Intraindividual variability in non-household contacts: a German longitudinal study, April 2020–December 2021

Intraindividual variability in non-household contacts: a German longitudinal study, April 2020–December 2021 | 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 Research Article Intraindividual variability in non-household contacts: a German longitudinal study, April 2020–December 2021 Chao Xu, Aleksandr Bryzgalov, Johannes Horn, Andrzej K. Jarynowski, and 11 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7796845/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 21 Feb, 2026 Read the published version in BMC Infectious Diseases → Version 1 posted 19 You are reading this latest preprint version Abstract Background Day-to-day variability in social contacts can shape transmission dynamics, yet is rarely quantified. We aimed to quantify intraindividual variability (IIV) in non-household contacts during the COVID-19 pandemic in Germany and assessed its associations with sociodemographic characteristics, vaccination, and policy stringency. Methods We analysed contact survey data with 33 waves (April 2020–December 2021; 7,845 participants; 59,462 observations). Pearson residuals from a mixed-effects negative binomial model were used to calculate the within-person standard deviation (riSD) for participants with ≥ 2 observations, serving as a proxy for IIV. We fitted Gamma regression models with log link to estimate mean ratios (MR) in three analyses: (1) sociodemographic characteristics (n = 6,251), (2) vaccination effects in participants observed both before and after their first dose within ± 100 days (n = 1,203), and (3) policy stringency effects in participants observed under both strong (index ≥ 70) and weak (< 70) conditions (n = 2,446). Results Children and adolescents (0–18 years) showed higher riSD than other age groups (MR = 1.13, 95% CI 1.10–1.16). Households with ≥ 3 members had slightly higher riSD (1.04, 95% CI 1.02–1.06) compared to single-person households. Retired (0.94, 95% CI 0.92–0.96), homemakers (0.88, 95% CI 0.85–0.91), and unemployed individuals (0.91, 95% CI 0.88–0.94) had lower riSD than those who were employed. Vaccination showed no overall association with riSD (0.99, 95% CI 0.93–1.06), though heterogeneity emerged by age and sex. Weaker stringency was strongly associated with higher riSD (1.34, 95% CI 1.31–1.37). Conclusions IIV in non-household contacts was shaped by age, household composition, and employment status, but not by vaccination status. Children and adolescents, living in larger households, and assessments during periods of weaker policy stringency exhibited greater IIV, while retired, housemakers, and unemployed individuals showed lower IIV. Vaccination did not have a consistent effect. Analyses relying solely on average contacts may misrepresent risk when IIV is high. Both models and policies should account for IIV, not just mean contact levels. contact rate contact variability infectious disease modelling vaccination policy stringency Figures Figure 1 Figure 2 Figure 3 Background Airborne infectious diseases, caused by bacteria or viruses that spread through the air via aerosols and respiratory droplets during contact, are particularly influenced by close interactions between people. Examples include measles, influenza, tuberculosis, and coronavirus disease 2019 (COVID-19). The probability of transmission and the spread of these infectious diseases through a population is mainly determined by how often and in what ways people interact. These patterns of people’s interactions are referred to as social contact patterns. Since 2006, many studies have demonstrated that contact patterns are pivotal in determining pathogen transmission, and many studies have been conducted to collect contact information [ 1 – 3 ] An example of contact pattern study is the POLYMOD study[ 3 ]. It used paper diaries collected from participants across eight European countries to record details of contacts made over a day. POLYMOD study provided the first large-scale cross-national description of mixing patterns and later became a critical baseline for assessing how contacts changed during the COVID-19 pandemic [ 4 – 7 ]. From early 2020 to 2023, many studies were conducted to assess contact patterns at different stages of the pandemic in countries including China, Germany, Canada, the UK, Netherland, Belgium, Italy, U.S., and Norway [ 4 , 6 – 16 ]. These studies reported reductions in contact rates during lockdowns, with variation by demographic group and setting. One study found that contact rates recovered differently across age groups once restrictions were lifted [ 10 ]. However, heterogeneity arose not only from demographic and setting factors, but also from vaccination status. A study reported that previously infected or vaccinated individuals in Germany had more non-household contacts than unvaccinated and uninfected individuals [ 17 ]. Similarly, a case-crossover study conducted in England and Wales between February and March 2021 found that within 14 days of their first COVID-19 vaccine, participants had significantly higher odds of reporting non-household contacts compared with the pre-vaccination period (39.6% vs. 32.2%; OR = 1.65, 95% CI 1.31–2.06) [ 18 ]. Epidemic models often rely on parameters derived from such contact surveys to simulate disease transmission, predict epidemic dynamics, and guide intervention strategies [ 16 , 19 – 24 ]. However, many epidemic models assume static contact parameters [ 21 , 22 , 25 ], or apply uniform behavioral rules [ 26 ], or treat an individual’s contact rate as constant over time. Age-structured models, for instance, typically use fixed contact matrices to represent between-group heterogeneity in mixing patterns [ 27 ]. By doing so, these models only capture differences between individuals, while overlooking variability within the same individual over time. In reality, individuals continuously change their contact behavior in response to internal and external factors, such as adaptation to epidemics [ 28 , 29 ]. Each person is a complex configuration of stable characteristics and fluctuating ones [ 30 ]. The stable characteristics may include age, sex, occupational status and family structure. The fluctuating factors may include time-related factors, such as seasonal effect and weekly cycle. As a result, contact rates exhibit short-term fluctuations (e.g., daily or weekly “wiggles”). During the pandemic, the adaptive behavior such as whether to make or reduce contact according to internal (e.g., self-perceived risk might be high for people with chronic diseases), and external circumstances (e.g. more people in the community got vaccinated) may also change the fluctuations of people’s contact behavior. Some of these fluctuations can be attributed to systematic changes (e.g., seasonality, lockdown measures, or adaptive behaviors). Yet not all variation in contact behavior can be explained by such systematic factors. The remaining, seemingly unpredictable variation has been conceptualized as intraindividual variability (IIV) [ 31 – 34 ]. In psychology and behavioral sciences, IIV has been widely studied [ 34 ]. Understanding patterns of IIV contributes to insights into human behavior in domains such as personality, cognitive performance, abilities, work values, and teaching performance [ 30 ]. Although the underlying concept of IIV is consistent across disciplines, its quantification depends on the variables of interest and study design. For instance, cognitive performance studies often measure variability in response times across repeated trials [ 35 , 36 ]. Despite the wide application of IIV in other disciplines, it has not yet been examined in the context of social contact behavior. However, if IIV in contact behavior is not merely random noise, ignoring it may lead to misleading epidemic models. For instance, two individuals may both average five daily contacts over a week, but one reports exactly five each day while the other alternates between very few and many. Although their weekly averages are identical, epidemic models informed by these patterns would yield different results. To address the gap that there is no study on IIV in relation to contact numbers during COVID-19, we use 33-wave longitudinal data from the German “COVID Pandemic: Social Contacts and Modelling” (COVIMOD) study to quantify IIV in non-household contact rates during the COVID-19 pandemic [ 4 ]. First, we examined whether IIV varied according to sociodemographic characteristics. Second, we assessed whether IIV changed following vaccination. Third, we examined whether IIV differed between periods of strong and weak policy stringency. Method The COVIMOD contact survey The COVIMOD contact survey is a 33-wave observational longitudinal study that collected participant contact information between April 2020 and December 2021 during the COVID-19 pandemic. Participants were asked to retrospectively report their contacts of the previous day. In addition, participants were asked to provide quarantine or self-isolation information in the past seven days, vaccination status, and their perceived seriousness of COVID-19. Age, sex, and other sociodemographic information were also collected. Recruitment was conducted by the market research firm Ipsos, which selected participants from the i-say.com online panel according to age, sex, and regional quotas, ensuring the study sample was representative of the German population in terms of sociodemographic characteristics. Adult participants with underage children in their households were invited to report information as proxies for their children, enabling data collection on contacts among children under 18 years of age. The questionnaire can be found in Additional file 1. More details about COVIMOD have also been described elsewhere [ 4 ]. COVIMOD used the POLYMOD contact definition: “people met in person with whom you exchanged at least a few words or had physical contact.” Participants reported all contacts from 5:00 a.m. on the previous day to 5:00 a.m. on the survey day, including both household and non-household contacts. Since household contacts were stable, this study focused only on non-household contacts [ 9 ]. To minimize potential bias arising from outlier data, we restricted each participant’s non-household contacts to a maximum of 100, consistent with previous studies [ 4 , 9 , 17 , 37 ]. Pre-existing health issues were defined as participants or household members either (i) being advised to receive the annual influenza vaccine (waves 1–13), or (ii) belonging to a medium- or high-risk health group (waves 14–33). The Stringency Index, provided by the Oxford COVID-19 Government Response Tracker (OxCGRT), was used as a proxy for the strictness of nationwide German government policies in response to the COVID-19 pandemic [ 38 ]. The index ranges from 0 to 100 and is based on multiple policy indicators, capturing changes in government responses over time. Higher values indicate stricter policies. “Vaccination-period” is a variable that combines information on vaccine availability and vaccination status. It was derived from self-reported vaccination dates. Observations were categorized into three periods: pre-vaccination (before 26 December 2020, when vaccines were unavailable), post-unvaccinated (after vaccines became available but the participant was not vaccinated), and post-vaccinated (after vaccines became available and the participant was vaccinated). Quarantine or isolation is a variable of whether the participant or a household member had been required to quarantine or isolate within the past seven days. Self-risk perception was defined based on responses to the question, “Coronavirus would be a serious illness for me.” Participants who agreed were classified as high, those who disagreed as low, and those who neither agreed nor disagreed as neutral. Sociodemographic variables included age group (0–18, 19–45, 46–65, ≥ 66 years), sex (female, male), household size (1, 2, ≥ 3 persons), and occupation status (employed, retired, homemaker, student, unemployed). Detailed definitions and sources of all variables are provided in Table S1 . Statistical analysis Compared to interindividual variability, IIV refers to variation within an individual over time. To quantify IIV, we applied the two-step residuals method [ 39 ]. Specifically, the baseline model was used to estimate each individual’s expected number of contacts at each time point. The deviations of the observed contact numbers from these expected values (i.e., the residuals) were calculated across multiple time points for the same individual and then standardized. The residual individual standard deviation (riSD) was subsequently derived as a measure of the magnitude of fluctuations an individual exhibited over a given study period. As shown in Fig. 1 . The formula for riSD is as below. $$\:{riSD}_{i}=\:\sqrt{\frac{1}{{n}_{i}-1}\:{\sum\:}_{t}^{}{({y}_{it}-\:{\widehat{y}}_{it})}^{2}}$$ 1 where \(\:{riSD}_{i}\) is the residual individual standard deviation for participant \(\:i\) . \(\:{y}_{it}\) is the observed number of non-household contacts for participant \(\:i\:\) at wave \(\:t\) , \(\:{\widehat{y}}_{it}\) is the model-predicted value, and \(\:{n}_{i}\) is the number of waves contributed by participant \(\:i\) . Unlike studies specifically designed to investigate IIV with day-to-day data collection over short periods, COVIMOD was designed to study changes in contact patterns. It consists of 33 survey waves conducted over two years, with contact data collected wave by wave. The average wave duration was 8 days (range: 4–14), and the average inter-wave interval was 11 days (range: 1–29). In each wave, when current participants dropped out and the number of respondents was smaller than the required number of participants, new participants were invited to join. Of all 7,845 participants, 1,594 (20.3%) reported contacts in only one wave, while 2,021 (25.8%) contributed data for more than 10 waves. Participants who dropped out were allowed to re-enter in later waves, resulting in unequal reporting intervals. Among participants with more than one wave, the median reporting interval was 16 days (SD = 32), with a maximum of 545 days. Additionally, the outcome of interest was the number of non-household contacts, which is count data and follows a right-skewed distribution (median = 0, interquartile range (IQR) = 0–2, range = 0–100, SD = 63.4). Given the long and unequal reporting intervals within individuals, and the fact that the variable of interest is count data, the baseline model was required to remove systematic patterns such as seasonal trends, weekly cycles, and changes in social distancing and other containment measures, as well as sociodemographic differences and adaptive behaviors [ 33 , 40 ]. Baseline model All 7,845 participants were included in the baseline model. Participants who joined only one wave were not excluded, as their data still contributed to the model and improved the accuracy of predictions, thereby yielding better residuals for participants who provided data in multiple waves and were included in the next step. As described above, sociodemographic covariates (age group, sex, household size, and occupation status), pre-existing health issues, vaccination period, perception of COVID-19 seriousness, quarantine or isolation, and stringency index were included as the covariates. Additionally, the cumulative number of waves a participant had already completed was included to account for survey fatigue [ 41 ]. The reporting date was converted as day of the year to control for seasonal effects, and weekday was included to account for weekly cycles. The reporting interval was included to account for unequal gaps between survey responses within individuals. Finally, participant ID was included as a random intercept. To illustrate the adjustment for time-related and policy effects, Fig. 2 presents predicted non-household contacts from the baseline model by day of year, weekday, and stringency index. Residuals Using the baseline model, we predicted the expected number of non-household contacts and calculated both raw and Pearson residuals. The raw residuals represent the difference between the observed and expected contact numbers, whereas the Pearson residuals are standardized by the model-predicted variance. In this study, we focused on Pearson residuals. This is because unlike raw residuals that only capture the difference between observed and expected contact numbers, Pearson residuals are standardized by the model-predicted variance, allowing comparability across individuals and model settings. As shown in Figure S1 , residuals fluctuate over time and across age groups, reflecting IIV beyond temporal and sociodemographic factors controlled for in the baseline model. To further assess the data structure, we calculated the intraclass correlation coefficient (ICC), which quantifies the proportion of total variance attributable to differences between individuals as opposed to within-individual variation over time. Residual individual standard deviation For each individual with at least two observations, riSD was calculated as a proxy for IIV in non-household contacts. The baseline model examined associations between covariates and mean trajectories, while the riSD models examined whether covariates were associated with the magnitude of individual fluctuations around expected means. Study objectives First, we examined whether riSD was associated with sociodemographic variables, thereby assessing whether within-individual variability differed systematically across groups. Second, we assessed whether riSD changed following vaccination. Participants were included if they reported contact behavior within 100 days before and after their first COVID-19 vaccination and contributed at least two observations on both sides of this window. Sensitivity analyses were conducted with narrower time windows (30, 40, and 50 days before and after vaccination, and − 50/+30 days). Third, we investigated whether riSD differed between periods of strong (stringency index ≥ 70) and weak (stringency index < 70) social distancing measures. The threshold of 70 was chosen because it coincided with the onset of German lockdown periods (Fig. 3 ). Sensitivity analyses were conducted using alternative thresholds of 65 and 80. All models were fitted using a Gamma distribution with a log link. To account for unequal precision in the variability estimates, we applied analytic weights corresponding to the number of repeated survey responses contributed by each participant. Participants with more observations provide more stable estimates of their IIV, while estimates based on only two or three responses are less reliable. Weighting by the number of observations therefore ensures that individuals with sparse data do not exert the same influence on model estimates as those with extensive follow-up. Multicollinearity was assessed using generalized variance inflation factors (GVIFs) from the fitted models. The proportion of missing data ranged from 0% (non-household contact number, household size, occupational status, survey date, vaccination period, stringency index) to 0.1% for age group and sex, 1.8% for quarantine or isolation, 7.3% for self-risk perception, and 8.0% for pre-existing health status. Missing values for age group, sex, self-risk perception, and pre-existing health status were imputed using information from the nearest available survey wave for the same participant. After imputation, missingness was reduced to 0% for age group, 0.08% for sex, 0.8% for self-risk perception, and 0.7% for pre-existing health status. Quarantine or isolation information was not imputed, as it reflects time-specific conditions that are unlikely to remain stable across waves. Several sensitivity analyses were performed, 1) All analyses were repeated restricting to participants with at least three or four observations. 2) Analyses were repeated using the dataset without computed values for missing data (complete data only). 3) Analyses were also repeatedly restricting the reporting interval to ≤ 100 days and ≤ 50 days. Stratified analyses (age and sex) and interaction analyses (for the vaccination and stringency analyses) were conducted. All statistical analyses were conducted using R (version 4.3.3, 2024-02-29) in Rstudio [ 42 ]. Modeling was performed using the “glmmTMB” [ 43 ]. Results Descriptive statistics The baseline population comprised 7,845 participants (median age 43 years, IQR 23–62; 52% female; median household size 2). Participants contributed 59,462 observations across 33 waves, with a median of 0 non-household contacts per day (IQR 0–2). For the analysis of sociodemographic predictors, 6,251 participants with ≥ 2 observations were included. For the vaccination effect on riSD, 1,203 participants who contributed observations within 100 days before and after their first vaccination were included. For the policy-stringency analysis, 2,446 participants contributed data in both periods of high (SI ≥ 70) and low (SI < 70) measures were included. Detailed descriptive statistics for each analysis population are presented in Table 1 . Table 1 Description of the study population for the baseline model and for the analytic subsets used in three analyses. Attribute Baseline Model 1st Analysis 2nd Analysis 3rd Analysis social demographic vaccine effect stringency effect unvaccinated vaccinated Weak stringency index < 70 Strong stringency index ≥ 70 Sample size: responses (participants) 59,462 (7,845) 57,868 (6,251) 3,955 (1,202) 4,343 (1,202) 27,357 (2,446) 11,477 (2,446) Analysis period April 2020 to December 2021 May 2020 to December 2021 September 2020 to November 2021 January to December 2021 May to December 2020, May to December 2021 April to May2020, December 2020 to May 2021 Non-household contact rate Mean (SD) 2.0 (6.6) 1.9 (6.3) 1.5 (5.2) 1.3 (4.4) 1.9 (6.3) 1.2 (4.7) Median (IQR) 0.0 (0.0–2.0) 0.0 (0.0–2.0) 0.0 (0.0–1.0) 0.0 (0.0–1.0) 0.0 (0.0–2.0) 0.0 (0.0–1.0) Min | Max 0.0 | 100.0 0.0 | 100.0 0.0 | 86.0 0.0 | 90.0 0.0 | 100.0 0.0 | 100.0 riSD Pearson Mean (SD) NA 0.7 (0.5) 0.6 (0.8) 0.5 (0.8) 0.8 (0.6) 0.6 (0.8) Median (IQR) NA 0.7 (0.4-1.0) 0.3 (0.0-0.8) 0.3 (0.0-0.8) 0.7 (0.4–1.1) 0.3 (0.0-0.9) Min | Max NA 0.0 | 3.1 0.0 | 5.9 0.0 | 10.1 0.0 | 3.7 0.0 | 5.7 Age Mean (SD) 41.8 (22.2) 44.1 (21.9) 51.2 (18.1) 46.2 (21.4) Median (IQR) 43.0 (23.0–62.0) 49.0 (26.0–63.0) 55.0 (37.0–66.0) 52.0 (29.0–66.0) Min | Max 0.0 | 93.0 0.0 | 93.0 0.0 | 92.0 0.0 | 92.0 Missing n 5 5 0 2 Sex Male 3,731 (48%) 3,047 (49%) 619 (51%) 1,227 (50%) Female 4,101 (52%) 3,195 (51%) 583 (49%) 1,219 (50%) Missing n 13 9 0 0 Household size group 1 2,283 (29%) 1,951 (31%) 394 (33%) 740 (30%) 2 2,728 (35%) 2,227 (36%) 468 (39%) 824 (34%) ≥ 3 2,834 (36%) 2,073 (33%) 340 (28%) 882 (36%) Occupational status Employed 4,775 (61%) 3,758 (60%) 710 (59%) 1,463 (60%) Retired 1,628 (21%) 1,420 (23%) 305 (25%) 551 (23%) Homemaker 377 (4.8%) 303 (4.8%) 57 (4.7%) 125 (5.1%) Student 581 (7.4%) 383 (6.1%) 61 (5.1%) 150 (6.1%) Unemployed 484 (6.2%) 387 (6.2%) 69 (5.7%) 157 (6.4%) Pre-existing health issues, yes 21,192 (36%) NA NA NA Missing n 226 Quarantine or isolation, yes 1,789 (3.1%) NA NA NA Missing n 1,073 Self-risk perception NA NA NA High 28,855 (49%) Neutral 15,702 (27%) Low 14,412 (24%) Missing n 493 Vaccination period NA NA NA pre-vaccination 21,620 (36%) post-vaccinated 23,112 (39%) post-unvaccinated 14,730 (25%) Number of Waves joined Mean (SD) 7.6 (7.0) 9.3 (6.9) 3.0 (1.0) 4.0 (1.0) 11.0 (5.0) 5.0 (2.0) Median (IQR) 5.0 (2.0–11.0) 7.9 (4.0–13.0) 3.0 (3.0–4.0) 4.0 (3.0–4.0) 11.0 (8.0–15.0) 5.0 (2.0–7.0) Min | Max 1.0 | 30.0 2.0 | 30.0 2.0 | 7.0 2.0 | 6.0 2.0 | 23.0 2.0 | 8.0 Note: Values are n (%) for categorical variables and mean (SD), median (IQR), and min–max for continuous variables. “Responses (participants)” counts survey responses and unique individuals; individuals may appear more than once across waves. Columns for the second analysis compare observations by vaccination status (pre-vaccination = observations prior to the first reported vaccine dose; post-vaccinated = observations after vaccination in participants who were vaccinated; post-unvaccinated = observations from participants who remained unvaccinated),while columns for the third analysis compare observations by Oxford Stringency Index (weak < 70 vs strong ≥ 70). riSD, residual individual-level standard deviation of Pearson residuals; IQR, interquartile range; SD, standard deviation. Intraclass correlation The intraclass correlation (ICC) quantifies the share of variance attributable to between-person versus within-person differences. In the baseline model, the adjusted ICC was 0.63, indicating that 63% of the variance in non-household contacts was explained by differences between individuals, and 37% by within-person variability over time. Because 1,594 of the 7,845 participants (20.3%) contributed data from only a single wave, they did not directly inform the within-person variance. Nevertheless, the ICC shows that a substantial proportion of variability occurred within individuals. Sociodemographic predictors of riSD We modeled the riSD using a Gamma regression with a log link, weighting by each participant’s number of observations. Exponentiated coefficients are reported as mean ratios (MR) of riSD. Compared with adults 19–45 years old, children/adolescents (0–18 years old) had higher within-person variability (MR = 1.13, 95% CI 1.10–1.16). Adults 46–65 years old and ≥ 66 years old were similar to the reference (0.99, 95% CI 0.97–1.01 and 1.00, 95% CI 0.98–1.02, respectively). Sex showed no difference (female vs male: 1.00, 95% CI 0.99–1.02). Relative to single-person households, 3 + persons had modestly higher riSD (1.04, 95% CI 1.02–1.07), while 2-person households did not differ (1.00, 95% CI 0.98–1.02). Versus the employed, riSD was lower for retired (0.94, 95% CI 0.92–0.96), homemaker (0.88, 95% CI 0.85–0.91), and unemployed (0.91, 95% CI 0.88–0.94); students were borderline lower (0.97, 95% CI 0.93–1.00). Results are shown in Table 2 . Table 2 Mean Ratios (MRs) and 95% confidence intervals (CIs) for the association between riSD and covariates from three analyses. 1st Analysis 2nd Analysis 3rd Analysis Variable Mean Ratio (95%CI) Age group: 19–45 ref ref ref Age group: 0–18 1.13 (1.10–1.16) 1.49 (1.27–1.75) 1.16 (1.11–1.20) Agegroup: 46–65 0.99 (0.97–1.01) 0.99 (0.91–1.07) 0.98 (0.95–1.01) Age group: 66+ 1.00 (0.98–1.02) 1.05 (0.95–1.16) 0.96 (0.93-1.00) Sex: Male ref ref ref Sex: Female 1.00 (0.99–1.02) 1.03 (0.96–1.09) 1.02 (1.00-1.04) Household size: 1 ref ref ref Household size: 2 1.00 (0.98–1.02) 0.91 (0.84–0.98) 1.00 (0.98–1.03) Household size: ≥3 1.04 (1.02–1.07) 0.89 (0.82–0.98) 1.01 (0.98–1.04) Occupational status: Employed ref ref ref Occupational status: Retired 0.94 (0.92–0.96) 1.02 (0.95–1.11) 0.97 (0.95-1.00) Occupational status: Homemaker 0.88 (0.85–0.91) 0.98 (0.84–1.14) 0.89 (0.84–0.93) Occupational status: Student 0.97 (0.93-1.00) 1.11 (0.95–1.30) 0.92 (0.88–0.97) Occupational status: Unemployed 0.91 (0.88–0.94) 0.99 (0.86–1.14) 0.92 (0.88–0.96) Vaccine: unvaccinated NA ref NA Vaccine: vaccinated NA 0.99 (0.93–1.06) NA Stringency index: ≥ 70 NA NA ref Stringency index: < 70 NA NA 1.34 (1.31–1.37) Note: The first analysis assessed sociodemographic predictors (age group, sex, household size, occupation). The second analysis assessed vaccination status, adjusted for the same sociodemographic variables. The third analysis assessed policy stringency (Oxford Stringency Index < 70 vs ≥ 70), also adjusted for sociodemographics. Abbreviations: ref = reference category; NA = not applicable; riSD = residual-based intra-individual standard deviation. Vaccination and riSD riSD did not differ by vaccination status (unvaccinated vs vaccinated: 0.99, 95% CI 0.93–1.06). Policy stringency and riSD Compared with days under strong measures (SI ≥ 70), weak stringency (SI < 70) was associated with substantially higher riSD (1.34, 95% CI 1.31–1.37). Sensitivity analyses Increasing the minimum number of observations used to compute riSD from ≥ 3 to ≥ 4 yielded estimates comparable to the main analysis. For the vaccination analyses, retaining the ≥ 2-observation criterion but narrowing the windows around vaccination (± 30, ± 40, ±50 days, and − 50/+30 days) led to the same conclusion: vaccination was not significantly associated with riSD. For the stringency analyses, using alternative SI thresholds (≥ 65 vs < 65 and ≥ 80 vs < 80) produced similar results. Complete-case analyses (no imputed missing values) were also consistent with the main findings. Finally, restricting the reporting interval to ≤ 100 days and ≤ 50 days did not materially change the estimates. However, sex became statistically significant in the sociodemographic and stringency models, with females showing slightly higher riSD than males. The magnitude of this effect was not statistically significant. Full results are provided in Supplementary Table S2 . Stratified analyses For age, two stratified analyses were conducted for ≥ 66 years and 19–65 years (Table S3). In the ≥ 66-year stratum, occupation categories with very small counts (student, unemployed, homemaker) were combined as “Other”. Patterns were broadly consistent with the main analyses results. Vaccination remained unassociated with riSD in both strata (≥ 66 years old: 1.07, 95% CI 0.96–1.19; 19–65 years old: 0.93, 95% CI 0.86–1.01). Weak stringency (SI < 70) was associated with higher riSD in both strata (≥ 66 years old: 1.29, 95% CI 1.24–1.34; 19–65 years old: 1.33, 95% CI 1.29–1.38). For sex, the analyses were conducted according to “male” and “female” respectively. Results were generally consistent with the main models (Table S3). For vaccination effect, estimates suggested modest heterogeneity by sex: among males, vaccination was associated with slightly higher within-person variability afterwards (1.09, 95% CI 1.00–1.19), whereas among females it was associated with slightly lower variability (0.91, 95% CI 0.83–0.99). For policy stringency analyses, periods with weak measures (SI < 70) showed higher riSD in both sexes (males: 1.36, 95% CI 1.32–1.41; females: 1.31, 95% CI 1.27–1.35). For the 1st analysis, the results were similar to the main analyses. Interaction analyses The vaccine effect on riSD showed modest heterogeneity: compared to adults 19–45 years old (vaccine 0.90, 95% CI 0.80–1.01), older adults (≥ 66 years old) had a higher vaccine-associated riSD (interaction 1.22, 95% CI 1.03–1.44; stratum MR ≈ 1.09), while 0–18 and 46–65 years old did not differ from the reference group. This indicates that older people shower higher variability after vaccination compared to themselves being unvaccinated. After vaccination males had slightly higher riSD (1.11, 95% CI 1.02–1.21) and females lower (modifier = 0.79, 95% CI 0.70–0.90; stratum MR ≈ 0.88). Weak measures (stringency index < 70) were associated with higher riSD across ages, but the effect reduced with age (base MR for 19–45 = 1.48, 95% CI 1.42–1.54; 46–65 = 0.84, 95% CI 0.79–0.89; ≥66 = 0.87, 95% CI 0.82–0.93) and showed no material sex modification (female modifier = 0.97, 95% CI 0.92–1.01). The results of interaction analyses can be found in Table S4. Discussion Using contact data from 7,845 participants, we quantified IIV in non-household contacts and examined its predictors. Children and adolescents consistently showed higher riSD than adults. Overall sex differences were small, although interaction analyses suggested that vaccination was associated with slightly higher riSD among males and slightly lower riSD among females. Similarly, while vaccination was not associated with riSD in the overall or stratified analyses, interaction analyses indicated modest heterogeneity by age, with older adults (≥ 66 years) showing a relatively higher vaccine-associated riSD compared with younger adults. By contrast, weaker policy stringency was consistently associated with much higher riSD across groups. Results were robust to multiple sensitivity analyses. We are not aware of prior studies that quantified within-person variability in contacts during the COVID-19 pandemic. Our findings demonstrate that contact behavior is not only heterogeneous between individuals but also exhibits patterns of fluctuation within individuals over time. Existing studies show that mean contact rates increased when distancing measures were relaxed and declined when they were tightened [ 6 , 8 , 17 , 44 ]. Our findings extend this literature by demonstrating that changes in stringency are also linked to greater day-to-day fluctuation within individuals, suggesting that loosened measures widen behavioral dispersion, not just raise average levels. Unlike between-person comparisons (e.g., vaccinated vs. unvaccinated groups), our analysis focused on within-person change over a defined window around each participant’s first vaccination [ 17 ]. We applied the same within-person logic to policy stringency by including only participants observed under both strong (SI ≥ 70) and weak (SI < 70) periods. This paired design helps isolate how the same individuals vary across contexts. A related study using a similar design reported higher odds of any non-household contact within 14 days after the first vaccine dose compared with the pre-vaccination period [ 18 ]. It is possible that contact behavior changes over shorter windows than we could capture. Because COVIMOD waves were spaced in time and computing riSD requires ≥ 2 observations per participant, pre/post windows ≤ 30 days yielded too few paired observations and unstable estimates; we therefore used a minimum 30-day window. Consequently, very transient post-vaccination changes may not have been captured in our estimates. Strengths include: (i) a large panel with highly repeated measures; (ii) a riSD that captures person-specific day-to-day fluctuation after accounting for time-related, lockdown effect; (iii) the paired, within-person design for vaccination and stringency analyses, which reduces confounding by stable, unmeasured traits inherent to between-person comparisons. However, this study also has several limitations. First, although riSD is a useful proxy for IIV, it may not fully capture qualitative variation in contact behavior. For example, if an individual consistently reports the same number of contacts across waves, the riSD is zero, even if the actual contact partners differ between waves. Second, the time span over which riSD is calculated varies across individuals. For those contributing many survey waves, the riSD may reflect variation across a long period, which limits temporal resolution and may conflate short- and long-term dynamics. To address this, we accounted for the reporting interval in the baseline model and conducted sensitivity analyses restricting the reporting interval to less than 100 days and 50 days. Third, in modeling the effect of vaccination on IIV, we treated vaccination as a binary status. This assumes a discrete behavioral shift before and after vaccination, which may oversimplify gradual or nonlinear changes, such as those observed in studies using “days since vaccination” as a continuous measure [ 45 ]. To address this, we computed riSD within a defined ± 100-day time window around the first vaccination and repeated the analyses with narrower windows (± 30, ± 40, and ± 50 days) as sensitivity checks, ensuring that our results were not driven by the choice of time frame. Fourth, about 20% of participants contributed only one wave and therefore did not directly inform within-person variance. Fifth, contacts were self-reported and are subject to recall bias; while riSD uses Pearson residuals to standardize counts, reporting errors may still inflate variability. Sixth, our stringency index thresholds (e.g., < 70 vs ≥ 70) were pragmatic and may not capture all nuances of policy intensity. Seventh, the observational design leaves room for residual confounding. Additionally, although time-related variables were included in the baseline model, other underlying trends may not have been fully controlled for. Finally, as our analyses are based on contact data from Germany, the results may not be directly generalizable to other countries with different cultural, demographic, or policy contexts. These results underscore that policy context shapes variability, not just mean contact levels. When restrictions are eased, individuals respond heterogeneously, producing larger within-person swings in contact behavior, which may influence transmission dynamics by increasing temporal dispersion in mixing. Age-targeted considerations are warranted: children/adolescents show higher day-to-day fluctuation, whereas older adults exhibit attenuated responses to policy relaxation and only small differences around vaccination. This has modeling implications: when a person’s contacts are relatively stable (riSD around 0), models that use a single contact rate can perform well; when contacts are unstable (high riSD), mean-based models tend to misrepresent risk unless within-person variability is explicitly modeled. Within-person variability in non-household contacts is substantial and is strongly associated with policy stringency and, to a lesser extent, age. Future work should link riSD to transmission outcomes and use day-to-day contact information to clarify behavioral mechanisms underlying IIV. Abbreviations CI: Confidence interval CoMix: A group of studies on contact behaviour during the COVID-19 pandemic conducted in several European countries COVID-19: Coronavirus disease caused by SARS-CoV-2 COVIMOD: A German study on contact behaviour during the COVID-19 pandemic ICC: Intraclass correlation IIV: Intraindividual variability Ipsos: A market research company IQR: Interquartile range MR: Mean ratios OxCGRT: Oxford COVID-19 Government Response Tracker POLYMOD: A landmark study on contact behaviour riSD: standard deviation of within individual residuals SI: Stringency index Declarations Ethics approval and consent to participate The data used in this study were obtained from the COVIMOD survey. Participation was voluntary and based on informed consent, in accordance with recognized ethical standards for survey research and the principles of the Declaration of Helsinki. Ethical approval for the COVIMOD study was granted by the Ethics Committee of the Medical Board Westfalen-Lippe and the University of Münster (reference number 2020–473-f-s). The data were anonymized, and no medical intervention or biological sampling was involved. Consent for publication Not applicable. Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Competing interests The authors declare that they have no conflicts of interest. Funding COVIMOD is funded by intramural funds of the Institute of Epidemiology and Social Medicine, University of Münster, and of the Institute of Medical Epidemiology, Biometry and Informatics, Martin Luther University Halle-Wittenberg, as well as by funds provided by the Robert Koch Institute, Berlin, the Helmholtz-Gemein- schaft Deutscher Forschungszentren e.V. via the HZEpiAdHoc "The Helmholtz Epidemiologic Response against the COVID-19 Pandemic" project, the Saxonian COVID-19 Research Consortium SaxoCOV (co-financed with tax funds on the basis of the budget passed by the Saxon state parliament), the Federal Ministry of Education and Research (BMBF) as part of the Network University Medicine (NUM) via the egePan Unimed project (funding code: 01KX2021) and the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation, project number 492390948). Authors' contributions RM, AK and VKJ designed the study. CX conducted the analyses. CX and RM wrote the first draft. VKJ and AK were involved in collecting and managing data. CX created the figures. All authors interpreted the data and the study findings, contributed to the writing of the manuscript, and approved the final version of the manuscript. Acknowledgements The authors thank Professor Andreas Wienke from Martin-Luther-Universität Halle-Wittenberg for his support in reviewing and confirming the statistical analyses. We also acknowledge the CoMix team for their valuable cooperation with the COVIMOD survey, including the opportunity to adapt the CoMix questionnaire for use in COVIMOD. Finally, we thank the team at Ipsos for implementing the COVIMOD survey, including adjustments to the questionnaire and sampling targets, as well as their careful attention to edge cases and technical details. References Wallinga J, Teunis P, Kretzschmar M. Using data on social contacts to estimate age-specific transmission parameters for respiratory-spread infectious agents. Am J Epidemiol. 2006;164:936–44. Mikolajczyk RT, Kretzschmar M. Collecting social contact data in the context of disease transmission: Prospective and retrospective study designs. Soc Networks. 2008;30:127–35. Mossong J, Hens N, Jit M, Beutels P, Auranen K, Mikolajczyk R, et al. Social contacts and mixing patterns relevant to the spread of infectious diseases. PLoS Med. 2008;5:e74. Tomori DV, Rübsamen N, Berger T, Scholz S, Walde J, Wittenberg I, et al. Individual social contact data and population mobility data as early markers of SARS-CoV-2 transmission dynamics during the first wave in Germany-an analysis based on the COVIMOD study. BMC Med. 2021;19:271. Harris T, Jayasundara P, Ragonnet R, Trauer J, Geard N, Zachreson C. Apparent structural changes in contact patterns during COVID-19 were driven by survey design and long-term demographic trends. arXiv [physics.soc-ph]. 2024. Veneti L, Robberstad B, Steens A, Forland F, Winje BA, Vestrheim DF, et al. Social contact patterns during the early COVID-19 pandemic in Norway: insights from a panel study, April to September 2020. BMC Public Health. 2024;24:1438. Wong KLM, Gimma A, Coletti P, CoMix Europe Working Group, Faes C, Beutels P, et al. Social contact patterns during the COVID-19 pandemic in 21 European countries - evidence from a two-year study. BMC Infect Dis. 2023;23:268. Zhang J, Litvinova M, Liang Y, Wang Y, Wang W, Zhao S, et al. Changes in contact patterns shape the dynamics of the COVID-19 outbreak in China. Science. 2020;368:1481–6. Phuong HT, Bartz A, Jarynowski AK, Lange B, Jarvis CI, Rübsamen N, et al. Changes in social contact patterns in Germany during the SARS-CoV-2 pandemic - an analysis based on the COVIMOD study. BMC Infect Dis. 2025;25:588. Backer JA, Mollema L, Vos ER, Klinkenberg D, van der Klis FR, de Melker HE, et al. Impact of physical distancing measures against COVID-19 on contacts and mixing patterns: repeated cross-sectional surveys, the Netherlands, 2016-17, April 2020 and June 2020. Euro Surveill. 2021;26. https://doi.org/10.2807/1560-7917.ES.2021.26.8.2000994. Coletti P, Wambua J, Gimma A, Willem L, Vercruysse S, Vanhoutte B, et al. CoMix: comparing mixing patterns in the Belgian population during and after lockdown. Sci Rep. 2020;10:21885. Brankston G, Merkley E, Fisman DN, Tuite AR, Poljak Z, Loewen PJ, et al. Quantifying contact patterns in response to COVID-19 public health measures in Canada. BMC Public Health. 2021;21:2040. Gimma A, Munday JD, Wong KLM, Coletti P, van Zandvoort K, Prem K, et al. Changes in social contacts in England during the COVID-19 pandemic between March 2020 and March 2021 as measured by the CoMix survey: A repeated cross-sectional study. PLoS Med. 2022;19:e1003907. Goodfellow L, Quilty BJ, van Zandvoort K, Edmunds WJ. Post-pandemic social contact patterns in the United Kingdom: the Reconnect survey. medRxiv. 2025. https://doi.org/10.1101/2025.08.13.25333584. Liu CY, Berlin J, Kiti MC, Del Fava E, Grow A, Zagheni E, et al. Rapid review of social contact patterns during the COVID-19 pandemic. Epidemiology. 2021;32:781–91. Trentini F, Manna A, Balbo N, Marziano V, Guzzetta G, O’Dell S, et al. Investigating the relationship between interventions, contact patterns, and SARS-CoV-2 transmissibility. Epidemics. 2022;40:100601. Böff L, Bartz A, Harries M, MuSPAD Consortium Group, COVIMOD Consortium Group, RESPINOW Consortium Group, et al. Dynamics of contact behaviour by self-reported COVID-19 vaccination and infection status during the COVID-19 pandemic in Germany: an analysis of two large population-based studies. BMC Med. 2025;23:406. Serisier A, Beale S, Boukari Y, Hoskins S, Nguyen V, Byrne T, et al. A case-crossover study of the effect of vaccination on SARS-CoV-2 transmission relevant behaviours during a period of national lockdown in England and Wales. Vaccine. 2023;41:511–8. Hoang T, Coletti P, Melegaro A, Wallinga J, Grijalva CG, Edmunds JW, et al. A systematic review of social contact surveys to inform transmission models of close-contact infections. Epidemiology. 2019;30:723–36. Davies NG, Barnard RC, Jarvis CI, Russell TW, Semple MG, Jit M, et al. Association of tiered restrictions and a second lockdown with COVID-19 deaths and hospital admissions in England: a modelling study. Lancet Infect Dis. 2021;21:482–92. Meyer S, Held L. Incorporating social contact data in spatio-temporal models for infectious disease spread. Biostatistics. 2016;:kxw051. Mistry D, Litvinova M, Pastore Y Piontti A, Chinazzi M, Fumanelli L, Gomes MFC, et al. Inferring high-resolution human mixing patterns for disease modeling. Nat Commun. 2021;12:323. Funk S, Bansal S, Bauch CT, Eames KTD, Edmunds WJ, Galvani AP, et al. Nine challenges in incorporating the dynamics of behaviour in infectious diseases models. Epidemics. 2015;10:21–5. Verelst F, Willem L, Beutels P. Behavioural change models for infectious disease transmission: a systematic review (2010-2015). J R Soc Interface. 2016;13:20160820. Funk S, Salathé M, Jansen VAA. Modelling the influence of human behaviour on the spread of infectious diseases: a review. J R Soc Interface. 2010;7:1247–56. Elie R, Hubert E, Turinici G. Contact rate epidemic control of COVID-19: an equilibrium view. Math Model Nat Phenom. 2020;15:35. Ram V, Schaposnik LP. A modified age-structured SIR model for COVID-19 type viruses. Sci Rep. 2021;11:15194. Ferguson N. Capturing human behaviour. Nature. 2007;446:733. Fenichel EP, Castillo-Chavez C, Ceddia MG, Chowell G, Parra PAG, Hickling GJ, et al. Adaptive human behavior in epidemiological models. Proc Natl Acad Sci U S A. 2011;108:6306–11. Nesselroade J, Ram N. Studying intraindividual variability: What we have learned that will help us understand lives in context. Res Hum Dev. 2004;1:9–29. Lerner RM, Nesselroade JR. Theory and method in the study of behavioral development: On the legacy of Joachim F. wohlwill. In: Annals of Theoretical Psychology. Boston, MA: Springer US; 1991. p. 177–89. Siegler RS. Cognitive variability: A key to understanding cognitive development. Curr Dir Psychol Sci. 1994;3:1–5. Salthouse TA. Implications of within-person variability in cognitive and neuropsychological functioning for the interpretation of change. Neuropsychology. 2007;21:401–11. Ram N, Gerstorf D. Time-structured and net intraindividual variability: tools for examining the development of dynamic characteristics and processes. Psychol Aging. 2009;24:778–91. Jutten RJ, Amariglio RE, Maruff P, Properzi MJ, Rentz DM, Johnson KA, et al. Increased intraindividual variability in reaction time performance is associated with emerging cognitive decline in cognitively unimpaired adults. Neuropsychology. 2024;38:184–97. Christ BU, Combrinck MI, Thomas KGF. Both reaction time and accuracy measures of intraindividual variability predict cognitive performance in Alzheimer’s disease. Front Hum Neurosci. 2018;12. https://doi.org/10.3389/fnhum.2018.00124. Walde J, Chaturvedi M, Berger T, Bartz A, Killewald R, Tomori DV, et al. Effect of risk status for severe COVID-19 on individual contact behaviour during the SARS-CoV-2 pandemic in 2020/2021-an analysis based on the German COVIMOD study. BMC Infect Dis. 2023;23:205. Hale T, Angrist N, Goldszmidt R, Kira B, Petherick A, Phillips T, et al. A global panel database of pandemic policies (Oxford COVID-19 Government Response Tracker). Nat Hum Behav. 2021;5:529–38. Stamps JA, Briffa M, Biro PA. Unpredictable animals: individual differences in intraindividual variability (IIV). Anim Behav. 2012;83:1325–34. Lund R. Time series analysis and its applications: With R examples. J Am Stat Assoc. 2007;102:1079–1079. Jeong D, Aggarwal S, Robinson J, Kumar N, Spearot A, Park DS. Exhaustive or exhausting? Evidence on respondent fatigue in long surveys. J Dev Econ. 2023;161:102992. The R project for statistical computing. https://www.R-project.org/. Accessed 2 Sept 2025. Brooks M, Kristensen K, Benthem K van, Magnusson A, Berg C, Nielsen A, et al. GlmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. R J. 2017;9:378. Backer JA, Bogaardt L, Beutels P, Coletti P, Edmunds WJ, Gimma A, et al. Dynamics of non-household contacts during the COVID-19 pandemic in 2020 and 2021 in the Netherlands. Sci Rep. 2023;13:5166. Buckell J, Jones J, Matthews PC, Diamond SI, Rourke E, Studley R, et al. COVID-19 vaccination, risk-compensatory behaviours, and contacts in the UK. Sci Rep. 2023;13:8441. Additional Declarations No competing interests reported. Supplementary Files Supplementaryfile1Questionnaire.docx Supplementaryfile2tablesandfigures.xlsx Cite Share Download PDF Status: Published Journal Publication published 21 Feb, 2026 Read the published version in BMC Infectious Diseases → Version 1 posted Editorial decision: Revision requested 15 Dec, 2025 Reviews received at journal 10 Dec, 2025 Reviews received at journal 02 Dec, 2025 Reviews received at journal 27 Nov, 2025 Reviews received at journal 26 Nov, 2025 Reviews received at journal 25 Nov, 2025 Reviews received at journal 25 Nov, 2025 Reviewers agreed at journal 19 Nov, 2025 Reviewers agreed at journal 19 Nov, 2025 Reviewers agreed at journal 17 Nov, 2025 Reviewers agreed at journal 17 Nov, 2025 Reviewers agreed at journal 13 Nov, 2025 Reviewers agreed at journal 11 Nov, 2025 Reviewers agreed at journal 10 Nov, 2025 Reviewers invited by journal 09 Nov, 2025 Editor invited by journal 10 Oct, 2025 Editor assigned by journal 09 Oct, 2025 Submission checks completed at journal 09 Oct, 2025 First submitted to journal 07 Oct, 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-7796845","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":547147223,"identity":"783b21cb-5722-448e-a6b0-3323fece0fd4","order_by":0,"name":"Chao Xu","email":"","orcid":"","institution":"Institute for Medical Epidemiology, Biometrics, and Informatics, Martin Luther University Halle-Wittenberg","correspondingAuthor":false,"prefix":"","firstName":"Chao","middleName":"","lastName":"Xu","suffix":""},{"id":547147224,"identity":"07845151-2dd6-4249-a55b-0f3a12ebe790","order_by":1,"name":"Aleksandr Bryzgalov","email":"","orcid":"","institution":"Institute for Medical Epidemiology, Biometrics, and Informatics, Martin Luther University Halle-Wittenberg","correspondingAuthor":false,"prefix":"","firstName":"Aleksandr","middleName":"","lastName":"Bryzgalov","suffix":""},{"id":547147225,"identity":"2e2d54bb-37e0-42a1-bd8f-9c74fbedfaa0","order_by":2,"name":"Johannes Horn","email":"","orcid":"","institution":"Institute for Medical Epidemiology, Biometrics, and Informatics, Martin Luther University Halle-Wittenberg","correspondingAuthor":false,"prefix":"","firstName":"Johannes","middleName":"","lastName":"Horn","suffix":""},{"id":547147226,"identity":"6a008c87-df5e-466f-856d-9800f6ae75e6","order_by":3,"name":"Andrzej K. Jarynowski","email":"","orcid":"","institution":"System Modelling Group, Institute of Veterinary Epidemiology and Biostatistics, Freie Universität Berlin","correspondingAuthor":false,"prefix":"","firstName":"Andrzej","middleName":"K.","lastName":"Jarynowski","suffix":""},{"id":547147227,"identity":"d3ca9265-10c1-4098-906b-2a75ab68250c","order_by":4,"name":"Vitaly Belik","email":"","orcid":"","institution":"System Modelling Group, Institute of Veterinary Epidemiology and Biostatistics, Freie Universität Berlin","correspondingAuthor":false,"prefix":"","firstName":"Vitaly","middleName":"","lastName":"Belik","suffix":""},{"id":547147228,"identity":"986b1971-600f-4c7d-9ecd-2349cf5988b0","order_by":5,"name":"Veronika K Jaeger","email":"","orcid":"","institution":"Institute of Epidemiology and Social Medicine, University of Münster","correspondingAuthor":false,"prefix":"","firstName":"Veronika","middleName":"K","lastName":"Jaeger","suffix":""},{"id":547147229,"identity":"2115afda-5bfc-479b-b44e-582a00b448f6","order_by":6,"name":"André Karch","email":"","orcid":"","institution":"Institute of Epidemiology and Social Medicine, University of Münster","correspondingAuthor":false,"prefix":"","firstName":"André","middleName":"","lastName":"Karch","suffix":""},{"id":547147230,"identity":"82588864-4271-4a53-aa8d-d08bbcf3a881","order_by":7,"name":"Huynh Thi Phuong","email":"","orcid":"","institution":"Institute of Epidemiology and Social Medicine, University of Münster","correspondingAuthor":false,"prefix":"","firstName":"Huynh","middleName":"Thi","lastName":"Phuong","suffix":""},{"id":547147231,"identity":"294d23a5-baf2-470a-9796-6187745dbc25","order_by":8,"name":"Janik Suer","email":"","orcid":"","institution":"Institute of Epidemiology and Social Medicine, University of Münster","correspondingAuthor":false,"prefix":"","firstName":"Janik","middleName":"","lastName":"Suer","suffix":""},{"id":547147232,"identity":"ff2e84ea-fd3f-4c00-9f40-d7f130f3e484","order_by":9,"name":"Marlli Zambrano","email":"","orcid":"","institution":"System Modelling Group, Institute of Veterinary Epidemiology and Biostatistics, Freie Universität Berlin","correspondingAuthor":false,"prefix":"","firstName":"Marlli","middleName":"","lastName":"Zambrano","suffix":""},{"id":547147233,"identity":"c60447a5-a0c6-43b0-b5ea-86006d83c7e7","order_by":10,"name":"Steven Schulz","email":"","orcid":"","institution":"Machine Learning Unit, Department of Engineering, NET CHECK GmbH","correspondingAuthor":false,"prefix":"","firstName":"Steven","middleName":"","lastName":"Schulz","suffix":""},{"id":547147234,"identity":"e4b56bcc-0382-4a15-9a92-5f8110562255","order_by":11,"name":"Alejandra Rincón Hidalgo","email":"","orcid":"","institution":"Machine Learning Unit, Department of Engineering, NET CHECK GmbH","correspondingAuthor":false,"prefix":"","firstName":"Alejandra","middleName":"Rincón","lastName":"Hidalgo","suffix":""},{"id":547147235,"identity":"9f7366b7-bba8-4665-b717-2c86f8f1ff1a","order_by":12,"name":"Ashish Thampi","email":"","orcid":"","institution":"Machine Learning Unit, Department of Engineering, NET CHECK GmbH","correspondingAuthor":false,"prefix":"","firstName":"Ashish","middleName":"","lastName":"Thampi","suffix":""},{"id":547147236,"identity":"104f5128-d28b-403a-beb9-d9c2eb738588","order_by":13,"name":"Richard Pastor","email":"","orcid":"","institution":"Machine Learning Unit, Department of Engineering, NET CHECK GmbH","correspondingAuthor":false,"prefix":"","firstName":"Richard","middleName":"","lastName":"Pastor","suffix":""},{"id":547147237,"identity":"483764d0-192f-4bc7-9189-d36ada0dbe0b","order_by":14,"name":"Rafael Mikolajczyk","email":"data:image/png;base64,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","orcid":"","institution":"Institute for Medical Epidemiology, Biometrics, and Informatics, Martin Luther University Halle-Wittenberg","correspondingAuthor":true,"prefix":"","firstName":"Rafael","middleName":"","lastName":"Mikolajczyk","suffix":""}],"badges":[],"createdAt":"2025-10-07 07:23:41","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7796845/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7796845/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12879-026-12940-4","type":"published","date":"2026-02-21T15:59:40+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":96354892,"identity":"965445f9-c2dd-4e00-9060-f2c4b87a0f2f","added_by":"auto","created_at":"2025-11-20 08:13:43","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4261251,"visible":true,"origin":"","legend":"","description":"","filename":"ContactVariationfinal1009.docx","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/17fb9eeb24cac9436d959d7c.docx"},{"id":96367274,"identity":"7fd6d8b6-9cb4-4f05-ae43-18ee3d60e10d","added_by":"auto","created_at":"2025-11-20 10:12:29","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":15435,"visible":true,"origin":"","legend":"","description":"","filename":"3845ab0378b84fe2b550b174c09f7d26.json","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/502f7f89f792434a77045b5e.json"},{"id":96366512,"identity":"73b68708-a0f7-4e03-8aae-86304862288d","added_by":"auto","created_at":"2025-11-20 10:11:31","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":24315,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryfile1Questionnaire.docx","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/068a4dafbaf6c1ea8ed6cfc2.docx"},{"id":96366755,"identity":"1d5454a0-f797-498f-b529-78dfc862d0fd","added_by":"auto","created_at":"2025-11-20 10:11:52","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":239720,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryfile2tablesandfigures.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/80308172e12fdf48f6f003ad.xlsx"},{"id":96354891,"identity":"796c76de-f1cb-4337-ac81-6ebec7e434bf","added_by":"auto","created_at":"2025-11-20 08:13:43","extension":"xml","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":145148,"visible":true,"origin":"","legend":"","description":"","filename":"3845ab0378b84fe2b550b174c09f7d261enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/ee2f667d7b782defa33f6125.xml"},{"id":96354882,"identity":"5849f412-eb56-45d0-bc6d-c83412280c55","added_by":"auto","created_at":"2025-11-20 08:13:43","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":112988,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/3940ea73a4290eafe348ecb5.png"},{"id":96354880,"identity":"4d363cc8-563e-49d3-bd64-58197d7bd61a","added_by":"auto","created_at":"2025-11-20 08:13:43","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":142483,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/1362483772c5ab2fbb8ab1f3.png"},{"id":96366733,"identity":"e4ebcb0f-c9de-4f50-954b-b90c5ced88eb","added_by":"auto","created_at":"2025-11-20 10:11:51","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":331227,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/e416c6bd2a28fbdb5f30a7e2.png"},{"id":96367233,"identity":"3c5457d4-09df-4285-a6b1-954a70a86119","added_by":"auto","created_at":"2025-11-20 10:12:21","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":32325,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/1a0ce8e4c1eb163524c8567b.png"},{"id":96354893,"identity":"11826919-4bb4-4a86-a19d-a19298757879","added_by":"auto","created_at":"2025-11-20 08:13:43","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":71523,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/0f6657f66164c6c1c582617c.png"},{"id":96367247,"identity":"9ad02c43-9497-47fd-a347-f33acc38aad7","added_by":"auto","created_at":"2025-11-20 10:12:22","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":60585,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/36f78aa17aa28c3caf07174c.png"},{"id":96354895,"identity":"f6d47da7-736f-42bd-9291-8bc15b88a086","added_by":"auto","created_at":"2025-11-20 08:13:43","extension":"xml","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":145677,"visible":true,"origin":"","legend":"","description":"","filename":"3845ab0378b84fe2b550b174c09f7d261structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/5f475dfe13ab503658aab887.xml"},{"id":96354894,"identity":"4cdfff18-3a3b-420c-b31f-212545a24748","added_by":"auto","created_at":"2025-11-20 08:13:43","extension":"html","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":156681,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/671bb7a5f9b0da203695d7c9.html"},{"id":96354884,"identity":"f6d3106a-ce3e-4ab1-b4b2-e25920feadcf","added_by":"auto","created_at":"2025-11-20 08:13:43","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":137024,"visible":true,"origin":"","legend":"\u003cp\u003eExample illustration of residual individual standard deviation (riSD). Both individuals have the same mean number of contacts, 5 as highlighted by the dashed line, but one is with low riSD (small variability around the mean), while the other shows high riSD (large variability around the mean).\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/c26c88135f7bd8ce2459ced8.png"},{"id":96354878,"identity":"01092054-87a3-4933-a041-fe3112cd4995","added_by":"auto","created_at":"2025-11-20 08:13:43","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":176671,"visible":true,"origin":"","legend":"\u003cp\u003ePredicted effects of temporal and policy factors on non-household contacts. (A) Effect of seasonal, (B) effect of weekly cycles, and (C) effect of stringency index. Predictions are from the baseline model, adjusted for sociodemographic covariates. Shaded areas show 95% confidence intervals. The dashed line in Panel C marks the stringency threshold (70) which was used for analyses.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/5d6056349f5ec0b8098f157b.png"},{"id":96366836,"identity":"0ce3d950-c968-4de6-a2b5-cfb3869b693a","added_by":"auto","created_at":"2025-11-20 10:11:56","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":351142,"visible":true,"origin":"","legend":"\u003cp\u003eOxford Stringency Index in Germany, April 2020–December 2021\u003cstrong\u003e. \u003c/strong\u003eThe black line shows the daily Stringency Index (0–100, higher = stricter measures). Background shading marks policy phases used for context: 1st Lockdown, 1st Relaxed, 2nd Lockdown, 2nd Relaxed, and Vaccination roll-out. The dashed red line at stringency index = 70 indicates the primary threshold used to classify periods as strong (≥70) versus weak (\u0026lt;70) stringency in the analysis. Dashed green lines at stringency index = 65 and stringency index = 80 denote thresholds used in sensitivity analyses.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/ddfc8902b743bd6aebbd27a0.png"},{"id":103251936,"identity":"019f134d-9b7c-43f4-917a-6ef3001793a3","added_by":"auto","created_at":"2026-02-23 16:12:17","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1745981,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/bab08eb0-c882-4532-8bf7-133b1ca521c1.pdf"},{"id":96354879,"identity":"98bcd492-b90f-48ae-8264-1633a229e8e5","added_by":"auto","created_at":"2025-11-20 08:13:43","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":24315,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryfile1Questionnaire.docx","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/6797c22093ad9f26c1475590.docx"},{"id":96354887,"identity":"e411c2c7-c214-4d74-a73a-3f677627ae48","added_by":"auto","created_at":"2025-11-20 08:13:43","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":239720,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryfile2tablesandfigures.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7796845/v1/626deeb87825072b2b59bfb2.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Intraindividual variability in non-household contacts: a German longitudinal study, April 2020–December 2021","fulltext":[{"header":"Background","content":"\u003cp\u003eAirborne infectious diseases, caused by bacteria or viruses that spread through the air via aerosols and respiratory droplets during contact, are particularly influenced by close interactions between people. Examples include measles, influenza, tuberculosis, and coronavirus disease 2019 (COVID-19). The probability of transmission and the spread of these infectious diseases through a population is mainly determined by how often and in what ways people interact. These patterns of people\u0026rsquo;s interactions are referred to as social contact patterns. Since 2006, many studies have demonstrated that contact patterns are pivotal in determining pathogen transmission, and many studies have been conducted to collect contact information [\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]\u003c/p\u003e\u003cp\u003eAn example of contact pattern study is the POLYMOD study[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. It used paper diaries collected from participants across eight European countries to record details of contacts made over a day. POLYMOD study provided the first large-scale cross-national description of mixing patterns and later became a critical baseline for assessing how contacts changed during the COVID-19 pandemic [\u003cspan additionalcitationids=\"CR5 CR6\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eFrom early 2020 to 2023, many studies were conducted to assess contact patterns at different stages of the pandemic in countries including China, Germany, Canada, the UK, Netherland, Belgium, Italy, U.S., and Norway [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan additionalcitationids=\"CR7 CR8 CR9 CR10 CR11 CR12 CR13 CR14 CR15\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. These studies reported reductions in contact rates during lockdowns, with variation by demographic group and setting. One study found that contact rates recovered differently across age groups once restrictions were lifted [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. However, heterogeneity arose not only from demographic and setting factors, but also from vaccination status. A study reported that previously infected or vaccinated individuals in Germany had more non-household contacts than unvaccinated and uninfected individuals [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Similarly, a case-crossover study conducted in England and Wales between February and March 2021 found that within 14 days of their first COVID-19 vaccine, participants had significantly higher odds of reporting non-household contacts compared with the pre-vaccination period (39.6% vs. 32.2%; OR\u0026thinsp;=\u0026thinsp;1.65, 95% CI 1.31\u0026ndash;2.06) [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eEpidemic models often rely on parameters derived from such contact surveys to simulate disease transmission, predict epidemic dynamics, and guide intervention strategies [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan additionalcitationids=\"CR20 CR21 CR22 CR23\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. However, many epidemic models assume static contact parameters [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e], or apply uniform behavioral rules [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e], or treat an individual\u0026rsquo;s contact rate as constant over time. Age-structured models, for instance, typically use fixed contact matrices to represent between-group heterogeneity in mixing patterns [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. By doing so, these models only capture differences between individuals, while overlooking variability within the same individual over time. In reality, individuals continuously change their contact behavior in response to internal and external factors, such as adaptation to epidemics [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eEach person is a complex configuration of stable characteristics and fluctuating ones [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. The stable characteristics may include age, sex, occupational status and family structure. The fluctuating factors may include time-related factors, such as seasonal effect and weekly cycle. As a result, contact rates exhibit short-term fluctuations (e.g., daily or weekly \u0026ldquo;wiggles\u0026rdquo;). During the pandemic, the adaptive behavior such as whether to make or reduce contact according to internal (e.g., self-perceived risk might be high for people with chronic diseases), and external circumstances (e.g. more people in the community got vaccinated) may also change the fluctuations of people\u0026rsquo;s contact behavior. Some of these fluctuations can be attributed to systematic changes (e.g., seasonality, lockdown measures, or adaptive behaviors). Yet not all variation in contact behavior can be explained by such systematic factors. The remaining, seemingly unpredictable variation has been conceptualized as intraindividual variability (IIV) [\u003cspan additionalcitationids=\"CR32 CR33\" citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eIn psychology and behavioral sciences, IIV has been widely studied [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Understanding patterns of IIV contributes to insights into human behavior in domains such as personality, cognitive performance, abilities, work values, and teaching performance [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Although the underlying concept of IIV is consistent across disciplines, its quantification depends on the variables of interest and study design. For instance, cognitive performance studies often measure variability in response times across repeated trials [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eDespite the wide application of IIV in other disciplines, it has not yet been examined in the context of social contact behavior. However, if IIV in contact behavior is not merely random noise, ignoring it may lead to misleading epidemic models. For instance, two individuals may both average five daily contacts over a week, but one reports exactly five each day while the other alternates between very few and many. Although their weekly averages are identical, epidemic models informed by these patterns would yield different results.\u003c/p\u003e\u003cp\u003eTo address the gap that there is no study on IIV in relation to contact numbers during COVID-19, we use 33-wave longitudinal data from the German \u0026ldquo;COVID Pandemic: Social Contacts and Modelling\u0026rdquo; (COVIMOD) study to quantify IIV in non-household contact rates during the COVID-19 pandemic [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. First, we examined whether IIV varied according to sociodemographic characteristics. Second, we assessed whether IIV changed following vaccination. Third, we examined whether IIV differed between periods of strong and weak policy stringency.\u003c/p\u003e"},{"header":"Method","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eThe COVIMOD contact survey\u003c/h2\u003e\u003cp\u003eThe COVIMOD contact survey is a 33-wave observational longitudinal study that collected participant contact information between April 2020 and December 2021 during the COVID-19 pandemic. Participants were asked to retrospectively report their contacts of the previous day. In addition, participants were asked to provide quarantine or self-isolation information in the past seven days, vaccination status, and their perceived seriousness of COVID-19. Age, sex, and other sociodemographic information were also collected. Recruitment was conducted by the market research firm Ipsos, which selected participants from the i-say.com online panel according to age, sex, and regional quotas, ensuring the study sample was representative of the German population in terms of sociodemographic characteristics. Adult participants with underage children in their households were invited to report information as proxies for their children, enabling data collection on contacts among children under 18 years of age.\u003c/p\u003e\u003cp\u003eThe questionnaire can be found in Additional file 1. More details about COVIMOD have also been described elsewhere [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e].\u003c/p\u003e\u003cp\u003e COVIMOD used the POLYMOD contact definition: \u0026ldquo;people met in person with whom you exchanged at least a few words or had physical contact.\u0026rdquo; Participants reported all contacts from 5:00 a.m. on the previous day to 5:00 a.m. on the survey day, including both household and non-household contacts. Since household contacts were stable, this study focused only on non-household contacts [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. To minimize potential bias arising from outlier data, we restricted each participant\u0026rsquo;s non-household contacts to a maximum of 100, consistent with previous studies [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e].\u003c/p\u003e\u003cp\u003ePre-existing health issues were defined as participants or household members either (i) being advised to receive the annual influenza vaccine (waves 1\u0026ndash;13), or (ii) belonging to a medium- or high-risk health group (waves 14\u0026ndash;33).\u003c/p\u003e\u003cp\u003eThe Stringency Index, provided by the Oxford COVID-19 Government Response Tracker (OxCGRT), was used as a proxy for the strictness of nationwide German government policies in response to the COVID-19 pandemic [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. The index ranges from 0 to 100 and is based on multiple policy indicators, capturing changes in government responses over time. Higher values indicate stricter policies.\u003c/p\u003e\u003cp\u003e\u0026ldquo;Vaccination-period\u0026rdquo; is a variable that combines information on vaccine availability and vaccination status. It was derived from self-reported vaccination dates. Observations were categorized into three periods: pre-vaccination (before 26 December 2020, when vaccines were unavailable), post-unvaccinated (after vaccines became available but the participant was not vaccinated), and post-vaccinated (after vaccines became available and the participant was vaccinated).\u003c/p\u003e\u003cp\u003eQuarantine or isolation is a variable of whether the participant or a household member had been required to quarantine or isolate within the past seven days.\u003c/p\u003e\u003cp\u003eSelf-risk perception was defined based on responses to the question, \u0026ldquo;Coronavirus would be a serious illness for me.\u0026rdquo; Participants who agreed were classified as high, those who disagreed as low, and those who neither agreed nor disagreed as neutral.\u003c/p\u003e\u003cp\u003eSociodemographic variables included age group (0\u0026ndash;18, 19\u0026ndash;45, 46\u0026ndash;65, \u0026ge;\u0026thinsp;66 years), sex (female, male), household size (1, 2, \u0026ge;\u0026thinsp;3 persons), and occupation status (employed, retired, homemaker, student, unemployed).\u003c/p\u003e\u003cp\u003eDetailed definitions and sources of all variables are provided in Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003eStatistical analysis\u003c/h2\u003e\u003cp\u003eCompared to interindividual variability, IIV refers to variation within an individual over time. To quantify IIV, we applied the two-step residuals method [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. Specifically, the baseline model was used to estimate each individual\u0026rsquo;s expected number of contacts at each time point. The deviations of the observed contact numbers from these expected values (i.e., the residuals) were calculated across multiple time points for the same individual and then standardized. The residual individual standard deviation (riSD) was subsequently derived as a measure of the magnitude of fluctuations an individual exhibited over a given study period. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. The formula for riSD is as below.\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:{riSD}_{i}=\\:\\sqrt{\\frac{1}{{n}_{i}-1}\\:{\\sum\\:}_{t}^{}{({y}_{it}-\\:{\\widehat{y}}_{it})}^{2}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{riSD}_{i}\\)\u003c/span\u003e\u003c/span\u003e is the residual individual standard deviation for participant \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{y}_{it}\\)\u003c/span\u003e\u003c/span\u003e is the observed number of non-household contacts for participant \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\:\\)\u003c/span\u003e\u003c/span\u003eat wave \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:t\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{y}}_{it}\\)\u003c/span\u003e\u003c/span\u003e is the model-predicted value, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{n}_{i}\\)\u003c/span\u003e\u003c/span\u003e is the number of waves contributed by participant \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eUnlike studies specifically designed to investigate IIV with day-to-day data collection over short periods, COVIMOD was designed to study changes in contact patterns. It consists of 33 survey waves conducted over two years, with contact data collected wave by wave. The average wave duration was 8 days (range: 4\u0026ndash;14), and the average inter-wave interval was 11 days (range: 1\u0026ndash;29). In each wave, when current participants dropped out and the number of respondents was smaller than the required number of participants, new participants were invited to join. Of all 7,845 participants, 1,594 (20.3%) reported contacts in only one wave, while 2,021 (25.8%) contributed data for more than 10 waves. Participants who dropped out were allowed to re-enter in later waves, resulting in unequal reporting intervals. Among participants with more than one wave, the median reporting interval was 16 days (SD\u0026thinsp;=\u0026thinsp;32), with a maximum of 545 days. Additionally, the outcome of interest was the number of non-household contacts, which is count data and follows a right-skewed distribution (median\u0026thinsp;=\u0026thinsp;0, interquartile range (IQR)\u0026thinsp;=\u0026thinsp;0\u0026ndash;2, range\u0026thinsp;=\u0026thinsp;0\u0026ndash;100, SD\u0026thinsp;=\u0026thinsp;63.4).\u003c/p\u003e\u003cp\u003eGiven the long and unequal reporting intervals within individuals, and the fact that the variable of interest is count data, the baseline model was required to remove systematic patterns such as seasonal trends, weekly cycles, and changes in social distancing and other containment measures, as well as sociodemographic differences and adaptive behaviors [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eBaseline model\u003c/h3\u003e\n\u003cp\u003eAll 7,845 participants were included in the baseline model. Participants who joined only one wave were not excluded, as their data still contributed to the model and improved the accuracy of predictions, thereby yielding better residuals for participants who provided data in multiple waves and were included in the next step.\u003c/p\u003e\u003cp\u003eAs described above, sociodemographic covariates (age group, sex, household size, and occupation status), pre-existing health issues, vaccination period, perception of COVID-19 seriousness, quarantine or isolation, and stringency index were included as the covariates. Additionally, the cumulative number of waves a participant had already completed was included to account for survey fatigue [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. The reporting date was converted as day of the year to control for seasonal effects, and weekday was included to account for weekly cycles. The reporting interval was included to account for unequal gaps between survey responses within individuals. Finally, participant ID was included as a random intercept. To illustrate the adjustment for time-related and policy effects, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e presents predicted non-household contacts from the baseline model by day of year, weekday, and stringency index.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\n\u003ch3\u003eResiduals\u003c/h3\u003e\n\u003cp\u003eUsing the baseline model, we predicted the expected number of non-household contacts and calculated both raw and Pearson residuals. The raw residuals represent the difference between the observed and expected contact numbers, whereas the Pearson residuals are standardized by the model-predicted variance. In this study, we focused on Pearson residuals. This is because unlike raw residuals that only capture the difference between observed and expected contact numbers, Pearson residuals are standardized by the model-predicted variance, allowing comparability across individuals and model settings. As shown in Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e, residuals fluctuate over time and across age groups, reflecting IIV beyond temporal and sociodemographic factors controlled for in the baseline model.\u003c/p\u003e\u003cp\u003eTo further assess the data structure, we calculated the intraclass correlation coefficient (ICC), which quantifies the proportion of total variance attributable to differences between individuals as opposed to within-individual variation over time.\u003c/p\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003eResidual individual standard deviation\u003c/h2\u003e\u003cp\u003eFor each individual with at least two observations, riSD was calculated as a proxy for IIV in non-household contacts. The baseline model examined associations between covariates and mean trajectories, while the riSD models examined whether covariates were associated with the magnitude of individual fluctuations around expected means.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eStudy objectives\u003c/h3\u003e\n\u003cp\u003eFirst, we examined whether riSD was associated with sociodemographic variables, thereby assessing whether within-individual variability differed systematically across groups.\u003c/p\u003e\u003cp\u003eSecond, we assessed whether riSD changed following vaccination. Participants were included if they reported contact behavior within 100 days before and after their first COVID-19 vaccination and contributed at least two observations on both sides of this window. Sensitivity analyses were conducted with narrower time windows (30, 40, and 50 days before and after vaccination, and \u0026minus;\u0026thinsp;50/+30 days).\u003c/p\u003e\u003cp\u003eThird, we investigated whether riSD differed between periods of strong (stringency index\u0026thinsp;\u0026ge;\u0026thinsp;70) and weak (stringency index\u0026thinsp;\u0026lt;\u0026thinsp;70) social distancing measures. The threshold of 70 was chosen because it coincided with the onset of German lockdown periods (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Sensitivity analyses were conducted using alternative thresholds of 65 and 80.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eAll models were fitted using a Gamma distribution with a log link. To account for unequal precision in the variability estimates, we applied analytic weights corresponding to the number of repeated survey responses contributed by each participant. Participants with more observations provide more stable estimates of their IIV, while estimates based on only two or three responses are less reliable. Weighting by the number of observations therefore ensures that individuals with sparse data do not exert the same influence on model estimates as those with extensive follow-up. Multicollinearity was assessed using generalized variance inflation factors (GVIFs) from the fitted models.\u003c/p\u003e\u003cp\u003eThe proportion of missing data ranged from 0% (non-household contact number, household size, occupational status, survey date, vaccination period, stringency index) to 0.1% for age group and sex, 1.8% for quarantine or isolation, 7.3% for self-risk perception, and 8.0% for pre-existing health status. Missing values for age group, sex, self-risk perception, and pre-existing health status were imputed using information from the nearest available survey wave for the same participant. After imputation, missingness was reduced to 0% for age group, 0.08% for sex, 0.8% for self-risk perception, and 0.7% for pre-existing health status. Quarantine or isolation information was not imputed, as it reflects time-specific conditions that are unlikely to remain stable across waves.\u003c/p\u003e\u003cp\u003eSeveral sensitivity analyses were performed, 1) All analyses were repeated restricting to participants with at least three or four observations. 2) Analyses were repeated using the dataset without computed values for missing data (complete data only). 3) Analyses were also repeatedly restricting the reporting interval to \u0026le;\u0026thinsp;100 days and \u0026le;\u0026thinsp;50 days.\u003c/p\u003e\u003cp\u003eStratified analyses (age and sex) and interaction analyses (for the vaccination and stringency analyses) were conducted.\u003c/p\u003e\u003cp\u003eAll statistical analyses were conducted using R (version 4.3.3, 2024-02-29) in Rstudio [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. Modeling was performed using the \u0026ldquo;glmmTMB\u0026rdquo; [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e].\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003eDescriptive statistics\u003c/h2\u003e\u003cp\u003eThe baseline population comprised 7,845 participants (median age 43 years, IQR 23\u0026ndash;62; 52% female; median household size 2). Participants contributed 59,462 observations across 33 waves, with a median of 0 non-household contacts per day (IQR 0\u0026ndash;2).\u003c/p\u003e\u003cp\u003eFor the analysis of sociodemographic predictors, 6,251 participants with \u0026ge;\u0026thinsp;2 observations were included. For the vaccination effect on riSD, 1,203 participants who contributed observations within 100 days before and after their first vaccination were included. For the policy-stringency analysis, 2,446 participants contributed data in both periods of high (SI\u0026thinsp;\u0026ge;\u0026thinsp;70) and low (SI\u0026thinsp;\u0026lt;\u0026thinsp;70) measures were included.\u003c/p\u003e\u003cp\u003eDetailed descriptive statistics for each analysis population are presented in Table \u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eDescription of the study population for the baseline model and for the analytic subsets used in three analyses.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAttribute\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBaseline Model\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1st Analysis\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e2nd Analysis\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e3rd Analysis\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003esocial demographic\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003evaccine effect\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003estringency effect\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eunvaccinated\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003evaccinated\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eWeak\u003c/p\u003e\u003cp\u003estringency index\u0026thinsp;\u0026lt;\u0026thinsp;70\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStrong\u003c/p\u003e\u003cp\u003estringency index\u0026thinsp;\u0026ge;\u0026thinsp;70\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSample size: responses (participants)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e59,462 (7,845)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e57,868\u003c/p\u003e\u003cp\u003e(6,251)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3,955\u003c/p\u003e\u003cp\u003e(1,202)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e4,343\u003c/p\u003e\u003cp\u003e(1,202)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e27,357 (2,446)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e11,477\u003c/p\u003e\u003cp\u003e(2,446)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAnalysis period\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eApril 2020 to December 2021\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eMay 2020 to December 2021\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSeptember 2020 to November 2021\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eJanuary to December 2021\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eMay to December 2020, May to December 2021\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eApril to May2020, December 2020 to May 2021\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNon-household contact rate\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMean (SD)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2.0 (6.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.9 (6.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.5 (5.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.3 (4.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e1.9 (6.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e1.2 (4.7)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMedian (IQR)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.0 (0.0\u0026ndash;2.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.0 (0.0\u0026ndash;2.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.0 (0.0\u0026ndash;1.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.0 (0.0\u0026ndash;1.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.0 (0.0\u0026ndash;2.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.0 (0.0\u0026ndash;1.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMin | Max\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.0 | 100.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.0 | 100.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.0 | 86.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.0 | 90.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.0 | 100.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.0 | 100.0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eriSD Pearson\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMean (SD)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.7 (0.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.6 (0.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.5 (0.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.8 (0.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.6 (0.8)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMedian (IQR)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.7 (0.4-1.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.3 (0.0-0.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.3 (0.0-0.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.7 (0.4\u0026ndash;1.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.3 (0.0-0.9)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMin | Max\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.0 | 3.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.0 | 5.9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.0 | 10.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.0 | 3.7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.0 | 5.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMean (SD)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e41.8 (22.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e44.1 (21.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e51.2 (18.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e46.2 (21.4)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMedian (IQR)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e43.0 (23.0\u0026ndash;62.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e49.0 (26.0\u0026ndash;63.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e55.0 (37.0\u0026ndash;66.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e52.0 (29.0\u0026ndash;66.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMin | Max\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.0 | 93.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.0 | 93.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e0.0 | 92.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e0.0 | 92.0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMissing n\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSex\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3,731 (48%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3,047 (49%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e619 (51%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e1,227 (50%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFemale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4,101 (52%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3,195 (51%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e583 (49%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e1,219 (50%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMissing n\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e13\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHousehold size group\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2,283 (29%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1,951 (31%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e394 (33%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e740 (30%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2,728 (35%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2,227 (36%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e468 (39%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e824 (34%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u0026ge;\u0026thinsp;3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2,834 (36%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2,073 (33%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e340 (28%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e882 (36%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOccupational status\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEmployed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4,775 (61%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3,758 (60%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e710 (59%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e1,463 (60%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRetired\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1,628 (21%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1,420 (23%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e305 (25%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e551 (23%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHomemaker\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e377 (4.8%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e303 (4.8%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e57 (4.7%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e125 (5.1%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStudent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e581 (7.4%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e383 (6.1%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e61 (5.1%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e150 (6.1%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUnemployed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e484 (6.2%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e387 (6.2%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e69 (5.7%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e157 (6.4%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePre-existing health issues, yes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e21,192 (36%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMissing n\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e226\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eQuarantine or isolation, yes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1,789 (3.1%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMissing n\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1,073\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSelf-risk perception\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHigh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e28,855 (49%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeutral\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e15,702 (27%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLow\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e14,412 (24%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMissing n\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e493\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVaccination period\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003epre-vaccination\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e21,620 (36%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003epost-vaccinated\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e23,112 (39%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003epost-unvaccinated\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e14,730 (25%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNumber of Waves joined\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMean (SD)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e7.6 (7.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e9.3 (6.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3.0 (1.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e4.0 (1.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e11.0 (5.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e5.0 (2.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMedian (IQR)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.0 (2.0\u0026ndash;11.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e7.9 (4.0\u0026ndash;13.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3.0 (3.0\u0026ndash;4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e4.0 (3.0\u0026ndash;4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e11.0 (8.0\u0026ndash;15.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e5.0 (2.0\u0026ndash;7.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMin | Max\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.0 | 30.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 | 30.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e2.0 | 7.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 | 6.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e2.0 | 23.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e2.0 | 8.0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"7\"\u003eNote: Values are n (%) for categorical variables and mean (SD), median (IQR), and min\u0026ndash;max for continuous variables. \u0026ldquo;Responses (participants)\u0026rdquo; counts survey responses and unique individuals; individuals may appear more than once across waves. Columns for the second analysis compare observations by vaccination status (pre-vaccination\u0026thinsp;=\u0026thinsp;observations prior to the first reported vaccine dose; post-vaccinated\u0026thinsp;=\u0026thinsp;observations after vaccination in participants who were vaccinated; post-unvaccinated\u0026thinsp;=\u0026thinsp;observations from participants who remained unvaccinated),while columns for the third analysis compare observations by Oxford Stringency Index (weak\u0026thinsp;\u0026lt;\u0026thinsp;70 vs strong\u0026thinsp;\u0026ge;\u0026thinsp;70). riSD, residual individual-level standard deviation of Pearson residuals; IQR, interquartile range; SD, standard deviation.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eIntraclass correlation\u003c/h2\u003e\u003cp\u003eThe intraclass correlation (ICC) quantifies the share of variance attributable to between-person versus within-person differences. In the baseline model, the adjusted ICC was 0.63, indicating that 63% of the variance in non-household contacts was explained by differences between individuals, and 37% by within-person variability over time. Because 1,594 of the 7,845 participants (20.3%) contributed data from only a single wave, they did not directly inform the within-person variance. Nevertheless, the ICC shows that a substantial proportion of variability occurred within individuals.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eSociodemographic predictors of riSD\u003c/h2\u003e\u003cp\u003e We modeled the riSD using a Gamma regression with a log link, weighting by each participant\u0026rsquo;s number of observations. Exponentiated coefficients are reported as mean ratios (MR) of riSD. Compared with adults 19\u0026ndash;45 years old, children/adolescents (0\u0026ndash;18 years old) had higher within-person variability (MR\u0026thinsp;=\u0026thinsp;1.13, 95% CI 1.10\u0026ndash;1.16). Adults 46\u0026ndash;65 years old and \u0026ge;\u0026thinsp;66 years old were similar to the reference (0.99, 95% CI 0.97\u0026ndash;1.01 and 1.00, 95% CI 0.98\u0026ndash;1.02, respectively). Sex showed no difference (female vs male: 1.00, 95% CI 0.99\u0026ndash;1.02). Relative to single-person households, 3\u0026thinsp;+\u0026thinsp;persons had modestly higher riSD (1.04, 95% CI 1.02\u0026ndash;1.07), while 2-person households did not differ (1.00, 95% CI 0.98\u0026ndash;1.02). Versus the employed, riSD was lower for retired (0.94, 95% CI 0.92\u0026ndash;0.96), homemaker (0.88, 95% CI 0.85\u0026ndash;0.91), and unemployed (0.91, 95% CI 0.88\u0026ndash;0.94); students were borderline lower (0.97, 95% CI 0.93\u0026ndash;1.00). Results are shown in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eMean Ratios (MRs) and 95% confidence intervals (CIs) for the association between riSD and covariates from three analyses.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1st Analysis\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2nd Analysis\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3rd Analysis\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariable\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e\u003cp\u003eMean Ratio (95%CI)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge group: 19\u0026ndash;45\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge group: 0\u0026ndash;18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.13 (1.10\u0026ndash;1.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.49 (1.27\u0026ndash;1.75)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.16 (1.11\u0026ndash;1.20)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAgegroup: 46\u0026ndash;65\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.99 (0.97\u0026ndash;1.01)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.99 (0.91\u0026ndash;1.07)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.98 (0.95\u0026ndash;1.01)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge group: 66+\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.00 (0.98\u0026ndash;1.02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.05 (0.95\u0026ndash;1.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.96 (0.93-1.00)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSex: Male\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSex: Female\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.00 (0.99\u0026ndash;1.02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.03 (0.96\u0026ndash;1.09)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.02 (1.00-1.04)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHousehold size: 1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHousehold size: 2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.00 (0.98\u0026ndash;1.02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.91 (0.84\u0026ndash;0.98)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.00 (0.98\u0026ndash;1.03)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHousehold size: \u0026ge;3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.04 (1.02\u0026ndash;1.07)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.89 (0.82\u0026ndash;0.98)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.01 (0.98\u0026ndash;1.04)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOccupational status: Employed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOccupational status: Retired\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.94 (0.92\u0026ndash;0.96)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.02 (0.95\u0026ndash;1.11)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.97 (0.95-1.00)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOccupational status: Homemaker\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.88 (0.85\u0026ndash;0.91)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.98 (0.84\u0026ndash;1.14)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.89 (0.84\u0026ndash;0.93)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOccupational status: Student\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.97 (0.93-1.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.11 (0.95\u0026ndash;1.30)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.92 (0.88\u0026ndash;0.97)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOccupational status: Unemployed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.91 (0.88\u0026ndash;0.94)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.99 (0.86\u0026ndash;1.14)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.92 (0.88\u0026ndash;0.96)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVaccine: unvaccinated\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVaccine: vaccinated\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.99 (0.93\u0026ndash;1.06)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStringency index: \u0026ge; 70\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eref\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStringency index: \u0026lt; 70\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.34 (1.31\u0026ndash;1.37)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"4\"\u003eNote: The first analysis assessed sociodemographic predictors (age group, sex, household size, occupation). The second analysis assessed vaccination status, adjusted for the same sociodemographic variables. The third analysis assessed policy stringency (Oxford Stringency Index\u0026thinsp;\u0026lt;\u0026thinsp;70 vs\u0026thinsp;\u0026ge;\u0026thinsp;70), also adjusted for sociodemographics. Abbreviations: ref\u0026thinsp;=\u0026thinsp;reference category; NA\u0026thinsp;=\u0026thinsp;not applicable; riSD\u0026thinsp;=\u0026thinsp;residual-based intra-individual standard deviation.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003eVaccination and riSD\u003c/h2\u003e\u003cp\u003eriSD did not differ by vaccination status (unvaccinated vs vaccinated: 0.99, 95% CI 0.93\u0026ndash;1.06).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003ePolicy stringency and riSD\u003c/h2\u003e\u003cp\u003eCompared with days under strong measures (SI\u0026thinsp;\u0026ge;\u0026thinsp;70), weak stringency (SI\u0026thinsp;\u0026lt;\u0026thinsp;70) was associated with substantially higher riSD (1.34, 95% CI 1.31\u0026ndash;1.37).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003eSensitivity analyses\u003c/h2\u003e\u003cp\u003eIncreasing the minimum number of observations used to compute riSD from \u0026ge;\u0026thinsp;3 to \u0026ge;\u0026thinsp;4 yielded estimates comparable to the main analysis. For the vaccination analyses, retaining the \u0026ge;\u0026thinsp;2-observation criterion but narrowing the windows around vaccination (\u0026plusmn;\u0026thinsp;30, \u0026plusmn;\u0026thinsp;40, \u0026plusmn;50 days, and \u0026minus;\u0026thinsp;50/+30 days) led to the same conclusion: vaccination was not significantly associated with riSD. For the stringency analyses, using alternative SI thresholds (\u0026ge;\u0026thinsp;65 vs\u0026thinsp;\u0026lt;\u0026thinsp;65 and \u0026ge;\u0026thinsp;80 vs\u0026thinsp;\u0026lt;\u0026thinsp;80) produced similar results. Complete-case analyses (no imputed missing values) were also consistent with the main findings. Finally, restricting the reporting interval to \u0026le;\u0026thinsp;100 days and \u0026le;\u0026thinsp;50 days did not materially change the estimates. However, sex became statistically significant in the sociodemographic and stringency models, with females showing slightly higher riSD than males. The magnitude of this effect was not statistically significant. Full results are provided in Supplementary Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003eStratified analyses\u003c/h2\u003e\u003cp\u003eFor age, two stratified analyses were conducted for \u0026ge;\u0026thinsp;66 years and 19\u0026ndash;65 years (Table S3). In the \u0026ge;\u0026thinsp;66-year stratum, occupation categories with very small counts (student, unemployed, homemaker) were combined as \u0026ldquo;Other\u0026rdquo;. Patterns were broadly consistent with the main analyses results. Vaccination remained unassociated with riSD in both strata (\u0026ge;\u0026thinsp;66 years old: 1.07, 95% CI 0.96\u0026ndash;1.19; 19\u0026ndash;65 years old: 0.93, 95% CI 0.86\u0026ndash;1.01). Weak stringency (SI\u0026thinsp;\u0026lt;\u0026thinsp;70) was associated with higher riSD in both strata (\u0026ge;\u0026thinsp;66 years old: 1.29, 95% CI 1.24\u0026ndash;1.34; 19\u0026ndash;65 years old: 1.33, 95% CI 1.29\u0026ndash;1.38).\u003c/p\u003e\u003cp\u003eFor sex, the analyses were conducted according to \u0026ldquo;male\u0026rdquo; and \u0026ldquo;female\u0026rdquo; respectively. Results were generally consistent with the main models (Table S3). For vaccination effect, estimates suggested modest heterogeneity by sex: among males, vaccination was associated with slightly higher within-person variability afterwards (1.09, 95% CI 1.00\u0026ndash;1.19), whereas among females it was associated with slightly lower variability (0.91, 95% CI 0.83\u0026ndash;0.99). For policy stringency analyses, periods with weak measures (SI\u0026thinsp;\u0026lt;\u0026thinsp;70) showed higher riSD in both sexes (males: 1.36, 95% CI 1.32\u0026ndash;1.41; females: 1.31, 95% CI 1.27\u0026ndash;1.35). For the 1st analysis, the results were similar to the main analyses.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\u003ch2\u003eInteraction analyses\u003c/h2\u003e\u003cp\u003eThe vaccine effect on riSD showed modest heterogeneity: compared to adults 19\u0026ndash;45 years old (vaccine 0.90, 95% CI 0.80\u0026ndash;1.01), older adults (\u0026ge;\u0026thinsp;66 years old) had a higher vaccine-associated riSD (interaction 1.22, 95% CI 1.03\u0026ndash;1.44; stratum MR\u0026thinsp;\u0026asymp;\u0026thinsp;1.09), while 0\u0026ndash;18 and 46\u0026ndash;65 years old did not differ from the reference group. This indicates that older people shower higher variability after vaccination compared to themselves being unvaccinated. After vaccination males had slightly higher riSD (1.11, 95% CI 1.02\u0026ndash;1.21) and females lower (modifier\u0026thinsp;=\u0026thinsp;0.79, 95% CI 0.70\u0026ndash;0.90; stratum MR\u0026thinsp;\u0026asymp;\u0026thinsp;0.88). Weak measures (stringency index\u0026thinsp;\u0026lt;\u0026thinsp;70) were associated with higher riSD across ages, but the effect reduced with age (base MR for 19\u0026ndash;45\u0026thinsp;=\u0026thinsp;1.48, 95% CI 1.42\u0026ndash;1.54; 46\u0026ndash;65\u0026thinsp;=\u0026thinsp;0.84, 95% CI 0.79\u0026ndash;0.89; \u0026ge;66\u0026thinsp;=\u0026thinsp;0.87, 95% CI 0.82\u0026ndash;0.93) and showed no material sex modification (female modifier\u0026thinsp;=\u0026thinsp;0.97, 95% CI 0.92\u0026ndash;1.01). The results of interaction analyses can be found in Table S4.\u003c/p\u003e\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eUsing contact data from 7,845 participants, we quantified IIV in non-household contacts and examined its predictors. Children and adolescents consistently showed higher riSD than adults. Overall sex differences were small, although interaction analyses suggested that vaccination was associated with slightly higher riSD among males and slightly lower riSD among females. Similarly, while vaccination was not associated with riSD in the overall or stratified analyses, interaction analyses indicated modest heterogeneity by age, with older adults (\u0026ge;\u0026thinsp;66 years) showing a relatively higher vaccine-associated riSD compared with younger adults. By contrast, weaker policy stringency was consistently associated with much higher riSD across groups. Results were robust to multiple sensitivity analyses.\u003c/p\u003e\u003cp\u003eWe are not aware of prior studies that quantified within-person variability in contacts during the COVID-19 pandemic. Our findings demonstrate that contact behavior is not only heterogeneous between individuals but also exhibits patterns of fluctuation within individuals over time.\u003c/p\u003e\u003cp\u003eExisting studies show that mean contact rates increased when distancing measures were relaxed and declined when they were tightened [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. Our findings extend this literature by demonstrating that changes in stringency are also linked to greater day-to-day fluctuation within individuals, suggesting that loosened measures widen behavioral dispersion, not just raise average levels. Unlike between-person comparisons (e.g., vaccinated vs. unvaccinated groups), our analysis focused on within-person change over a defined window around each participant\u0026rsquo;s first vaccination [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. We applied the same within-person logic to policy stringency by including only participants observed under both strong (SI\u0026thinsp;\u0026ge;\u0026thinsp;70) and weak (SI\u0026thinsp;\u0026lt;\u0026thinsp;70) periods. This paired design helps isolate how the same individuals vary across contexts.\u003c/p\u003e\u003cp\u003eA related study using a similar design reported higher odds of any non-household contact within 14 days after the first vaccine dose compared with the pre-vaccination period [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. It is possible that contact behavior changes over shorter windows than we could capture. Because COVIMOD waves were spaced in time and computing riSD requires\u0026thinsp;\u0026ge;\u0026thinsp;2 observations per participant, pre/post windows\u0026thinsp;\u0026le;\u0026thinsp;30 days yielded too few paired observations and unstable estimates; we therefore used a minimum 30-day window. Consequently, very transient post-vaccination changes may not have been captured in our estimates.\u003c/p\u003e\u003cp\u003eStrengths include: (i) a large panel with highly repeated measures; (ii) a riSD that captures person-specific day-to-day fluctuation after accounting for time-related, lockdown effect; (iii) the paired, within-person design for vaccination and stringency analyses, which reduces confounding by stable, unmeasured traits inherent to between-person comparisons.\u003c/p\u003e\u003cp\u003eHowever, this study also has several limitations. First, although riSD is a useful proxy for IIV, it may not fully capture qualitative variation in contact behavior. For example, if an individual consistently reports the same number of contacts across waves, the riSD is zero, even if the actual contact partners differ between waves. Second, the time span over which riSD is calculated varies across individuals. For those contributing many survey waves, the riSD may reflect variation across a long period, which limits temporal resolution and may conflate short- and long-term dynamics. To address this, we accounted for the reporting interval in the baseline model and conducted sensitivity analyses restricting the reporting interval to less than 100 days and 50 days. Third, in modeling the effect of vaccination on IIV, we treated vaccination as a binary status. This assumes a discrete behavioral shift before and after vaccination, which may oversimplify gradual or nonlinear changes, such as those observed in studies using \u0026ldquo;days since vaccination\u0026rdquo; as a continuous measure [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. To address this, we computed riSD within a defined\u0026thinsp;\u0026plusmn;\u0026thinsp;100-day time window around the first vaccination and repeated the analyses with narrower windows (\u0026plusmn;\u0026thinsp;30, \u0026plusmn;\u0026thinsp;40, and \u0026plusmn;\u0026thinsp;50 days) as sensitivity checks, ensuring that our results were not driven by the choice of time frame. Fourth, about 20% of participants contributed only one wave and therefore did not directly inform within-person variance. Fifth, contacts were self-reported and are subject to recall bias; while riSD uses Pearson residuals to standardize counts, reporting errors may still inflate variability. Sixth, our stringency index thresholds (e.g., \u0026lt;\u0026thinsp;70 vs\u0026thinsp;\u0026ge;\u0026thinsp;70) were pragmatic and may not capture all nuances of policy intensity. Seventh, the observational design leaves room for residual confounding. Additionally, although time-related variables were included in the baseline model, other underlying trends may not have been fully controlled for. Finally, as our analyses are based on contact data from Germany, the results may not be directly generalizable to other countries with different cultural, demographic, or policy contexts.\u003c/p\u003e\u003cp\u003eThese results underscore that policy context shapes variability, not just mean contact levels. When restrictions are eased, individuals respond heterogeneously, producing larger within-person swings in contact behavior, which may influence transmission dynamics by increasing temporal dispersion in mixing. Age-targeted considerations are warranted: children/adolescents show higher day-to-day fluctuation, whereas older adults exhibit attenuated responses to policy relaxation and only small differences around vaccination. This has modeling implications: when a person\u0026rsquo;s contacts are relatively stable (riSD around 0), models that use a single contact rate can perform well; when contacts are unstable (high riSD), mean-based models tend to misrepresent risk unless within-person variability is explicitly modeled. Within-person variability in non-household contacts is substantial and is strongly associated with policy stringency and, to a lesser extent, age. Future work should link riSD to transmission outcomes and use day-to-day contact information to clarify behavioral mechanisms underlying IIV.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eCI: Confidence interval\u003c/p\u003e\n\u003cp\u003eCoMix: A group of studies on contact behaviour during the COVID-19 pandemic conducted in several European countries\u003c/p\u003e\n\u003cp\u003eCOVID-19: Coronavirus disease caused by SARS-CoV-2\u003c/p\u003e\n\u003cp\u003eCOVIMOD: A German study on contact behaviour during the COVID-19 pandemic\u003c/p\u003e\n\u003cp\u003eICC: Intraclass correlation\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIIV: Intraindividual variability\u003c/p\u003e\n\u003cp\u003eIpsos: A market research company\u003c/p\u003e\n\u003cp\u003eIQR: Interquartile range\u003c/p\u003e\n\u003cp\u003eMR: Mean ratios\u003c/p\u003e\n\u003cp\u003eOxCGRT: Oxford COVID-19 Government Response Tracker\u003c/p\u003e\n\u003cp\u003ePOLYMOD: A landmark study on contact behaviour\u003c/p\u003e\n\u003cp\u003eriSD: standard deviation of within individual residuals\u003c/p\u003e\n\u003cp\u003eSI: Stringency index\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data used in this study were obtained from the COVIMOD survey. Participation was voluntary and based on informed consent, in accordance with recognized ethical standards for survey research and the principles of the Declaration of Helsinki. Ethical approval for the COVIMOD study was granted by the Ethics Committee of the Medical Board Westfalen-Lippe and the University of M\u0026uuml;nster (reference number 2020\u0026ndash;473-f-s). The data were anonymized, and no medical intervention or biological sampling was involved.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no conflicts of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCOVIMOD is funded by intramural funds of the Institute of Epidemiology and Social Medicine, University of M\u0026uuml;nster, and of the Institute of Medical Epidemiology, Biometry and Informatics, Martin Luther University Halle-Wittenberg, as well as by funds provided by the Robert Koch Institute, Berlin, the Helmholtz-Gemein- schaft Deutscher Forschungszentren e.V. via the HZEpiAdHoc \u0026quot;The Helmholtz Epidemiologic Response against the COVID-19 Pandemic\u0026quot; project, the Saxonian COVID-19 Research Consortium SaxoCOV (co-financed with tax funds on the basis of the budget passed by the Saxon state parliament), the Federal Ministry of Education and Research (BMBF) as part of the Network University Medicine (NUM) via the egePan Unimed project (funding code: 01KX2021) and the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation, project number 492390948).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRM, AK and VKJ designed the study. CX conducted the analyses. CX and RM wrote the first draft. VKJ and AK were involved in collecting and managing data. CX created the figures. All authors interpreted the data and the study findings, contributed to the writing of the manuscript, and approved the final version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors thank Professor Andreas Wienke from Martin-Luther-Universit\u0026auml;t Halle-Wittenberg for his support in reviewing and confirming the statistical analyses. We also acknowledge the CoMix team for their valuable cooperation with the COVIMOD survey, including the opportunity to adapt the CoMix questionnaire for use in COVIMOD. Finally, we thank the team at Ipsos for implementing the COVIMOD survey, including adjustments to the questionnaire and sampling targets, as well as their careful attention to edge cases and technical details.\u003c/p\u003e"},{"header":"References","content":"\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003eWallinga J, Teunis P, Kretzschmar M. Using data on social contacts to estimate age-specific transmission parameters for respiratory-spread infectious agents. Am J Epidemiol. 2006;164:936\u0026ndash;44.\u003c/li\u003e\n \u003cli\u003eMikolajczyk RT, Kretzschmar M. Collecting social contact data in the context of disease transmission: Prospective and retrospective study designs. Soc Networks. 2008;30:127\u0026ndash;35.\u003c/li\u003e\n \u003cli\u003eMossong J, Hens N, Jit M, Beutels P, Auranen K, Mikolajczyk R, et al. Social contacts and mixing patterns relevant to the spread of infectious diseases. PLoS Med. 2008;5:e74.\u003c/li\u003e\n \u003cli\u003eTomori DV, R\u0026uuml;bsamen N, Berger T, Scholz S, Walde J, Wittenberg I, et al. Individual social contact data and population mobility data as early markers of SARS-CoV-2 transmission dynamics during the first wave in Germany-an analysis based on the COVIMOD study. BMC Med. 2021;19:271.\u003c/li\u003e\n \u003cli\u003eHarris T, Jayasundara P, Ragonnet R, Trauer J, Geard N, Zachreson C. Apparent structural changes in contact patterns during COVID-19 were driven by survey design and long-term demographic trends. arXiv [physics.soc-ph]. 2024.\u003c/li\u003e\n \u003cli\u003eVeneti L, Robberstad B, Steens A, Forland F, Winje BA, Vestrheim DF, et al. Social contact patterns during the early COVID-19 pandemic in Norway: insights from a panel study, April to September 2020. BMC Public Health. 2024;24:1438.\u003c/li\u003e\n \u003cli\u003eWong KLM, Gimma A, Coletti P, CoMix Europe Working Group, Faes C, Beutels P, et al. Social contact patterns during the COVID-19 pandemic in 21 European countries - evidence from a two-year study. BMC Infect Dis. 2023;23:268.\u003c/li\u003e\n \u003cli\u003eZhang J, Litvinova M, Liang Y, Wang Y, Wang W, Zhao S, et al. Changes in contact patterns shape the dynamics of the COVID-19 outbreak in China. Science. 2020;368:1481\u0026ndash;6.\u003c/li\u003e\n \u003cli\u003ePhuong HT, Bartz A, Jarynowski AK, Lange B, Jarvis CI, R\u0026uuml;bsamen N, et al. Changes in social contact patterns in Germany during the SARS-CoV-2 pandemic - an analysis based on the COVIMOD study. BMC Infect Dis. 2025;25:588.\u003c/li\u003e\n \u003cli\u003eBacker JA, Mollema L, Vos ER, Klinkenberg D, van der Klis FR, de Melker HE, et al. Impact of physical distancing measures against COVID-19 on contacts and mixing patterns: repeated cross-sectional surveys, the Netherlands, 2016-17, April 2020 and June 2020. Euro Surveill. 2021;26. https://doi.org/10.2807/1560-7917.ES.2021.26.8.2000994.\u003c/li\u003e\n \u003cli\u003eColetti P, Wambua J, Gimma A, Willem L, Vercruysse S, Vanhoutte B, et al. CoMix: comparing mixing patterns in the Belgian population during and after lockdown. Sci Rep. 2020;10:21885.\u003c/li\u003e\n \u003cli\u003eBrankston G, Merkley E, Fisman DN, Tuite AR, Poljak Z, Loewen PJ, et al. Quantifying contact patterns in response to COVID-19 public health measures in Canada. BMC Public Health. 2021;21:2040.\u003c/li\u003e\n \u003cli\u003eGimma A, Munday JD, Wong KLM, Coletti P, van Zandvoort K, Prem K, et al. Changes in social contacts in England during the COVID-19 pandemic between March 2020 and March 2021 as measured by the CoMix survey: A repeated cross-sectional study. PLoS Med. 2022;19:e1003907.\u003c/li\u003e\n \u003cli\u003eGoodfellow L, Quilty BJ, van Zandvoort K, Edmunds WJ. Post-pandemic social contact patterns in the United Kingdom: the Reconnect survey. medRxiv. 2025. https://doi.org/10.1101/2025.08.13.25333584.\u003c/li\u003e\n \u003cli\u003eLiu CY, Berlin J, Kiti MC, Del Fava E, Grow A, Zagheni E, et al. Rapid review of social contact patterns during the COVID-19 pandemic. Epidemiology. 2021;32:781\u0026ndash;91.\u003c/li\u003e\n \u003cli\u003eTrentini F, Manna A, Balbo N, Marziano V, Guzzetta G, O\u0026rsquo;Dell S, et al. Investigating the relationship between interventions, contact patterns, and SARS-CoV-2 transmissibility. Epidemics. 2022;40:100601.\u003c/li\u003e\n \u003cli\u003eB\u0026ouml;ff L, Bartz A, Harries M, MuSPAD Consortium Group, COVIMOD Consortium Group, RESPINOW Consortium Group, et al. Dynamics of contact behaviour by self-reported COVID-19 vaccination and infection status during the COVID-19 pandemic in Germany: an analysis of two large population-based studies. BMC Med. 2025;23:406.\u003c/li\u003e\n \u003cli\u003eSerisier A, Beale S, Boukari Y, Hoskins S, Nguyen V, Byrne T, et al. A case-crossover study of the effect of vaccination on SARS-CoV-2 transmission relevant behaviours during a period of national lockdown in England and Wales. Vaccine. 2023;41:511\u0026ndash;8.\u003c/li\u003e\n \u003cli\u003eHoang T, Coletti P, Melegaro A, Wallinga J, Grijalva CG, Edmunds JW, et al. A systematic review of social contact surveys to inform transmission models of close-contact infections. Epidemiology. 2019;30:723\u0026ndash;36.\u003c/li\u003e\n \u003cli\u003eDavies NG, Barnard RC, Jarvis CI, Russell TW, Semple MG, Jit M, et al. Association of tiered restrictions and a second lockdown with COVID-19 deaths and hospital admissions in England: a modelling study. Lancet Infect Dis. 2021;21:482\u0026ndash;92.\u003c/li\u003e\n \u003cli\u003eMeyer S, Held L. Incorporating social contact data in spatio-temporal models for infectious disease spread. Biostatistics. 2016;:kxw051.\u003c/li\u003e\n \u003cli\u003eMistry D, Litvinova M, Pastore Y Piontti A, Chinazzi M, Fumanelli L, Gomes MFC, et al. Inferring high-resolution human mixing patterns for disease modeling. Nat Commun. 2021;12:323.\u003c/li\u003e\n \u003cli\u003eFunk S, Bansal S, Bauch CT, Eames KTD, Edmunds WJ, Galvani AP, et al. Nine challenges in incorporating the dynamics of behaviour in infectious diseases models. Epidemics. 2015;10:21\u0026ndash;5.\u003c/li\u003e\n \u003cli\u003eVerelst F, Willem L, Beutels P. Behavioural change models for infectious disease transmission: a systematic review (2010-2015). J R Soc Interface. 2016;13:20160820.\u003c/li\u003e\n \u003cli\u003eFunk S, Salath\u0026eacute; M, Jansen VAA. Modelling the influence of human behaviour on the spread of infectious diseases: a review. J R Soc Interface. 2010;7:1247\u0026ndash;56.\u003c/li\u003e\n \u003cli\u003eElie R, Hubert E, Turinici G. Contact rate epidemic control of COVID-19: an equilibrium view. Math Model Nat Phenom. 2020;15:35.\u003c/li\u003e\n \u003cli\u003eRam V, Schaposnik LP. A modified age-structured SIR model for COVID-19 type viruses. Sci Rep. 2021;11:15194.\u003c/li\u003e\n \u003cli\u003eFerguson N. Capturing human behaviour. Nature. 2007;446:733.\u003c/li\u003e\n \u003cli\u003eFenichel EP, Castillo-Chavez C, Ceddia MG, Chowell G, Parra PAG, Hickling GJ, et al. Adaptive human behavior in epidemiological models. Proc Natl Acad Sci U S A. 2011;108:6306\u0026ndash;11.\u003c/li\u003e\n \u003cli\u003eNesselroade J, Ram N. Studying intraindividual variability: What we have learned that will help us understand lives in context. Res Hum Dev. 2004;1:9\u0026ndash;29.\u003c/li\u003e\n \u003cli\u003eLerner RM, Nesselroade JR. Theory and method in the study of behavioral development: On the legacy of Joachim F. wohlwill. In: Annals of Theoretical Psychology. Boston, MA: Springer US; 1991. p. 177\u0026ndash;89.\u003c/li\u003e\n \u003cli\u003eSiegler RS. Cognitive variability: A key to understanding cognitive development. Curr Dir Psychol Sci. 1994;3:1\u0026ndash;5.\u003c/li\u003e\n \u003cli\u003eSalthouse TA. Implications of within-person variability in cognitive and neuropsychological functioning for the interpretation of change. Neuropsychology. 2007;21:401\u0026ndash;11.\u003c/li\u003e\n \u003cli\u003eRam N, Gerstorf D. Time-structured and net intraindividual variability: tools for examining the development of dynamic characteristics and processes. Psychol Aging. 2009;24:778\u0026ndash;91.\u003c/li\u003e\n \u003cli\u003e\u0026nbsp;Jutten RJ, Amariglio RE, Maruff P, Properzi MJ, Rentz DM, Johnson KA, et al. Increased intraindividual variability in reaction time performance is associated with emerging cognitive decline in cognitively unimpaired adults. Neuropsychology. 2024;38:184\u0026ndash;97.\u003c/li\u003e\n \u003cli\u003eChrist BU, Combrinck MI, Thomas KGF. Both reaction time and accuracy measures of intraindividual variability predict cognitive performance in Alzheimer\u0026rsquo;s disease. Front Hum Neurosci. 2018;12. https://doi.org/10.3389/fnhum.2018.00124.\u003c/li\u003e\n \u003cli\u003eWalde J, Chaturvedi M, Berger T, Bartz A, Killewald R, Tomori DV, et al. Effect of risk status for severe COVID-19 on individual contact behaviour during the SARS-CoV-2 pandemic in 2020/2021-an analysis based on the German COVIMOD study. BMC Infect Dis. 2023;23:205.\u003c/li\u003e\n \u003cli\u003eHale T, Angrist N, Goldszmidt R, Kira B, Petherick A, Phillips T, et al. A global panel database of pandemic policies (Oxford COVID-19 Government Response Tracker). Nat Hum Behav. 2021;5:529\u0026ndash;38.\u003c/li\u003e\n \u003cli\u003eStamps JA, Briffa M, Biro PA. Unpredictable animals: individual differences in intraindividual variability (IIV). Anim Behav. 2012;83:1325\u0026ndash;34.\u003c/li\u003e\n \u003cli\u003eLund R. Time series analysis and its applications: With R examples. J Am Stat Assoc. 2007;102:1079\u0026ndash;1079.\u003c/li\u003e\n \u003cli\u003eJeong D, Aggarwal S, Robinson J, Kumar N, Spearot A, Park DS. Exhaustive or exhausting? Evidence on respondent fatigue in long surveys. J Dev Econ. 2023;161:102992.\u003c/li\u003e\n \u003cli\u003eThe R project for statistical computing. https://www.R-project.org/. Accessed 2 Sept 2025.\u003c/li\u003e\n \u003cli\u003eBrooks M, Kristensen K, Benthem K van, Magnusson A, Berg C, Nielsen A, et al. GlmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. R J. 2017;9:378.\u003c/li\u003e\n \u003cli\u003eBacker JA, Bogaardt L, Beutels P, Coletti P, Edmunds WJ, Gimma A, et al. Dynamics of non-household contacts during the COVID-19 pandemic in 2020 and 2021 in the Netherlands. Sci Rep. 2023;13:5166.\u003c/li\u003e\n \u003cli\u003eBuckell J, Jones J, Matthews PC, Diamond SI, Rourke E, Studley R, et al. COVID-19 vaccination, risk-compensatory behaviours, and contacts in the UK. Sci Rep. 2023;13:8441.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-infectious-diseases","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"infd","sideBox":"Learn more about [BMC Infectious Diseases](http://bmcinfectdis.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/infd","title":"BMC Infectious Diseases","twitterHandle":"#bmcinfectdis","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"contact rate, contact variability, infectious disease modelling, vaccination, policy stringency","lastPublishedDoi":"10.21203/rs.3.rs-7796845/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7796845/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eDay-to-day variability in social contacts can shape transmission dynamics, yet is rarely quantified. We aimed to quantify intraindividual variability (IIV) in non-household contacts during the COVID-19 pandemic in Germany and assessed its associations with sociodemographic characteristics, vaccination, and policy stringency.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eWe analysed contact survey data with 33 waves (April 2020\u0026ndash;December 2021; 7,845 participants; 59,462 observations). Pearson residuals from a mixed-effects negative binomial model were used to calculate the within-person standard deviation (riSD) for participants with \u0026ge;\u0026thinsp;2 observations, serving as a proxy for IIV. We fitted Gamma regression models with log link to estimate mean ratios (MR) in three analyses: (1) sociodemographic characteristics (n\u0026thinsp;=\u0026thinsp;6,251), (2) vaccination effects in participants observed both before and after their first dose within \u0026plusmn;\u0026thinsp;100 days (n\u0026thinsp;=\u0026thinsp;1,203), and (3) policy stringency effects in participants observed under both strong (index\u0026thinsp;\u0026ge;\u0026thinsp;70) and weak (\u0026lt;\u0026thinsp;70) conditions (n\u0026thinsp;=\u0026thinsp;2,446).\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eChildren and adolescents (0\u0026ndash;18 years) showed higher riSD than other age groups (MR\u0026thinsp;=\u0026thinsp;1.13, 95% CI 1.10\u0026ndash;1.16). Households with \u0026ge;\u0026thinsp;3 members had slightly higher riSD (1.04, 95% CI 1.02\u0026ndash;1.06) compared to single-person households. Retired (0.94, 95% CI 0.92\u0026ndash;0.96), homemakers (0.88, 95% CI 0.85\u0026ndash;0.91), and unemployed individuals (0.91, 95% CI 0.88\u0026ndash;0.94) had lower riSD than those who were employed. Vaccination showed no overall association with riSD (0.99, 95% CI 0.93\u0026ndash;1.06), though heterogeneity emerged by age and sex. Weaker stringency was strongly associated with higher riSD (1.34, 95% CI 1.31\u0026ndash;1.37).\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e\u003cp\u003eIIV in non-household contacts was shaped by age, household composition, and employment status, but not by vaccination status. Children and adolescents, living in larger households, and assessments during periods of weaker policy stringency exhibited greater IIV, while retired, housemakers, and unemployed individuals showed lower IIV. Vaccination did not have a consistent effect. Analyses relying solely on average contacts may misrepresent risk when IIV is high. Both models and policies should account for IIV, not just mean contact levels.\u003c/p\u003e","manuscriptTitle":"Intraindividual variability in non-household contacts: a German longitudinal study, April 2020–December 2021","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-20 08:13:38","doi":"10.21203/rs.3.rs-7796845/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-12-15T11:25:07+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-10T07:30:27+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-02T16:14:49+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-28T02:42:46+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-26T05:30:30+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-25T16:24:21+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-25T07:42:52+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"207006771043615327732556747941305238783","date":"2025-11-19T20:47:27+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"263851175030055668031412708916467699013","date":"2025-11-19T11:49:06+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"74784936559643388565589954712180413171","date":"2025-11-17T14:02:40+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"279330634687207242215458383112316846875","date":"2025-11-17T09:02:27+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"325301611087644805904095354642098617358","date":"2025-11-13T14:27:48+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"48353358199929221360632019789556251066","date":"2025-11-12T04:52:57+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"92153579819451522257223931572467434963","date":"2025-11-10T22:18:57+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-11-10T03:36:13+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-10-10T16:55:10+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-10-09T23:56:39+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-10-09T23:56:22+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Infectious Diseases","date":"2025-10-07T07:12:12+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-infectious-diseases","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"infd","sideBox":"Learn more about [BMC Infectious Diseases](http://bmcinfectdis.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/infd","title":"BMC Infectious Diseases","twitterHandle":"#bmcinfectdis","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"5e5f99b6-e988-403a-b174-e5076bae7de2","owner":[],"postedDate":"November 20th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-02-23T16:08:21+00:00","versionOfRecord":{"articleIdentity":"rs-7796845","link":"https://doi.org/10.1186/s12879-026-12940-4","journal":{"identity":"bmc-infectious-diseases","isVorOnly":false,"title":"BMC Infectious Diseases"},"publishedOn":"2026-02-21 15:59:40","publishedOnDateReadable":"February 21st, 2026"},"versionCreatedAt":"2025-11-20 08:13:38","video":"","vorDoi":"10.1186/s12879-026-12940-4","vorDoiUrl":"https://doi.org/10.1186/s12879-026-12940-4","workflowStages":[]},"version":"v1","identity":"rs-7796845","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7796845","identity":"rs-7796845","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}