"\u003e\u003cstrong\u003ePC\u003c/strong\u003e: Percent change in annual mortality rate per one thousand US dollars increase in annual health spending. \u003cstrong\u003e95%CI\u003c/strong\u003e: 95% confidence interval of PC. \u003csup\u003e\u003cstrong\u003e1\u003c/strong\u003e\u003c/sup\u003e Includes England \u0026amp; Wales and Scotland for mortality and influenza activity and the entire United Kingdom for the rest of the variables. \u003csup\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sup\u003e Model 1: Fixed-effects multiple linear regression models of logarithm of age-standardized mortality rates on health spending in thousands of 2015 US dollars per capita, adjusting for country and year as a restricted cubic spline. \u003csup\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sup\u003e Model 2: Same as model 1, but additionally adjusting for social spending in thousands of 2015 US dollars per capita (SS). \u003csup\u003e\u003cstrong\u003e4\u003c/strong\u003e\u003c/sup\u003e Model 3: Same as model 1, but adjusting for heatwaves instead of SS. Model 4: Same as model 1, but adjusting for coldwaves instead of SS. \u003csup\u003e\u003cstrong\u003e6\u003c/strong\u003e\u003c/sup\u003e Model 5: Same as model 1, but adjusting for influenza activity instead of SS. \u003csup\u003e\u003cstrong\u003e7\u003c/strong\u003e\u003c/sup\u003e Model 6: Same as model 1, but adjusting for Gross Domestic Product per capita instead SS. \u003csup\u003e\u003cstrong\u003e8\u003c/strong\u003e\u003c/sup\u003e Includes Italy, Spain and Greece analysed together and applying country fixed-effects. \u003csup\u003e\u003cstrong\u003e9\u003c/strong\u003e\u003c/sup\u003e It refers to the number of points in regression models. \u003csup\u003e\u003cstrong\u003e9\u003c/strong\u003e\u003c/sup\u003e Model 7: Full model adjusted for all variables considered. This model was not calculated for the UK due to a lack of degrees of freedom.\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tfoot\u003e\n\u003c/table\u003e\n\u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n\u003ch2\u003e3.3. Sensitivity analyses\u003c/h2\u003e\n\u003cp\u003eThe association between health spending and annual mortality was quite similar when spending was measured in GDP percent (eTable 4) and when various accumulative lag structures of health and social spending were used (eTable 5). The forward entry models showed a considerable increase in the strength of the inverse association between health spending and mortality during 2002\u0026ndash;2010 when introducing social spending as an adjustment covariate (eTable 6), reflecting a direct association of this covariate with mortality during that period.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"4. DISCUSSION","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e4.1. Main findings\u003c/h2\u003e\u003cp\u003eThe trend analysis shows that the strong upward trend in government health spending was interrupted around 2010 (with a slowdown in the rise or even a decline, depending on the country), while mortality continued to decline, albeit at a slower pace. The regression analysis, after adjusting for heatwaves, coldwaves, influenza activity, economic growth and government social spending, does not show an inverse association between health spending and short-term mortality in Italy, Spain and Greece during 2011\u0026ndash;2019. However, a fairly strong inverse association was identified in 2002\u0026ndash;2010 in all four countries, which appears to have remained in the UK during 2011\u0026ndash;2019. Some heterogeneity in such associations was detected according to age and season. Main results remain introducing cumulative lagged values ​​of health and social spending (up to 3 years).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e4.2. On the explanatory factors of the slowdown in mortality decline\u003c/h2\u003e\u003cp\u003eWe had hypothesized that cuts in government health expenditure may have contributed to the slowdown in declining mortality in the countries analysed during 2011\u0026ndash;2019. The results would support the hypothesis for the UK, but not for the other three countries, where such cuts do not seem to be a relevant explanatory factor, requiring other time-varying explanatory factors to be considered(Jasilionis et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Raleigh, \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). As noted, we controled some of the time-varying confounders, such as influenza activity, coldwaves, heatwaves, economic growth or government social expenditure. Although we identified a positive association between influenza activity and mortality, as in previous studies(Murphy, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), no increase in influenza was detected in 2011\u0026ndash;2019, that could explain the slowdown in declining mortality, something previously ruled out for the UK and USA(Hiam et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Murphy \u0026amp; Grundy, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The same is true for coldwaves. Regarding government social spending, although previous research in high-income countries shows an inverse association with mortality (Koltai et al., 2021; Loopstra et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Martin et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Richardson et al., \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Seaman et al., \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Watkins et al., \u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), we did not find such association in Italy, Spain and Greece during 2011\u0026ndash;2019. Therefore, it does not seem to be a relevant factor in explaining the slowdown in mortality decline. On the contrary, global warming and heatwaves(Global Burden of Disease Greece, 2018; Lee et al., 2019) could help partially explain such slowdown. Thus, in our work, a clear positive association between heatwaves and mortality was found in Italy, Spain and Greece, stronger during 2011\u0026ndash;2019 than previously. Moreover, from international data(WBG, \u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) it is estimated that between 2002\u0026ndash;2010 and 2011\u0026ndash;2019, the average maximum air surface temperature across the quarter June-September increased 0.5\u0026ndash;0.6 degrees Celsius in Italy, Spain and Greece.\u003c/p\u003e\u003cp\u003eOther time-varying unmeasured factors may have contributed to confounding the association between health spending and mortality and explain the slowdown in declining mortality during 2011\u0026ndash;2019, such as obesity, alcohol and drug use, general austerity, changes in the productivity of health systems or even changes in the frequency or lethality of diseases and injuries. Although obesity has been proposed as a possible confounder(Walsh, Tod, et al., \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003csup\u003e,\u003c/sup\u003e(Murphy \u0026amp; Grundy, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), is unlikely to have been a relevant factor because a clear increase in its prevalence during 2008\u0026ndash;2020 in the countries studied has not been identified(WHO, \u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). However, some doubt remains, because obesity could have a very lagged impact on mortality and reports on obesity become less reliable as the response rate decreases(Walsh, Tod, et al., \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Per capita alcohol consumption may have contributed to the slowdown in mortality decline because a downward trend was observed during 2002\u0026ndash;2010, which, except in Greece, slowed or reversed starting in 2011.Thus, the annual decline in UK, Italy and Spain went from \u0026minus;\u0026thinsp;1.3%, -3.7% and \u0026minus;\u0026thinsp;2.4%, respectively, in 2002\u0026ndash;2010 to -0.2%, 1.2% and 1.6% in 2011\u0026ndash;2019(WHO, \u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). There are also reports from the UK and Spain that during the second decade of the 21st century the downward trend in alcohol-related mortality stabilised or reversed(Augarde et al., 2022; Donat et al., 2023). Regarding drug use, some UK research finds evidence of an inverse association between government spending on welfare and drug-related mortality(Friebel et al., 2022; Koltai et al., 2021), which can be explained by the increase in drug use due to coping with the stress attributable to austerity or the decrease in investments in prevention and harm reduction programs. However, the proportion of drug-related deaths in the analysed countries is low, so it is unlikely that drug use is a relevant explanatory factor for the slowdown in declining overall mortality. Previous research in high-income countries has found a direct association between mortality and general austerity (Alexiou et al., 2021; Golinelli et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; McCartney et al., 2022; Rajmil \u0026amp; Fernandez de Sanmamed, \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Walsh, Wyper, et al., 2022). However, we did not find this association in Italy, Spain and Greece during 2011\u0026ndash;2019, using the cyclically adjusted fiscal balance as an indicator (data not shown). Finally, another candidate explanatory factor to explain the slowdown in declining mortality is the increase in productivity of health systems due to changes in spending distribution, shifts in care models, technological innovations, changes in health services management or greater workforce's dedication(Garcia-Escribano, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Golinelli et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Hern\u0026aacute;ndez, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Jacques \u0026amp; Noel, 2022; ONS, \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). It is not easy to find evidence to clarify this issue. However, a recent study concludes that during 2012\u0026ndash;2015 the health systems productivity increased in most European countries, including Spain, Italy and Greece, but not in the UK, where it decreased. In addition, this increased productivity was positively associated with increased efficiency in health systems management, but not in health technology level(Gomez-Gallego, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). However, OECD data on diagnostic examinations with computerized axial tomography and magnetic resonance, show that mainly since 2016 the technological level increased considerably in all analysed countries, especially in Spain and the UK. Another study draws similar conclusions for Spain(Regidor, \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003e4.3. On the association between health spending and mortality\u003c/h2\u003e\u003cp\u003eThe inverse association between government health spending and short-term mortality during 2002\u0026ndash;2010 in all four countries and in UK during 2011\u0026ndash;2019 is consistent with the results of numerous previous studies (often from the UK)(Budhdeo et al., 2015; Hiam et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Leider et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Mackenbach et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Martin et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Mays \u0026amp; Smith, 2011; Watkins et al., \u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). This apparently supports the conclusion of some authors that the decrease in health spending worsen population health and vice versa. However, this does not seem too consistent with our finding that during 2011\u0026ndash;2019 mortality continued to decline at a good pace (between \u0026minus;\u0026thinsp;0.5% and \u0026minus;\u0026thinsp;1.1% annually) in all four countries, despite a radical change in health spending trends after 2010 (from a strong upward trend to a decline or slowdown in the rise, depending on the country). And it is also not consistent with the unexpected change in the direction of the health spending-mortality association, from inverse to direct in Italy, Spain and Greece during 2011\u0026ndash;2019, a period that includes years of strong austerity (2011\u0026ndash;2015) and declining government health spending. Taken together, these findings suggest that at the levels of government health spending analysed, the hypothesized inverse association between health spending and mortality in certain contexts could not exist or be weak and easily offset or reversed by unmeasured or poorly measured time-varying factors. These confounders may include increased productivity of health systems or other factors mentioned above. It should be noted that previous studies in high-income countries have also failed to find a clear inverse association(Borra et al., 2020; Franklin et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Lippi et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Tanaka et al., \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003e4.4. Heterogeneity in health spending-mortality association by age, sex and season of the year\u003c/h2\u003e\u003cp\u003eThe study results show a stronger association between health expenditure and mortality among older people (\u0026ge;\u0026thinsp;75 years), whether the association is inverse (2002\u0026ndash;2010) or direct (2011\u0026ndash;2019 in Italy, Spain and Greece). Some previous studies(Hiam et al., 2018; Loopstra et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Watkins et al., \u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), although not all(Toffolutti \u0026amp; Suhrcke, \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), have suggested an inverse association between general government austerity and mortality in older people, where health or social spending and mortality are concentrated(Kinge et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Our results do not show clear differences by sex in the association studied. There are hardly any studies on this topic, but a British study found some evidence that among more deprived populations, female mortality rates have worsened to a greater degree during the years of general austerity(Walsh, Dundas, et al., 2022).\u003c/p\u003e\u003cp\u003eFinally, our results suggest that when the association health spending-mortality was inverse (2002\u0026ndash;2010) it was stronger in autumn-winter, while when it was positive (2011\u0026ndash;2019 in Italy, Spain and Greece) the association was stronger in spring-summer. It is difficult to explain these differences and, to our knowledge, there are no previous studies focusing on this topic, although results suggest that health spending would preferentially affect treatable or preventable conditions that are more prevalent or that tend to worsen in autumn-winter.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e4.5. Strengths and limitations\u003c/h2\u003e\u003cp\u003eThe main strengths are the absorption through regression models of between-country differences in mean health spending to isolate the association between variations in health spending and mortality within each country. A strategy that would make it possible to attribute changes in mortality to changes in health spending with greater validity. Another strength is a more complete control strategy of possible time-varying confounders than in previous studies. Main limitations include insufficient quality of data on extreme temperatures and influenza and a short time series, which limits the statistical power for some stratified analysis, for example, by country and period. Although a few potential confounders were controlled, residual confounding possibly persists. Further control was not possible due to sample size limitations, difficulties in obtaining data, and the need to avoid overfitting. For example, adjusting for the unemployment rate would have meant overfitting because its changes are strongly associated with increase in GDP per capita. Private health expenditure is very unlikely to bias the results because it represents a modest proportion (20\u0026ndash;37%) of total health expenditure in the analysed countries with a very small temporal variation within countries (although during the financial crisis, its proportion decreased slightly). Finally, it is not convenient to adjust for possible mediators between health spending and mortality (e.g., waiting time for diagnosis or treatment, or practicing physicians per capita), because it could eliminate inadequately the health spending effect that passes through the mediator. There could also be numerator-denominator biases related to migration and tourism (probably small), as well as harvesting phenomena or mortality displacements, given that during economic crises (generally associated with decreased mortality) some deaths are delayed, while the opposite usually happens during periods of extreme temperatures or increased influenza activity. Our design does not allow inferring causal relationships. However, it seems unlikely that the slowdown in declining mortality caused the observed changes in health spending (e.g. stagnation or decline), because the latter generally began before the change in mortality trend. Furthermore, the associations remained after introducing cumulative lagged values ​​of health spending.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e4.6. Policy and research implications\u003c/h2\u003e\u003cp\u003eThe results during 2011\u0026ndash;2019 in three of four high-income countries analysed, suggest that the nature of the association health spending-mortality is very dependent on time-varying omitted covariates linked to the period and country. Therefore, as long as those omitted covariates that mediate or confound the association are not clearly identified, the generalisability of the results on such association is limited. To obtain more valid results, and avoid the influence of structural factors, it is advisable to focus only on within-country temporal variations in health spending and mortality. It is also very useful to have large time series with many geographical units and years, which allows controlling multiple time-varying confounders, including their possible delayed effects. The observation that stagnation or decrease in health spending is not always accompanied by an increase in mortality suggests that health system decision-makers, in addition to ensuring sufficient funding, should focus on improving other aspects (e.g., organizational or management) that increase the health systems' productivity.\u003c/p\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eDECLARATION OF INTEREST STATEMENT:\u0026nbsp;\u003c/strong\u003eThe authors have had no financial or personal relationships with other people or organizations that could inappropriately influence or bias their work.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eACKNOWLEDGEMENTS:\u0026nbsp;\u003c/strong\u003eThe authors would like to thank the statistical offices of the countries included in the study, specifically the Office for National Statistics (England \u0026amp; Wales), National Records of Scotland/Scotlands People (Scotland), Istituto Nazionale di Statistica -ISTAT- (Italy), Instituto Nacional de Estadística -INE- (Spain) and the Hellenic Statistical Authority (ELSTAT) for sending mortality data or for facilitating access to data.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eFUNDING\u003c/strong\u003e: This work was supported by a grant from the Carlos III Health Institute and the European Regional Development Fund (PI16/00455). The funding organization did not intervene or interfere in the design and conduction of the study; collection, management, analysis, and interpretation of the data; nor in the preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAUTHOR CONTRIBUTION\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization and study design: Julieta Politi, Marta Donat, Gregorio Barrio, Enrique Regidor, Mar\u0026iacute;a Jos\u0026eacute; Belza.\u003c/p\u003e\n\u003cp\u003eData acquisition and management: Julieta Politi, Lidia Herrero, Enrique Regidor, Juan Miguel Guerras.\u003c/p\u003e\n\u003cp\u003eData analysis and interpretation: Julieta Politi, Marta Donat, Gregorio Barrio, Roberto Pastor-Barriuso, Juan Miguel Guerras.\u003c/p\u003e\n\u003cp\u003eDrafting the manuscript: Julieta Politi, Marta Donat, Juan Miguel Guerras.\u003c/p\u003e\n\u003cp\u003eCritical revision for important intellectual content: All authors.\u003c/p\u003e\n\u003cp\u003eSupervision and senior authorship: Mar\u0026iacute;a Jos\u0026eacute; Belza, Enrique Regidor.\u003c/p\u003e\n\u003cp\u003eAll authors read and approved the final version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eETHICAL APPROVAL\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Review Committee of the Carlos III Health Institute approved the study proposal. Ethics Committee approval was not deemed necessary because the mortality microdata used, although comprised of individual records, do not contain variables that could directly or indirectly facilitate personal identification. The national statistical offices adopts the necessary logical, physical, and administrative measures to ensure the effective protection of confidential data, from collection to publication. In accordance with the Public Statistical Service and Regulation (EC) No 223/2009, it does not disseminate personal data from any source. Microdata files are distributed anonymously, taking precautions to avoid any indirect identification. For example, the deceased\u0026apos;s name, surname, Identification Document, address, death registration information, date of birth, or codes for municipalities with municipalities with a small population size are not provided. This study was based on aggregated, anonymized population-level data provided by national statistical offices (ONS, National Records of Scotland, ISTAT, INE, and ELSTAT). As the analyses did not involve identifiable individual-level data, approval from a Research Ethics Committee was not required according to national regulations.\u003c/del\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDATA AVAILABILITY\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated and/or analysed during the current study are available from the corresponding author upon reasonable request, for academic and non-commercial purposes. Mortality and population data were obtained from the national statistical offices listed above in aggregated and anonymized form, under data use agreements that do not permit public sharing. \u003c/ins\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAlexiou, A., Fahy, K., Mason, K., Bennett, D., Brown, H., Bambra, C., Taylor-Robinson, D., \u0026amp; Barr, B. (2021, Sep). Local government funding and life expectancy in England: a longitudinal ecological study. \u003cem\u003eLancet Public Health, 6\u003c/em\u003e(9), e641-e647. https://doi.org/10.1016/S2468-2667(21)00110-9\u003c/li\u003e\n\u003cli\u003eAntonakakis, N., \u0026amp; Collins, A. (2015, Nov). The impact of fiscal austerity on suicide mortality: Evidence across the 'Eurozone periphery'. \u003cem\u003eSoc Sci Med, 145\u003c/em\u003e, 63-78.\u0026nbsp;\u0026nbsp;https://doi.org/10.1016/j.socscimed.2015.09.033\u003c/li\u003e\n\u003cli\u003eArca, E., Principe, F., \u0026amp; Van Doorslaer, E. (2020, Dec). Death by austerity? The impact of cost containment on avoidable mortality in Italy. \u003cem\u003eHealth Econ, 29\u003c/em\u003e(12), 1500-1516. https://doi.org/10.1002/hec.4147\u003c/li\u003e\n\u003cli\u003eAugarde, E., Gunnell, D., Mars, B., \u0026amp; Hickman, M. (2022, Jun). An ecological study of temporal trends in 'deaths of despair' in England and Wales. \u003cem\u003eSoc Psychiatry Psychiatr Epidemiol, 57\u003c/em\u003e(6), 1135-1144. https://doi.org/10.1007/s00127-022-02251-9\u003c/li\u003e\n\u003cli\u003eBorra, C., Pons-Pons, J., \u0026amp; Vilar-Rodriguez, M. (2020, Apr). Austerity, healthcare provision, and health outcomes in Spain. \u003cem\u003eEur J Health Econ, 21\u003c/em\u003e(3), 409-423. https://doi.org/10.1007/s10198-019-01141-3\u003c/li\u003e\n\u003cli\u003eBroadbent, P., Walsh, D., Katikireddi, S. V., Gallagher, C., Dundas, R., \u0026amp; McCartney, G. (2024, Oct). Is Austerity Responsible for the Stalled Mortality Trends Across Many High-Income Countries? A Systematic Review. \u003cem\u003eInt J Soc Determinants Health Health Serv, 54\u003c/em\u003e(4), 362-379.\u0026nbsp;\u0026nbsp;https://doi.org/10.1177/27551938241255041\u003c/li\u003e\n\u003cli\u003eBudhdeo, S., Watkins, J., Atun, R., Williams, C., Zeltner, T., \u0026amp; Maruthappu, M. (2015, Dec). Changes in government spending on healthcare and population mortality in the European union, 1995-2010: a cross-sectional ecological study. \u003cem\u003eJ R Soc Med, 108\u003c/em\u003e(12), 490-498. https://doi.org/10.1177/0141076815600907\u003c/li\u003e\n\u003cli\u003eDarvas, Z. H., P.; De Sousa, C.; Terzi, A.; Tschekassin, O. . (2014). \u003cem\u003eAusterity and Poverty in the European Union\u003c/em\u003e. E. P. D. G. f. I. Policies. https://www.europarl.europa.eu/RegData/etudes/STUD/2014/518758/IPOL_STU(2014)518758_EN.pdf\u003c/li\u003e\n\u003cli\u003eDoetsch, J. N., Schlosser, C., Barros, H., Shaw, D., Krafft, T., \u0026amp; Pilot, E. (2023, Jan 5). A scoping review on the impact of austerity on healthcare access in the European Union: rethinking austerity for the most vulnerable. \u003cem\u003eInt J Equity Health, 22\u003c/em\u003e(1), 3. https://doi.org/10.1186/s12939-022-01806-1\u003c/li\u003e\n\u003cli\u003eDonat, M., Regidor, E., Barrio, G., Ambrosio, E., Sordo, L., Guerras, J. M., Politi, J., \u0026amp; Belza, M. J. (2023, Oct). Increase in educational inequalities in alcohol-related mortality in Spain during a period of economic growth. \u003cem\u003eAddiction, 118\u003c/em\u003e(10), 1920-1931. https://doi.org/10.1111/add.16264\u003c/li\u003e\n\u003cli\u003eFaurie, C., Varghese, B. M., Liu, J., \u0026amp; Bi, P. (2022, Dec 15). Association between high temperature and heatwaves with heat-related illnesses: A systematic review and meta analysis. \u003cem\u003eSci Total Environ, 852\u003c/em\u003e, 158332. https://doi.org/10.1016/j.scitotenv.2022.158332\u003c/li\u003e\n\u003cli\u003eFranklin, B., Hochlaf, D., \u0026amp; Holley-Moore, G. (2017). \u003cem\u003eEurope during the austerity years a research report from ILC-UK\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eFriebel, R., Yoo, K. J., \u0026amp; Maynou, L. (2022, Apr). Opioid abuse and austerity: Evidence on health service use and mortality in England. \u003cem\u003eSoc Sci Med, 298\u003c/em\u003e, 114511. https://doi.org/10.1016/j.socscimed.2021.114511\u003c/li\u003e\n\u003cli\u003eGarcia-Escribano, M. J., P. ; Mogues, T. (2022). \u003cem\u003ePatterns and Drivers of Health Spending Efficiency\u003c/em\u003e. I. M. Found.\u003c/li\u003e\n\u003cli\u003eGlobal Burden of Disease Greece Collaborators. (2018, Aug). The burden of disease in Greece, health loss, risk factors, and health financing, 2000-16: an analysis of the Global Burden of Disease Study 2016. \u003cem\u003eLancet Public Health, 3\u003c/em\u003e(8), e395-e406. https://doi.org/10.1016/S2468-2667(18)30130-0\u003c/li\u003e\n\u003cli\u003eGolinelli, D., Bucci, A., Toscano, F., Filicori, F., \u0026amp; Fantini, M. P. (2018, Aug 29). Real and predicted mortality under health spending constraints in Italy: a time trend analysis through artificial neural networks. \u003cem\u003eBMC Health Serv Res, 18\u003c/em\u003e(1), 671. https://doi.org/10.1186/s12913-018-3473-3\u003c/li\u003e\n\u003cli\u003eGolinelli, D., Toscano, F., Bucci, A., Lenzi, J., Fantini, M. P., Nante, N., \u0026amp; Messina, G. (2017, Dec). Health Expenditure and All-Cause Mortality in the 'Galaxy' of Italian Regional Healthcare Systems: A 15-Year Panel Data Analysis. \u003cem\u003eAppl Health Econ Health Policy, 15\u003c/em\u003e(6), 773-783.\u0026nbsp;https://doi.org/10.1007/s40258-017-0342-x\u003c/li\u003e\n\u003cli\u003eGomez-Gallego, J. C. G., J.F.; G\u0026oacute;mez-Gallego, M. (2019). Efiuciencia y productividad de los sistemas de salud de los pa\u0026iacute;ses de la Uni\u0026oacute;n Europea [Efficiency and productivity of the health systems of the countries of the European Union]. \u003cem\u003eEstudios de Econom\u0026iacute;a Aplicada, 37\u003c/em\u003e(2), 1-19.\u0026nbsp;https://repositorio.ucam.edu/handle/10952/5365\u003c/li\u003e\n\u003cli\u003eThe Greek health crisis. (2015, Jul 11). \u003cem\u003eLancet, 386\u003c/em\u003e(9989), 104. https://doi.org/10.1016/S0140-6736(15)61226-0\u003c/li\u003e\n\u003cli\u003eHern\u0026aacute;ndez, J. (2016). Productividad y gasto sanitario p\u0026uacute;blico durante la crisis: una aproximaci\u0026oacute;n al c\u0026aacute;lculo de la productividad agregada del SNS en el periodo 2009-2013. \u003cem\u003ePresupuesto y Gasto P\u0026uacute;blico, 85\u003c/em\u003e, 165-189.\u003c/li\u003e\n\u003cli\u003eHiam, L., Dorling, D., \u0026amp; McKee, M. (2021, Mar). Austerity, not influenza, caused the UK's health to deteriorate. Let's not make the same mistake again. \u003cem\u003eJ Epidemiol Community Health, 75\u003c/em\u003e(3), 312. https://doi.org/10.1136/jech-2020-215556\u003c/li\u003e\n\u003cli\u003eHiam, L., Harrison, D., McKee, M., \u0026amp; Dorling, D. (2018, May). Why is life expectancy in England and Wales 'stalling'? \u003cem\u003eJ Epidemiol Community Health, 72\u003c/em\u003e(5), 404-408. https://doi.org/10.1136/jech-2017-210401\u003c/li\u003e\n\u003cli\u003eHiam, L., McKee, M., \u0026amp; Dorling, D. (2024, Mar 13). Influenza: cause or excuse? An analysis of flu's influence on worsening mortality trends in England and Wales, 2010-19. \u003cem\u003eBr Med Bull, 149\u003c/em\u003e(1), 72-89. https://doi.org/10.1093/bmb/ldad028\u003c/li\u003e\n\u003cli\u003eJacques, O., \u0026amp; Noel, A. (2022, Sep). The politics of public health investments. \u003cem\u003eSoc Sci Med, 309\u003c/em\u003e, 115272. https://doi.org/10.1016/j.socscimed.2022.115272\u003c/li\u003e\n\u003cli\u003eJasilionis, D., van Raalte, A. A., Klusener, S., \u0026amp; Grigoriev, P. (2023, Aug). The underwhelming German life expectancy. \u003cem\u003eEur J Epidemiol, 38\u003c/em\u003e(8), 839-850. https://doi.org/10.1007/s10654-023-00995-5\u003c/li\u003e\n\u003cli\u003eKentikelenis, A., Karanikolos, M., Reeves, A., McKee, M., \u0026amp; Stuckler, D. (2014, Feb 22). Greece's health crisis: from austerity to denialism. \u003cem\u003eLancet, 383\u003c/em\u003e(9918), 748-753. https://doi.org/10.1016/S0140-6736(13)62291-6\u003c/li\u003e\n\u003cli\u003eKinge, J. M., Dieleman, J. L., Karlstad, O., Knudsen, A. K., Klitkou, S. T., Hay, S. I., Vos, T., Murray, C. J. L., \u0026amp; Vollset, S. E. (2023, Jun 6). Disease-specific health spending by age, sex, and type of care in Norway: a national health registry study. \u003cem\u003eBMC Med, 21\u003c/em\u003e(1), 201. https://doi.org/10.1186/s12916-023-02896-6\u003c/li\u003e\n\u003cli\u003eKoltai, J., McKee, M., \u0026amp; Stuckler, D. (2021, Sep). Association between disability-related budget reductions and increasing drug-related mortality across local authorities in Great Britain. \u003cem\u003eSoc Sci Med, 284\u003c/em\u003e, 114225. https://doi.org/10.1016/j.socscimed.2021.114225\u003c/li\u003e\n\u003cli\u003eLee, J. Y., Kim, H., Gasparrini, A., Armstrong, B., Bell, M. L., Sera, F., Lavigne, E., Abrutzky, R., Tong, S., Coelho, M., Saldiva, P. H. N., Correa, P. M., Ortega, N. V., Kan, H., Garcia, S. O., Kysely, J., Urban, A., Orru, H., Indermitte, E., Jaakkola, J. J. K., Ryti, N. R. I., Pascal, M., Goodman, P. G., Zeka, A., Michelozzi, P., Scortichini, M., Hashizume, M., Honda, Y., Hurtado, M., Cruz, J., Seposo, X., Nunes, B., Teixeira, J. P., Tobias, A., Iniguez, C., Forsberg, B., Astrom, C., Vicedo-Cabrera, A. M., Ragettli, M. S., Guo, Y. L., Chen, B. Y., Zanobetti, A., Schwartz, J., Dang, T. N., Do Van, D., Mayvaneh, F., Overcenco, A., Li, S., \u0026amp; Guo, Y. (2019, Oct). Predicted temperature-increase-induced global health burden and its regional variability. \u003cem\u003eEnviron Int, 131\u003c/em\u003e, 105027. https://doi.org/10.1016/j.envint.2019.105027\u003c/li\u003e\n\u003cli\u003eLeider, J. P., Alfonso, N., Resnick, B., Brady, E., McCullough, J. M., \u0026amp; Bishai, D. (2018, Apr). Assessing The Value Of 40 Years Of Local Public Expenditures On Health. \u003cem\u003eHealth Aff (Millwood), 37\u003c/em\u003e(4), 560-569. https://doi.org/10.1377/hlthaff.2017.1171\u003c/li\u003e\n\u003cli\u003eLippi, G., Mattiuzzi, C., \u0026amp; Cervellin, G. (2016, Jul). No correlation between health care expenditure and mortality in the European Union. \u003cem\u003eEur J Intern Med, 32\u003c/em\u003e, e13-14. https://doi.org/10.1016/j.ejim.2016.02.025\u003c/li\u003e\n\u003cli\u003eLoopstra, R., McKee, M., Katikireddi, S. V., Taylor-Robinson, D., Barr, B., \u0026amp; Stuckler, D. (2016, Mar). Austerity and old-age mortality in England: a longitudinal cross-local area analysis, 2007-2013. \u003cem\u003eJ R Soc Med, 109\u003c/em\u003e(3), 109-116. https://doi.org/10.1177/0141076816632215\u003c/li\u003e\n\u003cli\u003eMackenbach, J. P., Slobbe, L., Looman, C. W., van der Heide, A., Polder, J., \u0026amp; Garssen, J. (2011, Dec). Sharp upturn of life expectancy in the Netherlands: effect of more health care for the elderly? \u003cem\u003eEur J Epidemiol, 26\u003c/em\u003e(12), 903-914. https://doi.org/10.1007/s10654-011-9633-y\u003c/li\u003e\n\u003cli\u003eMartin, S., Longo, F., Lomas, J., \u0026amp; Claxton, K. (2021, Oct 15). Causal impact of social care, public health and healthcare expenditure on mortality in England: cross-sectional evidence for 2013/2014. \u003cem\u003eBMJ Open, 11\u003c/em\u003e(10), e046417. https://doi.org/10.1136/bmjopen-2020-046417\u003c/li\u003e\n\u003cli\u003eMays, G. P., \u0026amp; Smith, S. A. (2011, Aug). Evidence links increases in public health spending to declines in preventable deaths. \u003cem\u003eHealth Aff (Millwood), 30\u003c/em\u003e(8), 1585-1593.\u0026nbsp;\u0026nbsp;https://doi.org/10.1377/hlthaff.2011.0196\u003c/li\u003e\n\u003cli\u003eMcCartney, G., McMaster, R., Popham, F., Dundas, R., \u0026amp; Walsh, D. (2022, Nov). Is austerity a cause of slower improvements in mortality in high-income countries? A panel analysis. \u003cem\u003eSoc Sci Med, 313\u003c/em\u003e, 115397. https://doi.org/10.1016/j.socscimed.2022.115397\u003c/li\u003e\n\u003cli\u003eMurphy, M. (2021, May 5). Recent mortality in Britain: a review of trends and explanations. \u003cem\u003eAge Ageing, 50\u003c/em\u003e(3), 676-683. https://doi.org/10.1093/ageing/afab016\u003c/li\u003e\n\u003cli\u003eMurphy, M. J., \u0026amp; Grundy, E. M. D. (2022, May 27). Slowdown in Mortality Improvement in the Past Decade: A US/UK Comparison. \u003cem\u003eJ Gerontol B Psychol Sci Soc Sci, 77\u003c/em\u003e(Suppl_2), S138-S147.\u0026nbsp;https://doi.org/10.1093/geronb/gbab220\u003c/li\u003e\n\u003cli\u003eOECD. (2011). \u003cem\u003eOECD Economic Outlook. \u003c/em\u003e\u0026nbsp;https://doi.org/http://dx.doi.org/10.1787/eco_outlook-v2011-2-e\u003c/li\u003e\n\u003cli\u003eONS. (2022). Public service productivity: total, UK, 2019. https://www.ons.gov.uk/economy/economicoutputandproductivity/publicservicesproductivity/articles/publicservicesproductivityestimatestotalpublicservices/2019\u003c/li\u003e\n\u003cli\u003eParmar, D., Stavropoulou, C., \u0026amp; Ioannidis, J. P. (2016, Sep 6). Health outcomes during the 2008 financial crisis in Europe: systematic literature review. \u003cem\u003eBMJ, 354\u003c/em\u003e, i4588.\u0026nbsp;\u0026nbsp;https://doi.org/10.1136/bmj.i4588\u003c/li\u003e\n\u003cli\u003ePaulus, A. F., F.; Sutherland, H. (2017). The design of fiscal consolidation measures in the European Union: distributional effects and implications for macro-economic recovery. \u003cem\u003eOxford Economic Papers, 69\u003c/em\u003e(3), 632\u0026ndash;654. https://doi.org/10.1093/oep/gpw054\u003c/li\u003e\n\u003cli\u003ePHW-NHS. (2020). \u003cem\u003eLife expectancy and mortality in Wales\u003c/em\u003e. PHW-NHS-Trust. https://phw.nhs.wales/services-and-teams/observatory/data-and-analysis/life-expectancy-and-mortality-in-wales-2020/\u003c/li\u003e\n\u003cli\u003eRajmil, L., \u0026amp; Fernandez de Sanmamed, M. J. (2019, May). Austerity Policies and Mortality Rates in European Countries, 2011-2015. \u003cem\u003eAm J Public Health, 109\u003c/em\u003e(5), 768-770.\u003c/li\u003e\n\u003cli\u003ehttps://doi.org/10.2105/AJPH.2019.304997\u003c/li\u003e\n\u003cli\u003eRaleigh, V. S. (2019). \u003cem\u003eTrends in life expectancy in eu and other OECD countries: Why are improvements slowing?\u003c/em\u003e (DELSA/HEA/WD/HWP(2019)1). (OECD Health Working Paper No. 108.\u0026nbsp;https://www.oecd-ilibrary.org/docserver/223159ab-en.pdf?expires=1715622443\u0026amp;id=id\u0026amp;accname=guest\u0026amp;checksum=86F3E26040BFF8FD311BDDDC4AFFE8AC\u003c/li\u003e\n\u003cli\u003eRamsay, J., Minton, J., Fischbacher, C., Fenton, L., Kaye-Bardgett, M., Wyper, G. M. A., Richardson, E., \u0026amp; McCartney, G. (2020, Oct 7). How have changes in death by cause and age group contributed to the recent stalling of life expectancy gains in Scotland? Comparative decomposition analysis of mortality data, 2000-2002 to 2015-2017. \u003cem\u003eBMJ Open, 10\u003c/em\u003e(10), e036529. https://doi.org/10.1136/bmjopen-2019-036529\u003c/li\u003e\n\u003cli\u003eRegidor, E., Vallejo, F., Granados, J. A. T., Viciana-Fernandez, F. J., de la Fuente, L., \u0026amp; Barrio, G. (2016, Nov 26). Mortality decrease according to socioeconomic groups during the economic crisis in Spain: a cohort study of 36 million people. \u003cem\u003eLancet, 388\u003c/em\u003e(10060), 2642-2652.\u0026nbsp;https://doi.org/10.1016/S0140-6736(16)30446-9\u003c/li\u003e\n\u003cli\u003eRegidor, E. R., A. (2021). \u003cem\u003eCrisis econ\u0026oacute;mica- 08, Covid-19 y salud: palabras cruzadas sobre la pr\u0026aacute;ctica cient\u0026iacute;fica\u003c/em\u003e. Grupo Editorial C\u0026iacute;rculo Rojo, S.L. https://editorialcirculorojo.com/autores/enrique-regidor/\u003c/li\u003e\n\u003cli\u003eRichardson, E., Taulbut, M., Robinson, M., Pulford, A., \u0026amp; McCartney, G. (2020, Oct 20). The contribution of changes to tax and social security to stalled life expectancy trends in Scotland: a modelling study. \u003cem\u003eJ Epidemiol Community Health, 75\u003c/em\u003e(4), 365-370.\u0026nbsp;https://doi.org/10.1136/jech-2020-214770\u003c/li\u003e\n\u003cli\u003eSeaman, R., Walsh, D., Beatty, C., McCartney, G., \u0026amp; Dundas, R. (2023, Nov 7). Social security cuts and life expectancy: a longitudinal analysis of local authorities in England, Scotland and Wales. \u003cem\u003eJ Epidemiol Community Health, 78\u003c/em\u003e(2), 82-87. https://doi.org/10.1136/jech-2023-220328\u003c/li\u003e\n\u003cli\u003eTanaka, T., Okamoto, S., \u0026amp; Canning, D. (2022, Feb 19). National Health Spending, Health-Care Resources, Service Utilization, and Health Outcomes. \u003cem\u003eAm J Epidemiol, 191\u003c/em\u003e(3), 386-396.\u0026nbsp;https://doi.org/10.1093/aje/kwab179\u003c/li\u003e\n\u003cli\u003eTapia Granados, J. A., \u0026amp; Ionides, E. L. (2017, Dec). Population health and the economy: Mortality and the Great Recession in Europe. \u003cem\u003eHealth Econ, 26\u003c/em\u003e(12), e219-e235. https://doi.org/10.1002/hec.3495\u003c/li\u003e\n\u003cli\u003eTapia Granados, J. A., \u0026amp; Rodriguez, J. M. (2015, Jul). Health, economic crisis, and austerity: A comparison of Greece, Finland and Iceland. \u003cem\u003eHealth Policy, 119\u003c/em\u003e(7), 941-953. https://doi.org/10.1016/j.healthpol.2015.04.009\u003c/li\u003e\n\u003cli\u003eToffolutti, V., \u0026amp; Suhrcke, M. (2019, May). Does austerity really kill? \u003cem\u003eEcon Hum Biol, 33\u003c/em\u003e, 211-223.\u0026nbsp;\u0026nbsp;https://doi.org/10.1016/j.ehb.2019.03.002\u003c/li\u003e\n\u003cli\u003eWalsh, D., Dundas, R., McCartney, G., Gibson, M., \u0026amp; Seaman, R. (2022, Dec). Bearing the burden of austerity: how do changing mortality rates in the UK compare between men and women? \u003cem\u003eJ Epidemiol Community Health, 76\u003c/em\u003e(12), 1027-1033. https://doi.org/10.1136/jech-2022-219645\u003c/li\u003e\n\u003cli\u003eWalsh, D., Tod, E., McCartney, G., \u0026amp; Levin, K. A. (2022, Dec 14). How much of the stalled mortality trends in Scotland and England can be attributed to obesity? \u003cem\u003eBMJ Open, 12\u003c/em\u003e(12), e067310.\u0026nbsp;\u0026nbsp;https://doi.org/10.1136/bmjopen-2022-067310\u003c/li\u003e\n\u003cli\u003eWalsh, D., Wyper, G. M. A., \u0026amp; McCartney, G. (2022, Jun 6). Trends in healthy life expectancy in the age of austerity. \u003cem\u003eJ Epidemiol Community Health, 76\u003c/em\u003e(8), 743-745. https://doi.org/10.1136/jech-2022-219011\u003c/li\u003e\n\u003cli\u003eWatkins, J., Wulaningsih, W., Da Zhou, C., Marshall, D. C., Sylianteng, G. D. C., Dela Rosa, P. G., Miguel, V. A., Raine, R., King, L. P., \u0026amp; Maruthappu, M. (2017, Nov 15). Effects of health and social care spending constraints on mortality in England: a time trend analysis. \u003cem\u003eBMJ Open, 7\u003c/em\u003e(11), e017722. https://doi.org/10.1136/bmjopen-2017-017722\u003c/li\u003e\n\u003cli\u003eWBG. (2024). \u003cem\u003eThe Climate Change Knowledge Portal. Country. Current climate. Trends within variability. Variability and trends of average mean surface air temperature across seasonal cycle, 1951-2020\u003c/em\u003e https://climateknowledgeportal.worldbank.org/\u003c/li\u003e\n\u003cli\u003eWHO. (2024). \u003cem\u003eThe Global Health Observatory (GHO). Data. GHO. Indicators. Indicators index \u003c/em\u003ehttps://www.who.int/data/gho/data/indicators/indicators-index\u003c/li\u003e\n\u003cli\u003eWooldridge, J. (2001). \u003cem\u003eEconometric Analysis of Cross Section and Panel Data\u003c/em\u003e. MIT Press.\u003c/li\u003e\n\u003cli\u003eXu, Z., FitzGerald, G., Guo, Y., Jalaludin, B., \u0026amp; Tong, S. (2016, Apr-May). Impact of heatwave on mortality under different heatwave definitions: A systematic review and meta-analysis. \u003cem\u003eEnviron Int, 89-90\u003c/em\u003e, 193-203. https://doi.org/10.1016/j.envint.2016.02.007\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Health spending. Mortality. Time series analysis. Europe","lastPublishedDoi":"10.21203/rs.3.rs-7772794/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7772794/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eThe recent slowdown in mortality decline in high-income countries is often attributed to decreasing in health spending.\u003c/p\u003e\u003ch2\u003eObjective\u003c/h2\u003e\u003cp\u003eTo characterize the association between government health spending and short-term mortality risk in UK, Italy, Spain, and Greece.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eCross-sectional time-series design using annual data from all residents in those countries during 2002\u0026ndash;2019. The exposure was the government health spending in thousands of constant dollars per capita and the main outcome was the age-standardized mortality rate. We examined time trends in exposure and outcome. Then, we applied linear regression models (country-fixed-effects -FE- and first-differences -FD-), adjusting for year, country, heatwave, coldwave, influenza, income per capita, and government social spending to estimate the percent change (PC) in mortality rate per unit increase in health spending.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eMortality rates declined and health spending increased sharply during 2002\u0026ndash;2010 in all four countries. Subsequently, mortality slowed its decline, while health spending either slowed its rise (UK) or decreased (other countries). Regression models show a significant association (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) between health spending and mortality in the Italy-Spain-Greece group: inverse in 2002\u0026ndash;2010 (PC-FE: -7.4%, PC-FD: -9.5%) and direct during 2010\u0026ndash;2019 (PC-FE: 7.4%, PC-FD: 6.0%), while in the UK, the association probably was inverse throughout the period 2002\u0026ndash;2019. Main results remain introducing cumulative lagged values ​​of health spending.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e\u003cp\u003eThe health spending-mortality association went from inverse to direct between 2002\u0026ndash;2010 and 2011\u0026ndash;2019 in the Italy-Spain-Greece group, so the decline in health spending could not adequately explain the slowdown in mortality decline during 2011\u0026ndash;2019 in these countries.\u003c/p\u003e","manuscriptTitle":"Changes in government health spending and short-term mortality in four European countries in the 21st century","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-09 17:33:33","doi":"10.21203/rs.3.rs-7772794/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-03-13T18:32:12+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-09T21:00:06+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"225191844112814474193329448689792755630","date":"2026-03-09T07:47:13+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"310476605790309066724925518257201066530","date":"2026-03-09T06:53:58+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-07T09:43:17+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"230907543407403725554216059471642159602","date":"2026-02-06T03:38:05+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"266096454089064857089059659694167448177","date":"2026-02-05T08:18:02+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-30T22:16:35+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"124714092386589680809538470544753287383","date":"2025-12-22T21:45:18+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-12-05T13:29:07+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-12-05T13:25:56+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-12-05T13:18:54+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-11-14T09:27:25+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-11-14T09:23:14+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"5da823bd-f2a3-4fb0-b838-628724ea621c","owner":[],"postedDate":"December 9th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":59280248,"name":"Health sciences/Diseases"},{"id":59280249,"name":"Health sciences/Health care"},{"id":59280250,"name":"Health sciences/Medical research"}],"tags":[],"updatedAt":"2026-04-27T08:23:59+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-09 17:33:33","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7772794","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7772794","identity":"rs-7772794","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}