{"paper_id":"d53ef098-3864-4648-8faa-a46df9c07269","body_text":"Persistent Multi-System Impairments Detected by Wearable Monitoring in Long COVID Cases reporting Fatigue up to Three Years Post-Infection | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Persistent Multi-System Impairments Detected by Wearable Monitoring in Long COVID Cases reporting Fatigue up to Three Years Post-Infection Michele Orini, Callum Stuart, Alexandra Jamieson, Alba Fernandez-Sanles, and 18 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7876232/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Long COVID remains poorly understood and previous studies have mainly focused on describing symptoms or single-system health deficits in severe cases. In the CONVALESCENCE case-control study, 306 community-based participants (aged 57 [IQR 41–63], 80% female) were enrolled 2.0 [1.7, 2.5] years post-infection and were monitored for 12 months using a smartwatch and apps to assess mental health, cognitive function, physical activity, exercise capacity, cardiorespiratory function, autonomic function and sleep. Compared to controls (N = 152), long COVID cases with fatigue-cluster symptoms (N = 50) presented small persistent deficits across all systems with no evidence of recovery over 12 months. Participants with fatigue-cluster symptoms not attributable to long COVID (N = 28) showed a similar trend, while participants with prior long COVID but no fatigue-cluster symptoms (N = 76) were not different from controls. The lack of significant recovery between years two and three post-infection demonstrates protracted multi-system deficits for fatigue-cluster cases and highlights ongoing needs to better understand and manage long COVID. Health sciences/Medical research/Epidemiology Health sciences/Diseases/Infectious diseases Post-acute sequelae of COVID-19 long COVID wearable mobile health cognitive function physical activity cardiovascular function mental health sleep Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Main Long COVID is a complex, multisystem, condition characterized by a constellation of symptoms following SARS-CoV-2 infection. It may affect up to 400 million individuals globally and has an estimated economic impact equivalent to about 1% of the global economy 1 , 2 . No diagnostic test or effective treatment is available and a better understanding of health trajectories in long COVID is needed to improve patient care and support. Studies investigating long COVID have mainly focused on characterizing or tracking symptoms through surveys 3 – 6 and those that have investigated physiological parameters have mainly focused on single system deficits (e.g. mental health or cognitive function) 7 – 10 often in severe, hospitalised, cases 7 , 11 – 17 . However, around 90% of long COVID cases had a SARS-CoV-2 infection for which they were not hospitalised 1 , 18 . This limits our understanding of long COVID and poses the risk of constructing an evidence base biased towards more severe or specific initial presentations and less focused on non-hospitalised individuals with symptoms clustered around fatigue, cognitive function and mental health 19 . The lack of multi-domain prospective longitudinal data means that the long-term trajectory of long COVID following community-managed SARS-CoV-2 infection is unclear. The prognosis for such cases is also uncertain. Data from surveys have shown evidence of deficits in mental health, and in perceived physical and cognitive functions up to two years post-infection 20 , with some symptoms worsening over time 7 . Persisting cognitive function deficits were also reported in large population-based samples in which long COVID cases and controls were followed for two years post-infection 9 , 10 . Very limited data are available on health trajectories beyond 2 years after infection, and current knowledge is based on observational studies assessing risk of adverse health outcomes rather than continuous monitoring of health 21 , 22 . As most of the available data come from questionnaires, it is still unclear which physiological process underpins the symptomatology of long COVID, and this limits our understanding of and possible response to this condition. CONVALESCENCE (“COroNaVirus post-Acute Long-term EffectS: Constructing an evidENCE base”) is a case-control study of long COVID that continuously monitored multi-system function and trajectories of health in people suffering from SARS-CoV-2 infection who were not hospitalised. An innovative mobile health framework including a smartwatch, mobile phone applications, and frequent remote tests and surveys 10 , 23 , 24 was developed to simultaneously and continuously monitor mental health, cognitive function, sleep, cardiorespiratory health, physical activity, exercise capacity and autonomic nervous system function (Fig. 1 ) for 12 months, covering on average the period from the second to the third year post-infection. Long COVID cases presenting with fatigue-cluster symptoms were compared to three control groups defined by the presence or absence of previous SARS-CoV-2 infection and symptoms. Overall, this study captures trajectories of health across multiple systems with unprecedented granularity and demonstrates persistent multiple health deficits three years post infection in long COVID cases with a history of fatigue. Results The CONVALESCENCE study recruited from two established population-based cohorts in the UK, the Avon Longitudinal Study of Parents and Children 25 , 26 (ALSPAC; REC: 21/SC/0030) and TwinsUK 27 (REC: 19/NW/0187). 349 adults were invited to clinical assessments between September 2021 and May 2023, and 306 (80% female, median 57 [IQR 41–63] years) participated in the study (Fig. 1 ). Participants were divided into four groups depending on whether they had reported having long COVID (LC + or LC-) before presenting to the clinic visit and whether they presented fatigue symptoms (fatigue, or shortness of breath, or muscle ache) when they entered the study at the clinic visit (S + or S-). Out of 141 participants who reported LC symptoms during the recruitment phase (Supplementary Fig. 9), 50 presented with persistent fatigue-cluster symptoms (LC + S+) at the clinic visit and were considered cases. The primary control group (LC-S-, N = 152) did not report having had long COVID and remained free of fatigue-cluster symptoms. Other control groups included those who recovered from long COVID before assessments, i.e. those who reported long COVID in the past but were free of fatigue symptoms at the clinic visit (LC + S-, N = 76), and those who presented in clinic with fatigue-cluster symptoms not attributable to long COVID (LC-S+, N = 28). Initial assessment The description of single and aggregate outcome measures captured by the smartwatch, the Mass Science app and the Cognitron web-based cognitive test is reported in Supplementary Table 1–2. On average, participants had estimated first infections by SARS-CoV-2 virus 2.0 (1.7, 2.5) years before the clinic visit, between January 2020 and November 2021, when alpha and delta were the dominant variants in the UK. Among participants who reported long COVID, only 1 had been vaccinated before infection, and only 4 remained unvaccinated at the time of the clinic visit. Groups were generally similar in terms of age, sex, vaccination status, and comorbidities, although LC + S + participants had 12% higher BMI than controls (p = 0.01) (Table 1 ). Cardiovascular disease and diabetes were rare (< 5 in all groups). The initial assessment period included the first 30 days of monitoring. During this period, median smartwatch wearing time was 93% and 90% in LC-S- controls and LC + S + cases, respectively (p = 0.04) (Table 1 ). Compared to controls (LC-S-), cases (LC + S+) presented differences in at least one of each system’s outcome measure (Table 1 ), with large median differences for intensity of depression and anxiety and moderate to small median differences for cognitive function (accuracy − 77%), sleep duration variability (+ 23%), heart rate variability (8%), heart rate reserve (-6%), number of steps per day (-5%), walk speed (-5%) and resting heart rate (+ 2 bpm). Table 1 Demographic data and outcome measures at initial assessment. Data is provided as median (interquartile range) or number (%). BMI: Body mass index; HR: Heart rate; RR: Respiratory rate; WT: walk test; HRV: Heart rate variability; Wear. T.: Wearing time. Data reported in bold show p < 0.05 with respect to LC-S- (two-tailed Mann Whitney U test) . Characteristics LC- S- (152) LC- S+ (28) LC + S- (76) LC + S+ (50) Age (years) 58 (31, 63) 59 (53, 66) 56 (46, 62) 55 (47, 61) Sex (Male) 35 (23%) 6 (21%) 12 (16%) 8 (16%) BMI (kg/m2) 25 (22, 29) 27 (25, 28) 25 (22, 28) 28 (23, 34) Vaccination 144 (95%) 25 (89%) 71 (93%) 48 (96%) Ever smoker 56 (37%) 12 (43%) 30 (39%) 16 (32%) Asthma 28 (21%) 5 (20%) 11 (16%) 11 (23%) Hypertension 21 (17%) 6 (27%) 13 (20%) 11 (24%) Higher education 93 (61%) 15 (54%) 45 (59%) 27 (54%) Mental Health LC- S- (143) LC- S+ (27) LC + S- (73) LC + S+ (50) Satisfaction (%) 40.0 (35.2, 44.5) 37.5 (32.5, 40.0) 40.0 (34.3, 43.8) 37.8 (33.0, 41.7) Anxiety (%) 8.93 (1.90, 19.52) 11.11 (4.76, 23.81) 8.73 (1.90, 23.69) 13.7 (7.14, 31.8) Depression (%) 8.33 (4.17, 19.31) 22.2 (10.0, 29.2) 11.25 (4.17, 23.26) 22.9 (10.4, 35.0) Cognitive Function LC- S- (125) LC- S+ (25) LC + S- (66) LC + S+ (47) Resp. Time (nu) -0.41 (-0.86, 0.12) 0.07 (-0.68, 0.37) -0.33 (-0.81, 0.02) -0.26 (-0.77, 0.62) Accuracy (nu) 0.35 (-0.06, 0.76) -0.06 (-0.43, 0.54) 0.32 (-0.09, 0.68) 0.08 (-0.52, 0.53) Sleep LC- S- (136) LC- S+ (25) LC + S- (70) LC + S+ (48) Out of Range (%) 42.6 (28.6, 55.6) 47.4 (36.7,61.1) 43.9 (27.6, 59.1) 50.0 (30.4, 59.2) Variability (h) 1.03 (0.84, 1.34) 1.12 (0.94, 1.31) 1.09 (0.85, 1.39) 1.27 (0.92, 1.59) Exercise Capacity LC- S- (125) LC- S+ (21) LC + S- (61) LC + S+ (40) HR Reserve (bpm) 88.0 (79.6, 101.6) 84.7 (72.5, 93.0) 87.8 (75.0, 102.5) 82.9 (71.3, 94.0) Steps in WT (n) 756 (722, 788) 693 (671, 765) 750 (705, 799) 717 (669, 741) Walk speed (mps) 1.84 (1.70, 1.99) 1.88 (1.67, 1.97) 1.83 (1.66, 1.94) 1.75 (1.56, 1.85) Physical Activity LC- S- (137) LC- S+ (25) LC + S- (70) LC + S+ (47) Steps/day (K) 8.26 (6.62, 10.7) 7.10 (4.83, 9.53) 9.77 (6.56, 11.6) 7.32 (5.18, 9.59) Active Time (%) 7.40 (5.64, 9.70) 7.07 (4.68, 9.19) 7.51 (5.97, 8.95) 6.14 (4.94, 8.25) Cardiorespiratory LC- S- (145) LC- S+ (26) LC + S- (73) LC + S+ (48) Rest HR (bpm) 50.0 (46.0, 54.0) 52.8 (49.0, 59.0) 49.0 (45.8, 52.0) 52.0 (48.2, 54.5) Rest RR (brpm) 8.05 ( 5.90, 9.31) 8.06 ( 5.54, 8.99) 7.98 ( 6.40, 9.19) 8.62 ( 7.68, 9.36) Autonomic Function LC- S- (145) LC- S+ (26) LC + S- (73) LC + S+ (48) HRV SDANN (bpm) 13.0 (11.7, 14.3) 12.7 (11.2, 14.7) 12.7 (11.6, 13.9) 12.0 (10.9, 13.5) HRV SDSD (bpm) 9.25 (8.65, 9.77) 8.78 (8.29, 9.26) 9.38 (8.77, 10.09) 8.79 (8.21, 9.87) Validity LC- S- (145) LC- S+ (26) LC + S- (73) LC + S+ (48) Wear. T Day (%) 92.9 (85.5, 95.5) 93.2 (88.7, 95.1) 91.2 (81.2, 95.3) 90.1 (83.7, 92.7) Wear. T Night (%) 89.8 (76.5, 97.1) 89.9 (74.6, 97.2) 90.3 (71.2, 96.6) 89.1 (77.4, 95.0) Correlation heatmaps between outcome measures during the initial assessment revealed distinct patterns between cases (LC + S+) and controls (LC-S-) (Supplementary Fig. 1). For example, in cases, the response time to the cognitive tests was moderately inversely correlated to heart rate reserve, walking speed and heart rate variability, suggesting a possible association between cognitive function, exercise capacity and autonomic regulation. Associations between outcome measures and participant status were assessed using linear regressions. After statistically adjusting for age, sex, BMI, education (cognitive function only) and mean heart rate (cardiac autonomic nervous system only), cases (LC + S+) presented deficits with respect to controls (LC-S-) in all health systems (Fig. 2 , Supplementary Table 3). Aggregate mental health outcomes combining anxiety, depression and satisfaction were the most reduced (β: -0.53 (-0.82, -0.24), p < 0.01), but deficits were also present in aggregate outcomes for cognitive function (β: -0.39 (-0.60, -0.18), p < 0.01), exercise capacity (β: -0.31 (-0.55, -0.08), p < 0.01), cardiorespiratory function (β: -0.39 (-0.60, -0.18), p < 0.01), sleep (β: -0.34 (-0.63, -0.04), p = 0.02), physical activity (β: -0.34 (-0.63, -0.05), p = 0.02), and autonomic function (β: -0.24 (-0.49, 0.01), p = 0.06). Similarly, LC-S + participants tended to show deficits across all aggregate outcome measures, although with smaller effect size and wider confidence intervals, whereas LC + S- participants were similar to controls (Fig. 2 ). These system deficits were underpinned by differences in most single outcome measures (Fig. 3 , Supplementary Table 4). Compared to controls, all components of mental health (anxiety, depression and reduced satisfaction) were more adverse, while all components of cognitive function (aggregate response time and accuracy), and exercise capacity (heart rate reserve, walking speed and number of steps during walk tests) were decreased. Most of the individual cognitive tasks were performed with lower accuracy, or longer response time, or both (Supplementary Table 5). Differences were detected in motor control, immediate and delayed memory, processing speed and visuo-spatial abilities, but not in mental manipulation and verbal reasoning. Deficits in cardiorespiratory function, autonomic nervous system, physical activity and sleep were mainly driven by a higher resting heart rate, smaller long-term heart rate variability (SDANN), reduced active time, and increased sleep duration variability, respectively. Trajectories of health in long COVID On average, participants were continuously monitored using the smartwatch for 12 (9, 12) months, during which they conducted 9 (3, 16) walk tests, 11 (2, 26) surveys for mental health and 12 (5, 22) cognitive tests (Supplementary Table 6). No difference was found in smartwatch wearing time or in the number of tests/surveys between the four groups. Adjusted linear regressions using data collected at 12 months showed that, with respect to LC-S- controls, LC + S + cases still presented deficits across most of the health systems (Fig. 2 , supplementary Table 3). Aggregate outcome measures remained significantly reduced for mental health (β: -0.49 (-0.92, -0.05), p = 0.03), exercise capacity (β: -0.49 (-0.82, -0.15), p < 0.01), and cognitive function (β: -0.43 (-0.83, -0.03), p = 0.04). Aggregate outcome measures for sleep (β: -0.36 (-0.77, 0.04), p = 0.08), autonomic function (β: -0.25 (-0.58, 0.08), p = 0.13) and cardiorespiratory system (β: -0.17 (-0.49, 0.14), p = 0.28) were still reduced to similar extents, although differences were no longer statistically significant probably on account of a reduced sample size. No differences were observed in physical activity (β: -0.13 (-0.52, 0.27), p = 0.53). These persisting system deficits were mainly driven by increased depressive symptoms, reduced heart rate reserve, increased sleep variability, increased resting heart rate and reduced heart rate variability (Fig. 3 , Supplementary Table 4). Trajectories of outcome measures were predicted using adjusted linear mixed-effect regressions. Aggregate outcome measures for cases did not converge over time towards control values (Fig. 4 ). The regression coefficients representing the interaction between participant status and time (i.e. the trajectory slope) provided no convincing evidence of recovery for any of the outcome measures monitored during the study. On the contrary, a trend towards a more marked deterioration in sleep, and possibly in exercise capacity and autonomic function was observed in cases as compared to controls (Figs. 3 & 4 ). Temporal variability in outcome measures Long COVID patients often report temporal fluctuations in their symptoms. Temporal variability in psychological, cognitive and physiological outcome measures was assessed by measuring the standard deviation of the linear mixed-effect model residuals. Linear regressions adjusted for age, sex, BMI and education showed that temporal variability in cognitive function was larger in LC + S + cases compared with LC-S- controls (Supplementary Table 7). These differences were however attenuated after adjusting for the mean level of cognitive function over the monitoring period, which was reduced in cases. Individuals who recovered from long COVID (LC + S-) did not show differences in temporal variability of outcome measures, while participants with fatigue-cluster symptoms not attributable to long COVID (LC-S+) showed larger temporal variability in psychological outcome measures. This difference was also attenuated after adjusting for the mean value of psychological outcome measures over the monitoring period. Pairwise, within-subject, correlation coefficients between outcome measures were estimated to assess the strength of the coupling between the temporal fluctuations in outcome measures (Fig. 5 ). Only participants who completed the 12-month monitoring period were included in the analysis to capture consistent patterns. Outcome measures representing physical activity, exercise capacity and autonomic function were positively coupled in both cases and controls. For example, among controls, steps per day correlated with both heart rate variability and heart rate reserve (median [interquartile range] correlation coefficients equal to 0.33 [0.10, 0.56], p < 0.01, and 0.17 [-0.05, 0.44], p < 0.01, respectively). Some correlations were statistically significant among LC + S + cases but not among LC-S- controls (Fig. 5 A). The accuracy of cognitive tests was inversely correlated to depressive symptoms (cc = -0.54 [-0.66, -0.16], p = 0.02, Fig. 5 B) and resting heart rate (cc=-0.18 [-0.41,-0.10], p = 0.02); depression was also negatively correlated with steps per day (cc=-0.28 [-0.45, 0.02, p = 0.04]), and fluctuations in sleep duration was inversely correlated with heart rate variability (cc=-0.26 [-0.41, 0.04], p = 0.01). The median correlation coefficient between accuracy of cognitive tests and anxiety (cc=-0.32 [-0.48, 0.11], p = 0.29) and satisfaction (cc = 0.33 [-0.10, 0.50], p = 0.30) was relatively high but did not reach statistical significance, probably on account of small sample size (N = 11). Sensitivity analyses Several sensitivity analyses were conducted. Differences between aggregate outcome measures were assessed after matching cases and controls for BMI (by identifying the control subgroup which minimizes BMI differences). Results were very similar to those obtained in the primary analysis (Supplementary Fig. 2), demonstrating that a slightly larger BMI in cases does not explain health deficits. We assessed the impact of attrition during the monitoring period. Restricting the analysis to study participants who were continuously monitored for 12 months (N = 202) did not substantially modify the results. In this subgroup, aggregate and single outcome measures at the initial assessment and at 12 months, as well as their predicted trends over time showed patterns similar to those observed in the primary analysis, albeit with larger confidence intervals attributable to smaller sample size (Supplementary Fig. 3–5, supplementary table 8–9). We then compared trajectories of outcome measures when the starting point was defined as the date of the reported infection, rather than the date of the clinic visit. No evidence of significant recovery in aggregate outcome measures was found in cases (LC + S+) as compared to controls (LC-S-) (Supplementary Fig. 6). Qualitatively, in cases, cognitive function slightly improved over the years, while exercise capacity, autonomic function and sleep tended to get worse. Finally, we assessed how the definition of cases and controls could have impacted the results. Broadening the spectrum of symptoms used to define cases by including participants that reported confusion or joint pain in addition to fatigue, shortness of breath and muscle pain, did not alter the results and only slightly increased difference in mental health outcomes (Supplementary Fig. 7–8, supplementary table 10–11). This is probably because 86% of study participants reporting confusion or joint pain also reported fatigue cluster symptoms. Discussion Compared to controls with no history of long COVID or fatigue, long COVID cases that reported persistent fatigue cluster symptoms two years after infection presented ongoing deficits across all investigated systems (mental health, cognitive function, exercise capacity, physical activity, autonomic nervous system, cardiorespiratory function and sleep). No evidence of recovery was identified in the trajectory of outcome measures over the following year with most of the deficits persisting after 12 months. Participants who had recovered from long COVID or did not report fatigue cluster symptoms at recruitment (LC + S-) did not show any discernible health deficits. In contrast, health deficits in cases (LC + S+) were similar to those in participants presenting with fatigue cluster symptoms not attributed to long COVID (LC-S+). Our data suggest that for long COVID cases that remained symptomatic 2 years post infection, fatigue cluster symptoms coexist not only with worse mental health and cognitive outcomes, as recently reported in hospitalised cases 7 , but also with impaired physiological outcome measures, namely exercise capacity, physical activity, cardiac autonomic and cardiorespiratory function and sleep quality. Importantly, in cases reporting fatigue, the lack of recovery in most outcome measures over 12 months strongly suggests that time may not be a ‘cure’, and further investigation on possible treatment strategies is urgently needed. Several studies have found persisting cognitive and mental health symptoms at least 1 year after infection in post-hospitalised 7 , 11 , 28 and community-based samples 9 , 10 , 29 , and our data demonstrate that these deficits may persist for over 3 years. Contrary to a previous report that assessed subjective cognitive function 2 years post infection 20 , we did not observe cognitive deficits in those who had recovered from long COVID (LC + S-). This discrepancy, however, may be partially explained by differences between subjective and objective cognitive assessment. Objectively measured deficits in global cognition in individuals who recovered from long COVID were also identified in a large cross-sectional community sample study, in some cases up to two years post-infection 10 . The magnitude of these deficits was, however, small and their detection may require much larger samples than used in our study. Reduced exercise capacity 13 , 29 – 32 and diminished levels of objectively measured physical activity 33 – 35 in long COVID have been previously reported and are considered hallmarks of long COVID. Our data demonstrate that in those reporting fatigue cluster symptoms two years post infection, reduced exercise capacity, as assessed by maximum walk speed, number of steps in a 6-minute walk test and heart rate reserve, is likely to persist for at least another year. The magnitude of deficits in exercise capacity was however relatively small. Interestingly, after 12 months, cases and controls showed similar aggregate outcome measures of physical activity, but the difference in aggregate outcome measures of exercise capacity remained evident, suggesting that reduced exercise capacity may not be fully explained by reduced physical activity. Cardiorespiratory function also differed between cases and controls, with both resting heart rate and respiratory rate being modestly higher in cases than controls. This is consistent with the results of a large longitudinal wearable study that demonstrated an increase in resting heart rate following SARS-CoV-2 infection 24 . Heart rate variability, a non-invasive and qualitative measure of cardiac autonomic function, was also reduced in cases as compared to controls, and this difference was maintained during the monitoring period. Differences remained statistically significant after adjusting for mean heart rate, therefore reducing the possibility that these differences were driven by increased resting heart rate. Dysautonomia is a frequent finding in long COVID cases 32 , 36 – 38 , and our data demonstrate a protracted effect in those individuals with long-term fatigue cluster symptoms. Compared to controls, cases also showed greater objectively measured sleep irregularity, which persisted through the entire monitoring period. This is in agreement with data from surveys of individuals with long COVID, which indicate a prevalence of poor sleep quality above 50% 39 , with irregular sleep being mentioned as the most commonly reported disturbance 40 and poor sleep quality correlating with symptom severity and reduced quality of life 41 . Sleep irregularity is an important and perhaps overlooked consequence of long COVID, which has been associated with reduced cardiometabolic health and cognitive function 42 , 43 . While in the long COVID cases monitored in this study sleep variability was not associated with reduced cognitive function, sleep duration trends were inversely correlated with heart rate variability, suggesting that periods of increased sleep irregularity coincided with periods of reduced cardiac autonomic function. The multimodal longitudinal design of the study allowed us to analyse outcome measure fluctuations over the 12-month monitoring period and their correlations. Cognitive function was the only outcome measure showing larger fluctuations in long COVID cases as compared to controls. This difference was attenuated after adjusting by the mean cognitive outcome measure suggesting that cognitive function in long COVID cases is not consistently reduced but rather fluctuates around lower levels. The strength of the coupling between fluctuations in outcome measures was estimated using correlation analysis. A small-to-moderate positive correlation between measures of physical activity, exercise capacity, and cardiac autonomic function was identified in both cases and controls. In long COVID cases, cognitive function was inversely coupled with depressive symptoms and resting heart rate, and sleep was inversely coupled with autonomic function. The strength of these correlations was low-to-moderate and based on a small number of cases (only participants with complete outcome measures over the 12-month monitoring period were included). While these correlations do not imply causal links, they may reflect a dynamic impact of long COVID on multiple physiological, cognitive and psychological measures which warrants further investigation. Although this study focused on physiological metrics captured via remote monitoring, emerging molecular evidence provides complementary insight into the biological mechanisms underlying long COVID. Recent work has identified distinct inflammatory signatures (including neuroinflammation) associated with persistent symptoms such as breathlessness and fatigue 44 . Additionally, altered microRNA profiles have been implicated in immune dysregulation and cellular stress responses 45 . While molecular data were not examined in this study, our findings may reflect downstream effects of sustained inflammation or immune dysregulation. Future studies integrating wearable-derived phenotypes with molecular profiling could help delineate pathophysiological subtypes and inform targeted interventions. To the extent of our knowledge, this is the first study to simultaneously and continuously monitor multiple physiological, cognitive and psychological health outcomes in long COVID for up to one year. The study utilised a smartwatch which has been extensively validated for monitoring of physiological parameters including heart rate, steps, distance, physical activity, and sleep 46 – 48 , a battery of state-of-the-art cognitive tests commonly employed in long COVID studies 7 – 11 and an app specifically designed to communicate with study participants and to collect survey data 24 , 49 . Other strengths include the case-control design and recruitment of community-based participants. The main limitations include a moderate sample size and inclusion of individuals infected mainly by alpha and delta variants circulating before the roll out of vaccination programmes. However, because vaccination reduced both the number of infections and the prevalence of long COVID, our sample may be representative of the majority of long COVID cases with persistent symptoms, who were infected before mass vaccination. Another important limitation is the lack of data characterising health outcome measures before the initial assessment. Although participants were carefully recruited and separated into case and control groups based on recently reported symptoms attributable or not attributable to long COVID (and with serological testing for SARS-Cov-2 infections), we cannot exclude that at least part of the health deficits identified in this study may have already existed before the initial assessment. Methods Study participants The CONVALESCENCE study (REC: 21/SC/0235) is described in detail elsewhere 50 . In brief, the study was set up in 2021 to establish an evidence base for long COVID and it recruited from two established population-based cohorts, the Avon Longitudinal Study of Parents and Children (ALSPAC; REC: 21/SC/0030; Supplementary text) 25 , 26 and Twins UK study (REC: 19/NW/0187) 51 . 349 study participants from these two UK cohorts completed a survey to establish their health status and were invited to a clinic visit at the Bloomsbury Centre for Clinical Phenotyping (BCCP) in London. Visits were held between October 2021 and May 2023. During the visit, participants were offered a smartwatch (Garmin Vivoactive 4) and the possibility of installing two apps on their phones: Garmin Connect, which automatically collects data from the watches, and the Mass Science app 24 , 49 , which was used to collect survey data and send notifications to study participants as well as to enable the use of the Cognitron test, a platform for quantitative remote assessment of cognitive function 10 . Participants were asked to wear the watch as much as possible, to complete mental health and quality of life surveys and 6-minute walking tests (6MWT) frequently (every fortnight for the first 6 weeks and every 4 weeks afterwards). Participants were included in the study if they gave informed consent and if they provided at least one valid data point, defined as completing at least a survey, or a cognitive test, or wearing the watch for at least 6 hours during the daytime for a minimum of 7 days. Three-hundred and six (N = 306) participants complied with this requirement and were included in the study (Fig. 1 ). Long COVID case-control definitions Before the clinic visit, participants were classified into those who experienced and did not experience long COVID (LC + and LC+). Long COVID was defined as evidence of SARS-CoV-2 infection (self-reported positive PCR test or self-reported suspected SARS-CoV-2 infection with later evidence of natural infection from consideration of antibody testing and vaccination status) and self-reporting of persistent long COVID symptoms including any of fatigue, respiratory, cognitive, cardiovascular, and gastrointestinal cluster symptoms lasting more than 4 weeks and attributable to the infection (Supplementary Fig. 9). Those who did not experience long COVID (LC-), either did not report long COVID symptoms or, in case of reporting symptoms, they had no evidence of SARS-CoV-2 infection. Among participants reporting SARS-CoV-2 infection, 90% were infected between January 2020 and November 2021, and the last infection was recorded in April 2022. Participants were seen in clinic between October 2021 and May 2023 where they were divided into four groups based on whether they had long COVID (LC + and LC-) and whether they presented fatigue cluster symptoms (fatigue, or breathlessness or muscle ache) (Fig. 1 ). Cases were defined as those LC + participants still presenting fatigue cluster symptoms (N = 50, LC + S+). Controls were defined as participants who reported neither long COVID nor fatigue cluster symptoms (N = 152, LC-S-). This included those with no infection or infection and symptoms lasting < 4 weeks. Additionally, two other control groups were considered: Participants who had reported long COVID but did not present fatigue cluster symptoms in clinic (N = 76, LC + S-), including those with LC characterised by symptoms unrelated to fatigue and those no longer reporting fatigue cluster symptoms, and Participants who reported not having had long COVID, with a negative antibody test, but who presented fatigue cluster symptoms assumed not to be attributable to SARS-CoV-2 infection (N = 28, LC-S+). Demographics and anthropometric measurements Participant age, sex, long COVID symptomatology, medical history and current medications were collected by questionnaire in a secure electronic Case Report Form (eCRF) using REDCap (Research Electronic Data Capture). Height was measured using a stadiometer (Seca217, Seca, Germany) to the closest centimetre and weight was measured in kilograms using digital bio-impedance scales (BC-418 or MC-780MA, Tanita, USA) to calculate body mass index (BMI). Smartwatch measurements Data was transferred to a Trusted Research Environment through an IT infrastructure that shares the backend of RADAR BASE 23 . Supplementary Table 1 provides detailed definitions of outcomes based on smartwatch measurements. The watch recorded heart rate every 15 seconds, respiratory rate, number of steps and intensity of physical activity every 15 minutes, and sleep duration once a day. Smartwatch measurements were used to derive metrics representing four distinct systems: Physical activity, exercise capacity, cardiorespiratory health, cardiac autonomic function, and sleep. Physical activity was quantified by the number of steps per day and the proportion of time spent in active or very active activity as defined by Garmin’s proprietary software. Exercise capacity was assessed using the number of steps covered during the remote 6-minute walk tests, the maximum speed during any walking activity lasting between 4 and 60 minutes, and by heart rate reserve 52 , defined as the difference between maximum and minimum heart rate between 7 AM and 11 PM (night-time values were not considered because some participants did not use the device overnight). The number of steps instead of distance covered during the remote 6-minute walk test was used because number of steps were found to be measured more accurately than distance using the Garmin watch 46 . Walk activities for which the maximum speed was > 10 Km/h, or number of steps was < 100, or total elevation loss/gain was > 50 m were not considered. Cardiorespiratory function was captured by resting heart rate and respiratory rate (minimum daily values). Cardiac autonomic function was captured by two measures of heart rate variability: The standard deviation of 5-minute average heart rate (SDANN) and the standard deviation of successive 15-second differences measured daily. Mental health and wellbeing Study participants were asked to complete standard questionnaires to assess anxiety (the Generalised Anxiety Disorder, GAD-7), depression (Patient Health Questionnaire, PHQ-8) and life satisfaction (Office National Statistics, ONS-2) through the Mass Science app 24 , 49 . Questionnaires were presented to participants every 4 weeks, and results were rescaled from 0 to 100. Cognitive assessment Cognitive function was quantified through Cognitron 7 , 9 – 11 , a battery of web-based tasks which were presented through the Mass Science app. More information is available on the website https://www.cognitron.co.uk/ . Eight tasks were presented to assess memory (immediate and delayed), visuospatial abilities/attention, processing speed, verbal reasoning and spatial planning. Information related to performance and compliance was collected automatically and combined in a normalized summary score summarizing accuracy and median response time, which were used to characterize cognitive function. Data processing The time series representing the temporal trajectory of each outcome measure over the 12-month period was computed by averaging measurements collected over 30 (for all smartwatch measures, except for those derived from walk tests) or 60 (for mental health and cognitive function) consecutive days. In the case of mental health, cognitive function and walk tests, a longer time frame was necessary because surveys and tests were collected less frequently than smartwatch data. To assess health deficits, an aggregate outcome measure per physiological domain was computed by averaging normalized single outcome measures (formula shown in Supplementary Table 2). Single outcome measures that were inversely associated with the global outcomes (e.g. anxiety, depression, response time) were inverted before averaging. Statistical Analysis Data were summarized as median (interquartile range) or number (percentage) and differences between the 4 groups and pairwise comparisons were assessed using Kruskal Wallis and Wilcoxon ranksum tests, respectively. Associations between participant status (LC + S+, LC + S-, LC-S+, and LC-S-) and outcome measures at the beginning of the study (first data point in the time-series) and at 12 months (last data point in the time series) were investigated using linear regressions, adjusted for age, sex, and BMI, which were considered confounders. Cognitive outcome measures were further adjusted for education (having/not having higher education or professional equivalent), while cardiac autonomic outcomes (heart rate variability) were further adjusted for mean heart rate. Associations between outcome measures and status are reported using controls (LC-S-) as reference category. Associations between participant status and the trajectory of outcome measures were investigated using linear mixed-effect models, where status, time and status-time interaction were treated as fixed effects and uncorrelated random intercepts and slopes were allowed per participants and time. The model was also adjusted for age, sex, BMI, education (cognitive outcomes only) and mean heart rate (autonomic outcomes only). Trajectory of aggregate outcome measures were predicted based on this model, with estimates and 95% confidence interval plotted against time. To account for a learning effect after repeating the battery of cognitive tests multiple times, the trajectories of cognitive outcome measures were corrected by subtracting the controls’ slope (LC-S-). In primary analysis, time zero was defined as the date of the clinic visit, whereas in sensitivity analysis, time zero was defined as the reported date of COVID infection. Participants who did not report a COVID infection were assigned a random time zero from a normal distribution with mean and standard deviation equal to that of the visit date distribution in the latter analysis. Associations between status and the temporal variability in the outcome measures were measured by regressing the standard deviation of the conditional residuals of the mixed-effect model on status and number of valid data points. Participants with less than 3 data points were excluded from this analysis. The strength of the coupling between outcome measure fluctuations was assessed measuring within subject Spearman’s correlation coefficients between outcome measures. In this analysis, data averaged over 30 or 60 days (as described above) was used. Outcome measures averaged over 60 days were resampled every 30 days using nearest value interpolation so that every outcome measure was described by a time-series of at most 12 data-points. To assess whether the distributions of within-subject correlation coefficients were statistically different from zero, the Wilcoxon signrank test was used. Given the exploratory nature of the study and multiple distinct domains being investigated, no statistical correction for multiple comparisons were applied. Declarations Acknowledgements We are extremely grateful to all the people who took part in this study, and to the past and present members of the research team who collected and managed the data including Lidia Nigrelli, Fintan McArdle, Chelsea Beckford, Clare Davie, Suzanne Williams, Uhuru Lambert, Felicia Huang and Imran Shah. We are extremely grateful to the all participants and Department of Twin Research staff. We are extremely grateful to all the families who took part in this study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes data collection staff, data and administrations staff, technical managers and the technical staff with the Bristol Bioresource Laboratory, based within the University of Bristol. Conflict of Interest NC receives funds from AstraZeneca for serving on data safety and monitoring committees for clinical trials Funding Statement The Characterisation, determinants, mechanisms and consequences of the long-term effects of COVID-19: providing the evidence base for health care services (CONVALESCENCE, COV-LT-0009\\MC_PC20051\\MC_PC_20059) study was jointly funded by the National Institute for Health and Care Research (NIHR) and UK Research and Innovation (UKRI). The views expressed in this publication are those of the author(s) and not necessarily those of NIHR, The Department of Health and Social Care or UKRI. AJ was supported by a British Heart Foundation 4-year PhD studentship (FS/19/63/34902) awarded to UCL. The UK Medical Research Council and Wellcome (Grant ref: 217065/Z/19/Z) and the University of Bristol provide core support for ALSPAC. This publication is the work of the authors and will serve as guarantors for the contents of this paper. A comprehensive list of grants funding is available on the ALSPAC website (http://www.bristol.ac.uk/alspac/external/documents/grant-acknowledgements.pdf). TwinsUK is funded by the Wellcome Trust, Medical Research Council, Versus Arthritis, European Union Horizon 2020, Chronic Disease Research Foundation (CDRF), Zoe Ltd, the National Institute for Health and Care Research (NIHR) Clinical Research Network (CRN) and Biomedical Research Centre based at Guy’s and St Thomas’ NHS Foundation Trust in partnership with King’s College London. BCCP received infrastructure support from the NIHR UCLH Biomedical Research Centre and the BHF. NC and ADH work in a unit that receives support from the UK Medical Research Council (grant number MC_UU_12019/1). NJC was supported by the CONVALESCENCE study [COV-LT-0009]. EJT acknowledges funding from the Wellcome Trust (WT212904/Z/18/Z) and NIHR (CONVALESCENCE grant COV-LT-0009). Data Availability Declaration Due to the sensitive nature of the data collected for this study, data cannot be made publicly available, but requests to access the dataset from qualified researchers trained in human subject confidentiality protocols may be sent to [email protected] . The informed consent obtained from ALSPAC (Avon Longitudinal Study of Parents and Children) participants does not allow the data to be made available through any third party maintained public repository. Supporting data are available from ALSPAC on request under the approved proposal number, B3666. Full instructions for applying for data access can be found here: http://www.bristol.ac.uk/alspac/researchers/access/. The ALSPAC study website contains details of all available data (http://www.bristol.ac.uk/alspac/researchers/our-data/). References Al-Aly Z et al (2024) Long COVID science, research and policy. Nat Med 30:2148–2164 Al-Aly Z (2023) Prevention of long COVID: progress and challenges. Lancet Infect Dis 23:776–777 Shen Q et al (2023) COVID-19 illness severity and 2-year prevalence of physical symptoms: an observational study in Iceland, Sweden, Norway and Denmark. Lancet Reg Health - Europe 35 Whitaker M et al (2022) Persistent COVID-19 symptoms in a community study of 606,434 people in England. Nat Commun 13 Servier C, Porcher R, Pane I, Ravaud P, Tran V-T (2023) Trajectories of the evolution of post-COVID-19 condition, up to two years after symptoms onset. Int J Infect Dis 133:67–74 Tran VT, Porcher R, Pane I, Ravaud P (2022) Course of post COVID-19 disease symptoms over time in the ComPaRe long COVID prospective e-cohort. Nat Commun 13 Taquet M et al (2024) Cognitive and psychiatric symptom trajectories 2–3 years after hospital admission for COVID-19: a longitudinal, prospective cohort study in the UK. Lancet Psychiatry 11:696–708 Hampshire A et al (2021) Cognitive deficits in people who have recovered from COVID-19. EClinicalMedicine 39 Cheetham NJ et al (2023) The effects of COVID-19 on cognitive performance in a community-based cohort: a COVID symptom study biobank prospective cohort study. EClinicalMedicine 62 Hampshire A et al (2024) Cognition and Memory after Covid-19 in a Large Community Sample. N Engl J Med 390:806–818 Wood GK et al (2025) Posthospitalization COVID-19 cognitive deficits at 1 year are global and associated with elevated brain injury markers and gray matter volume reduction. Nat Med 31:245–257 Routen A, Khunti K (2024) Long-term outcomes in hospitalised COVID-19 survivors and future research priorities. Lancet Respir Med 12:7–8 Huang L et al (2022) Health outcomes in people 2 years after surviving hospitalisation with COVID-19: a longitudinal cohort study. Lancet Respir Med 10:863–876 Fernández-de-las-Peñas C, Arias-Navalón JA, Martín-Guerrero JD, Pellicer-Valero OJ, Cigarán-Méndez M (2024) Trajectory of anxiety/depressive symptoms and sleep quality in individuals who had been hospitalized by COVID-19: The LONG-COVID-EXP multicenter study. J Psychosom Res 179 Fernández-De-Las-Peñas C et al (2023) Trajectory curves of post-COVID anxiety/depressive symptoms and sleep quality in previously hospitalized COVID-19 survivors: The LONG-COVID-EXP-CM multicenter study. Psychol Med 53:4298–4299 Fernández-de-las-Peñas C et al (2023) Trajectory of post-COVID brain fog, memory loss, and concentration loss in previously hospitalized COVID-19 survivors: the LONG-COVID-EXP multicenter study. Front Hum Neurosci 17 Fernández-de-las-Peñas C et al (2023) Trajectory of Post-COVID Self-Reported Fatigue and Dyspnoea in Individuals Who Had Been Hospitalized by COVID-19: The LONG-COVID-EXP Multicenter Study. Biomedicines 11 Wulf Hanson S et al (2022) Estimated Global Proportions of Individuals With Persistent Fatigue, Cognitive, and Respiratory Symptom Clusters Following Symptomatic COVID-19 in 2020 and 2021. JAMA 328:1604 Davis HE, McCorkell L, Vogel JM, Topol EJ (2023) Long COVID: major findings, mechanisms and recommendations. Nat Rev Microbiol 21:133–146 Szewczyk W et al (2024) Long COVID and recovery from Long COVID: quality of life impairments and subjective cognitive decline at a median of 2 years after initial infection. BMC Infect Dis 24:1241 Zhang H et al (2024) 3-year outcomes of discharged survivors of COVID-19 following the SARS-CoV-2 omicron (B.1.1.529) wave in 2022 in China: a longitudinal cohort study. Lancet Respir Med 12:55–66 Cai M, Xie Y, Topol EJ, Al-Aly Z (2024) Three-year outcomes of post-acute sequelae of COVID-19. Nat Med 30:1564–1573 Ranjan Y et al (2019) RADAR-Base: Open Source Mobile Health Platform for Collecting, Monitoring, and Analyzing Data Using Sensors, Wearables, and Mobile Devices. JMIR Mhealth Uhealth 7:e11734 Stewart C et al (2024) Physiological presentation and risk factors of long COVID in the UK using smartphones and wearable devices: a longitudinal, citizen science, case–control study. Lancet Digit Health 6:e640–e650 Fraser A et al (2013) Cohort Profile: The Avon Longitudinal Study of Parents and Children: ALSPAC mothers cohort. Int J Epidemiol 42:97–110 Boyd A et al (2013) Cohort Profile: The ‘Children of the 90s’—the index offspring of the Avon Longitudinal Study of Parents and Children. Int J Epidemiol 42:111–127 Verdi S et al (2019) TwinsUK: The UK Adult Twin Registry Update. Twin Res Hum Genet 22:523–529 Wahlgren C et al (2023) Two-year follow-up of patients with post-COVID-19 condition in Sweden: a prospective cohort study. Lancet Reg Health - Europe 28:100595 Peter RS et al (2025) Persistent symptoms and clinical findings in adults with post-acute sequelae of COVID-19/post-COVID-19 syndrome in the second year after acute infection: A population-based, nested case-control study. PLoS Med 22:e1004511 Zheng C et al (2024) Physical exercise-related manifestations of long COVID: A systematic review and meta-analysis. J Exerc Sci Fit 22:341–349 Durstenfeld MS et al (2022) Use of Cardiopulmonary Exercise Testing to Evaluate Long COVID-19 Symptoms in Adults. JAMA Netw Open 5:e2236057 Jamieson A et al (2024) Mechanisms underlying exercise intolerance in long < scp > COVID : An accumulation of multisystem dysfunction. Physiol Rep 12 Diciolla NS et al (2025) Physical Activity and Sedentary Behaviour in People with Long COVID: A Follow-Up from 12 to 18 Months After Discharge. J Clin Med 14:3641 Rosa-Souza FJ et al (2024) Association of physical symptoms with accelerometer-measured movement behaviors and functional capacity in individuals with Long COVID. Sci Rep 14:20652 Plekhanova T et al (2022) Device-assessed sleep and physical activity in individuals recovering from a hospital admission for COVID-19: a multicentre study. Int J Behav Nutr Phys Activity 19:94 Dani M et al (2021) Autonomic dysfunction in ‘long COVID’: rationale, physiology and management strategies. Clin Med 21:e63–e67 Raj SR et al (2021) Long-COVID postural tachycardia syndrome: an American Autonomic Society statement. Clin Auton Res 31:365–368 Ladlow P et al (2022) Dysautonomia following COVID-19 is not associated with subjective limitations or symptoms but is associated with objective functional limitations. Heart Rhythm 19:613–620 Chinvararak C, Chalder T (2023) Prevalence of sleep disturbances in patients with long COVID assessed by standardised questionnaires and diagnostic criteria: A systematic review and meta-analysis. J Psychosom Res 175:111535 Tański W, Tomasiewicz A (2024) Jankowska-Polańska, B. Sleep Disturbances as a Consequence of Long COVID-19: Insights from Actigraphy and Clinimetric Examinations—An Uncontrolled Prospective Observational Pilot Study. J Clin Med 13:839 Alzueta E et al (2022) An international study of post-COVID sleep health. Sleep Health 8:684–690 Zhu B et al (2022) Associations between sleep variability and cardiometabolic health: A systematic review. Sleep Med Rev 66:101688 Qin S, Chee M (2024) The Emerging Importance of Sleep Regularity on Cardiovascular Health and Cognitive Impairment in Older Adults: A Review of the Literature. Nat Sci Sleep Volume 16:585–597 Gao Y et al (2025) Identification of soluble biomarkers that associate with distinct manifestations of long COVID. Nat Immunol 26:692–705 Paval N-E et al (2025) MicroRNAs in long COVID: roles, diagnostic biomarker potential and detection. Hum Genomics 19:90 Jamieson A, Jones S, Chaturvedi N, Hughes AD, Orini M (2024) Accuracy of smartwatches for the remote assessment of exercise capacity. Sci Rep 14:22994 Williams K, Jamieson A, Chaturvedi N, Hughes A, Orini M Validation of Wearable Derived Heart Rate Variability and Oxygen Saturation from the Garmin’s Health Snapshot. Comput Cardiol ( (2010)) (2023)) (2023) Miller DJ, Sargent C, Roach GD (2022) A Validation of Six Wearable Devices for Estimating Sleep, Heart Rate and Heart Rate Variability in Healthy Adults. Sensors 22:6317 Stewart C et al (2021) Investigating the Use of Digital Health Technology to Monitor COVID-19 and Its Effects: Protocol for an Observational Study (Covid Collab Study). JMIR Res Protoc 10:e32587 Jamieson A et al (2025) Cohort profile: characterisation, determinants, mechanisms and consequences of the long-term effects of COVID-19 – providing the evidence base for health care services (CONVALESCENCE) in the UK. BMJ Open 15:e094760 Moayyeri A, Hammond CJ, Valdes AM, Spector TD (2013) Cohort profile: Twinsuk and healthy ageing twin study. Int J Epidemiol 42:76–85 SWAIN, D. P., LEUTHOLTZ, B. C., KING, M. E., HAAS, L. A. & BRANCH, J. D. Relationship between% heart rate reserve and%??VO2reserve in treadmill exercise. Medicine & Science in Sports & Exercise 30, 318–321 (1998) Additional Declarations There is NO Competing Interest. Supplementary Files Supplementary251014.docx Supplementary Figures and Tables Cite Share Download PDF Status: Under Review Version 1 posted 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-7876232\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":false,\"archivedVersions\":[],\"articleType\":\"Article\",\"associatedPublications\":[],\"authors\":[{\"id\":536461363,\"identity\":\"8f79479a-00ba-42bf-9bb4-dd16f7b92f61\",\"order_by\":0,\"name\":\"Michele Orini\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAtklEQVRIiWNgGAWjYLCCBAYJOQkG5gYwR4JYLcYSDIykaAGCxBlEa9FtYD724eEOi/SZ7QcbGH7UMCTObCCgxewAW/KMxDMSubN5EhsYe44xJM4mZIvZAR5jhsQ2idx5EkCH8TYwJM4jVku6HFAL419StCRIA7Uwg2wh7LDDbMkgLYYzexIbDssckzAm7P3jzYcZf7bVyUscP3zw4ZsaG9kZBwhZw4zEPkBCRI6CUTAKRsEowAcA7Cs45SsZIRIAAAAASUVORK5CYII=\",\"orcid\":\"https://orcid.org/0000-0001-5773-0344\",\"institution\":\"King's College London\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Michele\",\"middleName\":\"\",\"lastName\":\"Orini\",\"suffix\":\"\"},{\"id\":536461364,\"identity\":\"9b0d1982-a117-4ca4-9866-719d9f026cd9\",\"order_by\":1,\"name\":\"Callum Stuart\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"KCL\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Callum\",\"middleName\":\"\",\"lastName\":\"Stuart\",\"suffix\":\"\"},{\"id\":536461365,\"identity\":\"b8e4d21c-46aa-4422-af5f-aa95345465c0\",\"order_by\":2,\"name\":\"Alexandra Jamieson\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"UCL\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Alexandra\",\"middleName\":\"\",\"lastName\":\"Jamieson\",\"suffix\":\"\"},{\"id\":536461366,\"identity\":\"7476277a-fe00-4345-806c-375d4f789476\",\"order_by\":3,\"name\":\"Alba Fernandez-Sanles\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-3587-8177\",\"institution\":\"UCL\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Alba\",\"middleName\":\"\",\"lastName\":\"Fernandez-Sanles\",\"suffix\":\"\"},{\"id\":536461367,\"identity\":\"39c221a0-36f1-49cc-84e1-0f3ef2f0d90f\",\"order_by\":4,\"name\":\"Yatharth Ranjan\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"KCL\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Yatharth\",\"middleName\":\"\",\"lastName\":\"Ranjan\",\"suffix\":\"\"},{\"id\":536461368,\"identity\":\"8958aa08-2ef2-43eb-a991-cb1f4420806d\",\"order_by\":5,\"name\":\"Dylan Williams\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-3825-2487\",\"institution\":\"University College London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Dylan\",\"middleName\":\"\",\"lastName\":\"Williams\",\"suffix\":\"\"},{\"id\":536461369,\"identity\":\"e8ccf1f0-fb72-4536-9ba5-4620fd1afe47\",\"order_by\":6,\"name\":\"Lee Howes\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"UCL\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Lee\",\"middleName\":\"\",\"lastName\":\"Howes\",\"suffix\":\"\"},{\"id\":536461370,\"identity\":\"d6ba994a-f1ec-45f2-9f20-6f218a149d08\",\"order_by\":7,\"name\":\"Andrew Wong\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-2079-4779\",\"institution\":\"University College London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Andrew\",\"middleName\":\"\",\"lastName\":\"Wong\",\"suffix\":\"\"},{\"id\":536461371,\"identity\":\"b26f785b-d7bd-4440-ba32-efabf5c1531a\",\"order_by\":8,\"name\":\"Valentina Giunchiglia\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Department of Brain Sciences, Faculty of Medicine, Imperial College London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Valentina\",\"middleName\":\"\",\"lastName\":\"Giunchiglia\",\"suffix\":\"\"},{\"id\":536461372,\"identity\":\"a130349a-a09b-47dd-81e3-0aad2741f148\",\"order_by\":9,\"name\":\"William Trender\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-3947-5532\",\"institution\":\"Imperial College London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"William\",\"middleName\":\"\",\"lastName\":\"Trender\",\"suffix\":\"\"},{\"id\":536461373,\"identity\":\"1d9c6631-68ea-4de6-9049-f7bb02bae73a\",\"order_by\":10,\"name\":\"Peter Hellyer\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"King's College London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Peter\",\"middleName\":\"\",\"lastName\":\"Hellyer\",\"suffix\":\"\"},{\"id\":536461374,\"identity\":\"c989a328-4384-4d41-a87b-a7c58376b51e\",\"order_by\":11,\"name\":\"Nathan J Cheetham\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"KCL\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Nathan\",\"middleName\":\"J\",\"lastName\":\"Cheetham\",\"suffix\":\"\"},{\"id\":536461375,\"identity\":\"41119345-ac8d-42f1-912c-f0eda2201329\",\"order_by\":12,\"name\":\"Ellen Thompson\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of Sussex\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Ellen\",\"middleName\":\"\",\"lastName\":\"Thompson\",\"suffix\":\"\"},{\"id\":536461376,\"identity\":\"8b4ad23a-440f-4a7e-a2ee-80773950ee81\",\"order_by\":13,\"name\":\"Lucy Goudswaard\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of Bristol\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Lucy\",\"middleName\":\"\",\"lastName\":\"Goudswaard\",\"suffix\":\"\"},{\"id\":536461377,\"identity\":\"941ba812-5dca-491f-b738-5b7a90eac7fc\",\"order_by\":14,\"name\":\"Praveetha Patelay\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"UCL\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Praveetha\",\"middleName\":\"\",\"lastName\":\"Patelay\",\"suffix\":\"\"},{\"id\":536461378,\"identity\":\"2b155b9f-8e7b-4030-8428-f909382aca81\",\"order_by\":15,\"name\":\"Amos Folarin\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-0333-1927\",\"institution\":\"Department of Biostatistics \\u0026 Health Informatics King's College London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Amos\",\"middleName\":\"\",\"lastName\":\"Folarin\",\"suffix\":\"\"},{\"id\":536461379,\"identity\":\"6dec8792-dcd0-490d-99e8-e4f90dd170ff\",\"order_by\":16,\"name\":\"Richard Dobson\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-4224-9245\",\"institution\":\"Kings College London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Richard\",\"middleName\":\"\",\"lastName\":\"Dobson\",\"suffix\":\"\"},{\"id\":536461380,\"identity\":\"a8c21c5c-c046-47af-8b07-03a70527d784\",\"order_by\":17,\"name\":\"Nicholas (J) Timpson\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-7141-9189\",\"institution\":\"University of Bristol\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Nicholas\",\"middleName\":\"(J)\",\"lastName\":\"Timpson\",\"suffix\":\"\"},{\"id\":536461381,\"identity\":\"fc531e2e-86f6-4014-ad45-e0bbbe807464\",\"order_by\":18,\"name\":\"Claire Steves\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-4910-0489\",\"institution\":\"King's College London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Claire\",\"middleName\":\"\",\"lastName\":\"Steves\",\"suffix\":\"\"},{\"id\":536461382,\"identity\":\"f2992d1b-21ae-463d-ab12-4cfa796b5ab0\",\"order_by\":19,\"name\":\"Adam Hampshire\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"King's College London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Adam\",\"middleName\":\"\",\"lastName\":\"Hampshire\",\"suffix\":\"\"},{\"id\":536461383,\"identity\":\"202a2cd9-b0e3-4048-bcde-a6b1c9d349b8\",\"order_by\":20,\"name\":\"Nish Chaturvedi\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-6211-2775\",\"institution\":\"University College London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Nish\",\"middleName\":\"\",\"lastName\":\"Chaturvedi\",\"suffix\":\"\"},{\"id\":536461384,\"identity\":\"70e30d64-2fd3-4562-9300-967fc13b838c\",\"order_by\":21,\"name\":\"Alun Hughes\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0001-5432-5271\",\"institution\":\"UCL\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Alun\",\"middleName\":\"\",\"lastName\":\"Hughes\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2025-10-16 10:21:10\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-7876232/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-7876232/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":95354000,\"identity\":\"ffd16713-e45e-429b-bd00-d39e8a99d37b\",\"added_by\":\"auto\",\"created_at\":\"2025-11-07 06:02:08\",\"extension\":\"jpeg\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":334818,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eFlowchart of the study.\\u003c/p\\u003e\\n\\u003cp\\u003eLC+S+: Cases, presenting evidence of long COVID and persistent fatigue cluster symptoms at the initial assessment. LC-S- : Control group 1, presenting no evidence of long COVID (no SARS-CoV-2 infection or infection without prolonged symptoms); LC+S- : Control group 2 (recovered), participants presenting evidence of long COVID at recruitment but no fatigue cluster symptoms at initial assessment; LC-S+ : Control group 3, participants presenting fatigue cluster symptoms not attributable to SARS-CoV-2 infection. Single outcome measures are described next to each physiological system.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"image1.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7876232/v1/0799e465e87fcb248d9e1a58.jpeg\"},{\"id\":95353997,\"identity\":\"697ab52b-6803-4cfc-8493-d5c5b7a972d3\",\"added_by\":\"auto\",\"created_at\":\"2025-11-07 06:02:08\",\"extension\":\"jpeg\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":239710,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eDeficits in aggregate outcome measures.\\u003c/p\\u003e\\n\\u003cp\\u003eSpider plots showing differences between controls (LC-S-) and cases (LC+S+), and groups LC-S+ and LC+S-. Circles represent standardized β coefficients for the association between participant status and aggregate outcome measures at initial assessment (upper graphs) and after 12 months (lower graphs). Shaded areas represent 95% confidence intervals. Models were adjusted for age, sex, body mass index, education (cognitive function only) and mean heart rate (autonomic function only).\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"image2.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7876232/v1/41a0210d1fdc06c7974b59be.jpeg\"},{\"id\":95354001,\"identity\":\"cc7d79e1-40a8-43fe-9ff9-0bd50aa953ae\",\"added_by\":\"auto\",\"created_at\":\"2025-11-07 06:02:08\",\"extension\":\"jpeg\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":394888,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eAssociation between single outcome measures and status.\\u003c/p\\u003e\\n\\u003cp\\u003eForest plots reporting standardized β coefficients and 95% confidence intervals describing the association between participant status (LC+S+, LC-S+, LC+S- compared to LC-S-) and single outcome measures, grouped by physiological system. In each graph, the first 2 columns represent associations at baseline and after 12 months from linear multiple regressions, while the third column reports the standardized coefficient describing the trajectory slope (interaction between time and status in mixed-effect models). Models were adjusted for age, sex and body mass index. Cognitive function and HRV models were further adjusted for education and mean HR, respectively. HR: Heart rate; HRV: HR variability; RR: Respiratory Rate; WT: Walk test.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"image3.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7876232/v1/27e89457837c8a24cb245c59.jpeg\"},{\"id\":95525300,\"identity\":\"382c09a0-899d-45c8-ad67-6ccb65f5183d\",\"added_by\":\"auto\",\"created_at\":\"2025-11-10 10:04:45\",\"extension\":\"jpeg\",\"order_by\":4,\"title\":\"Figure 4\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":217407,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eTrajectories of health in cases and controls.\\u003c/p\\u003e\\n\\u003cp\\u003eTrajectories of aggregate system scores over time were estimated from adjusted linear mixed-effect models. The lines and shaded area represent the predicted outcome and the 95% confidence interval. Models were adjusted for age, sex and body mass index at baseline. Cognitive function and HRV models were further adjusted for education and mean heart rate, respectively. To reduce the effect of learning, the trajectories of cognitive function have been corrected by subtracting the LC-S- slope.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"image4.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7876232/v1/09a2f97a47240eb865c21084.jpeg\"},{\"id\":95353999,\"identity\":\"a6dff896-b947-4b02-b8c0-20e12117482a\",\"added_by\":\"auto\",\"created_at\":\"2025-11-07 06:02:08\",\"extension\":\"jpeg\",\"order_by\":5,\"title\":\"Figure 5\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":266246,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eCorrelations between outcome measure fluctuations\\u003c/p\\u003e\\n\\u003cp\\u003eA: Heatmap showing the median, pairwise, within-subject, Spearman’s correlation coefficient between outcome measure fluctuations. Only participants monitored for ≥12 months were included. The upper and lower triangles show results for controls (LC-S-) and cases (LC+S+), respectively. Crosses indicate correlation coefficients whose distribution is different from zero (p\\u0026lt;0.05, signrank test). Correlations which were statistically significant in cases but not in controls are reported in bold. B: Time-series showing inverse correlation between accuracy in cognitive tests (CF Accuracy) and depression in 11 long COVID cases. For better visualization, outcome measures are normalized and inverted for depression. CF: Cognitive function. HR: Heart rate. RR: respiratory rate. HRV: Heart rate variability.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"image5.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7876232/v1/d62c32de2951e42e0ac94e33.jpeg\"},{\"id\":95530757,\"identity\":\"e892cfc1-9cdd-4558-a7e9-320940b08a84\",\"added_by\":\"auto\",\"created_at\":\"2025-11-10 10:21:48\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":2583635,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7876232/v1/ccacc2a2-bbc4-4188-8343-34291c24a706.pdf\"},{\"id\":95354002,\"identity\":\"bec09245-0ebd-46e4-94b8-f85c3c1f66ab\",\"added_by\":\"auto\",\"created_at\":\"2025-11-07 06:02:08\",\"extension\":\"docx\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":2735537,\"visible\":true,\"origin\":\"\",\"legend\":\"Supplementary Figures and Tables\",\"description\":\"\",\"filename\":\"Supplementary251014.docx\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7876232/v1/725998afbb9c21904ed7119a.docx\"}],\"financialInterests\":\"There is \\u003cb\\u003eNO\\u003c/b\\u003e Competing Interest.\",\"formattedTitle\":\"Persistent Multi-System Impairments Detected by Wearable Monitoring in Long COVID Cases reporting Fatigue up to Three Years Post-Infection\",\"fulltext\":[{\"header\":\"Main\",\"content\":\"\\u003cp\\u003eLong COVID is a complex, multisystem, condition characterized by a constellation of symptoms following SARS-CoV-2 infection. It may affect up to 400\\u0026nbsp;million individuals globally and has an estimated economic impact equivalent to about 1% of the global economy \\u003csup\\u003e\\u003cspan citationid=\\\"CR1\\\" class=\\\"CitationRef\\\"\\u003e1\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e\\u003c/sup\\u003e. No diagnostic test or effective treatment is available and a better understanding of health trajectories in long COVID is needed to improve patient care and support. Studies investigating long COVID have mainly focused on characterizing or tracking symptoms through surveys\\u003csup\\u003e\\u003cspan additionalcitationids=\\\"CR4 CR5\\\" citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR6\\\" class=\\\"CitationRef\\\"\\u003e6\\u003c/span\\u003e\\u003c/sup\\u003e and those that have investigated physiological parameters have mainly focused on single system deficits (e.g. mental health or cognitive function)\\u003csup\\u003e\\u003cspan additionalcitationids=\\\"CR8 CR9\\\" citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e\\u003c/sup\\u003e often in severe, hospitalised, cases \\u003csup\\u003e\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e,\\u003cspan additionalcitationids=\\\"CR12 CR13 CR14 CR15 CR16\\\" citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR17\\\" class=\\\"CitationRef\\\"\\u003e17\\u003c/span\\u003e\\u003c/sup\\u003e. However, around 90% of long COVID cases had a SARS-CoV-2 infection for which they were not hospitalised \\u003csup\\u003e\\u003cspan citationid=\\\"CR1\\\" class=\\\"CitationRef\\\"\\u003e1\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR18\\\" class=\\\"CitationRef\\\"\\u003e18\\u003c/span\\u003e\\u003c/sup\\u003e. This limits our understanding of long COVID and poses the risk of constructing an evidence base biased towards more severe or specific initial presentations and less focused on non-hospitalised individuals with symptoms clustered around fatigue, cognitive function and mental health \\u003csup\\u003e\\u003cspan citationid=\\\"CR19\\\" class=\\\"CitationRef\\\"\\u003e19\\u003c/span\\u003e\\u003c/sup\\u003e. The lack of multi-domain prospective longitudinal data means that the long-term trajectory of long COVID following community-managed SARS-CoV-2 infection is unclear. The prognosis for such cases is also uncertain. Data from surveys have shown evidence of deficits in mental health, and in perceived physical and cognitive functions up to two years post-infection \\u003csup\\u003e\\u003cspan citationid=\\\"CR20\\\" class=\\\"CitationRef\\\"\\u003e20\\u003c/span\\u003e\\u003c/sup\\u003e, with some symptoms worsening over time \\u003csup\\u003e\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e\\u003c/sup\\u003e. Persisting cognitive function deficits were also reported in large population-based samples in which long COVID cases and controls were followed for two years post-infection \\u003csup\\u003e\\u003cspan citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e\\u003c/sup\\u003e. Very limited data are available on health trajectories beyond 2 years after infection, and current knowledge is based on observational studies assessing risk of adverse health outcomes rather than continuous monitoring of health \\u003csup\\u003e\\u003cspan citationid=\\\"CR21\\\" class=\\\"CitationRef\\\"\\u003e21\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR22\\\" class=\\\"CitationRef\\\"\\u003e22\\u003c/span\\u003e\\u003c/sup\\u003e. As most of the available data come from questionnaires, it is still unclear which physiological process underpins the symptomatology of long COVID, and this limits our understanding of and possible response to this condition.\\u003c/p\\u003e\\u003cp\\u003eCONVALESCENCE (\\u0026ldquo;COroNaVirus post-Acute Long-term EffectS: Constructing an evidENCE base\\u0026rdquo;) is a case-control study of long COVID that continuously monitored multi-system function and trajectories of health in people suffering from SARS-CoV-2 infection who were not hospitalised. An innovative mobile health framework including a smartwatch, mobile phone applications, and frequent remote tests and surveys \\u003csup\\u003e\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR23\\\" class=\\\"CitationRef\\\"\\u003e23\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e\\u003c/sup\\u003e was developed to simultaneously and continuously monitor mental health, cognitive function, sleep, cardiorespiratory health, physical activity, exercise capacity and autonomic nervous system function (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e) for 12 months, covering on average the period from the second to the third year post-infection. Long COVID cases presenting with fatigue-cluster symptoms were compared to three control groups defined by the presence or absence of previous SARS-CoV-2 infection and symptoms. Overall, this study captures trajectories of health across multiple systems with unprecedented granularity and demonstrates persistent multiple health deficits three years post infection in long COVID cases with a history of fatigue.\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\"},{\"header\":\"Results\",\"content\":\"\\u003cp\\u003eThe CONVALESCENCE study recruited from two established population-based cohorts in the UK, the Avon Longitudinal Study of Parents and Children \\u003csup\\u003e\\u003cspan citationid=\\\"CR25\\\" class=\\\"CitationRef\\\"\\u003e25\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR26\\\" class=\\\"CitationRef\\\"\\u003e26\\u003c/span\\u003e\\u003c/sup\\u003e (ALSPAC; REC: 21/SC/0030) and TwinsUK \\u003csup\\u003e\\u003cspan citationid=\\\"CR27\\\" class=\\\"CitationRef\\\"\\u003e27\\u003c/span\\u003e\\u003c/sup\\u003e (REC: 19/NW/0187). 349 adults were invited to clinical assessments between September 2021 and May 2023, and 306 (80% female, median 57 [IQR 41\\u0026ndash;63] years) participated in the study (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e). Participants were divided into four groups depending on whether they had reported having long COVID (LC\\u0026thinsp;+\\u0026thinsp;or LC-) before presenting to the clinic visit and whether they presented fatigue symptoms (fatigue, or shortness of breath, or muscle ache) when they entered the study at the clinic visit (S\\u0026thinsp;+\\u0026thinsp;or S-). Out of 141 participants who reported LC symptoms during the recruitment phase (Supplementary Fig.\\u0026nbsp;9), 50 presented with persistent fatigue-cluster symptoms (LC\\u0026thinsp;+\\u0026thinsp;S+) at the clinic visit and were considered cases. The primary control group (LC-S-, N\\u0026thinsp;=\\u0026thinsp;152) did not report having had long COVID and remained free of fatigue-cluster symptoms. Other control groups included those who recovered from long COVID before assessments, i.e. those who reported long COVID in the past but were free of fatigue symptoms at the clinic visit (LC\\u0026thinsp;+\\u0026thinsp;S-, N\\u0026thinsp;=\\u0026thinsp;76), and those who presented in clinic with fatigue-cluster symptoms not attributable to long COVID (LC-S+, N\\u0026thinsp;=\\u0026thinsp;28).\\u003c/p\\u003e\\u003cdiv id=\\\"Sec3\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003eInitial assessment\\u003c/h2\\u003e\\u003cp\\u003eThe description of single and aggregate outcome measures captured by the smartwatch, the Mass Science app and the Cognitron web-based cognitive test is reported in Supplementary Table\\u0026nbsp;1\\u0026ndash;2.\\u003c/p\\u003e\\u003cp\\u003eOn average, participants had estimated first infections by SARS-CoV-2 virus 2.0 (1.7, 2.5) years before the clinic visit, between January 2020 and November 2021, when alpha and delta were the dominant variants in the UK. Among participants who reported long COVID, only 1 had been vaccinated before infection, and only 4 remained unvaccinated at the time of the clinic visit. Groups were generally similar in terms of age, sex, vaccination status, and comorbidities, although LC\\u0026thinsp;+\\u0026thinsp;S\\u0026thinsp;+\\u0026thinsp;participants had 12% higher BMI than controls (p\\u0026thinsp;=\\u0026thinsp;0.01) (Table\\u0026nbsp;\\u003cspan refid=\\\"Tab1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e). Cardiovascular disease and diabetes were rare (\\u0026lt;\\u0026thinsp;5 in all groups). The initial assessment period included the first 30 days of monitoring. During this period, median smartwatch wearing time was 93% and 90% in LC-S- controls and LC\\u0026thinsp;+\\u0026thinsp;S\\u0026thinsp;+\\u0026thinsp;cases, respectively (p\\u0026thinsp;=\\u0026thinsp;0.04) (Table\\u0026nbsp;\\u003cspan refid=\\\"Tab1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e). Compared to controls (LC-S-), cases (LC\\u0026thinsp;+\\u0026thinsp;S+) presented differences in at least one of each system\\u0026rsquo;s outcome measure (Table\\u0026nbsp;\\u003cspan refid=\\\"Tab1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e), with large median differences for intensity of depression and anxiety and moderate to small median differences for cognitive function (accuracy \\u0026minus;\\u0026thinsp;77%), sleep duration variability (+\\u0026thinsp;23%), heart rate variability (8%), heart rate reserve (-6%), number of steps per day (-5%), walk speed (-5%) and resting heart rate (+\\u0026thinsp;2 bpm).\\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\\u003e\\u003cb\\u003eDemographic data and outcome measures at initial assessment.\\u003c/b\\u003e Data is provided as median (interquartile range) or number (%). BMI: Body mass index; HR: Heart rate; RR: Respiratory rate; WT: walk test; HRV: Heart rate variability; Wear. T.: Wearing time. Data reported in bold show p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05 with respect to LC-S- (two-tailed Mann Whitney U test) .\\u003c/p\\u003e\\u003c/div\\u003e\\u003c/caption\\u003e\\u003ccolgroup cols=\\\"5\\\"\\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\\u003cthead\\u003e\\u003ctr\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eCharacteristics\\u003c/p\\u003e\\u003c/th\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003eLC- S- (152)\\u003c/p\\u003e\\u003c/th\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003eLC- S+ (28)\\u003c/p\\u003e\\u003c/th\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003eLC\\u0026thinsp;+\\u0026thinsp;S- (76)\\u003c/p\\u003e\\u003c/th\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003eLC\\u0026thinsp;+\\u0026thinsp;S+ (50)\\u003c/p\\u003e\\u003c/th\\u003e\\u003c/tr\\u003e\\u003c/thead\\u003e\\u003ctbody\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eAge (years)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e58 (31, 63)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e59 (53, 66)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e56 (46, 62)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e55 (47, 61)\\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\\u003e35 (23%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e6 (21%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e12 (16%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e8 (16%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eBMI (kg/m2)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e25 (22, 29)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e27 (25, 28)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e25 (22, 28)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e28 (23, 34)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eVaccination\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e144 (95%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e25 (89%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e71 (93%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e48 (96%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eEver smoker\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e56 (37%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e12 (43%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e30 (39%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e16 (32%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eAsthma\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e28 (21%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e5 (20%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e11 (16%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e11 (23%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eHypertension\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e21 (17%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e6 (27%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e13 (20%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e11 (24%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eHigher education\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e93 (61%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e15 (54%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e45 (59%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e27 (54%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eMental Health\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S-\\u003c/b\\u003e (143)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S+\\u003c/b\\u003e (27)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S-\\u003c/b\\u003e (73)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S+\\u003c/b\\u003e (50)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eSatisfaction (%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e40.0 (35.2, 44.5)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e37.5 (32.5, 40.0)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e40.0 (34.3, 43.8)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e37.8 (33.0, 41.7)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eAnxiety (%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e8.93 (1.90, 19.52)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e11.11 (4.76, 23.81)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e8.73 (1.90, 23.69)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e13.7 (7.14, 31.8)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eDepression (%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e8.33 (4.17, 19.31)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e22.2 (10.0, 29.2)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e11.25 (4.17, 23.26)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e22.9 (10.4, 35.0)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eCognitive Function\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S-\\u003c/b\\u003e (125)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S+\\u003c/b\\u003e (25)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S-\\u003c/b\\u003e (66)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S+\\u003c/b\\u003e (47)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eResp. Time (nu)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e-0.41 (-0.86, 0.12)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e0.07 (-0.68, 0.37)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e-0.33 (-0.81, 0.02)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e-0.26 (-0.77, 0.62)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eAccuracy (nu)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e0.35 (-0.06, 0.76)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e-0.06 (-0.43, 0.54)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e0.32 (-0.09, 0.68)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e0.08 (-0.52, 0.53)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eSleep\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S-\\u003c/b\\u003e (136)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S+\\u003c/b\\u003e (25)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S-\\u003c/b\\u003e (70)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S+\\u003c/b\\u003e (48)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eOut of Range (%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e42.6 (28.6, 55.6)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e47.4 (36.7,61.1)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e43.9 (27.6, 59.1)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e50.0 (30.4, 59.2)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eVariability (h)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e1.03 (0.84, 1.34)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e1.12 (0.94, 1.31)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e1.09 (0.85, 1.39)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e1.27 (0.92, 1.59)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eExercise Capacity\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S-\\u003c/b\\u003e (125)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S+\\u003c/b\\u003e (21)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S-\\u003c/b\\u003e (61)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S+\\u003c/b\\u003e (40)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eHR Reserve (bpm)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e88.0 (79.6, 101.6)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e84.7 (72.5, 93.0)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e87.8 (75.0, 102.5)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e82.9 (71.3, 94.0)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eSteps in WT (n)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e756 (722, 788)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e693 (671, 765)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e750 (705, 799)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e717 (669, 741)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eWalk speed (mps)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e1.84 (1.70, 1.99)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e1.88 (1.67, 1.97)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e1.83 (1.66, 1.94)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e1.75 (1.56, 1.85)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003ePhysical Activity\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S-\\u003c/b\\u003e (137)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S+\\u003c/b\\u003e (25)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S-\\u003c/b\\u003e (70)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S+\\u003c/b\\u003e (47)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eSteps/day (K)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e8.26 (6.62, 10.7)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e7.10 (4.83, 9.53)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e9.77 (6.56, 11.6)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e7.32 (5.18, 9.59)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eActive Time (%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e7.40 (5.64, 9.70)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e7.07 (4.68, 9.19)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e7.51 (5.97, 8.95)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e6.14 (4.94, 8.25)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eCardiorespiratory\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S-\\u003c/b\\u003e (145)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S+\\u003c/b\\u003e (26)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S-\\u003c/b\\u003e (73)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S+\\u003c/b\\u003e (48)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eRest HR (bpm)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e50.0 (46.0, 54.0)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e52.8 (49.0, 59.0)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e49.0 (45.8, 52.0)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e52.0 (48.2, 54.5)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eRest RR (brpm)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e8.05 ( 5.90, 9.31)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e8.06 ( 5.54, 8.99)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e7.98 ( 6.40, 9.19)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e8.62 ( 7.68, 9.36)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eAutonomic Function\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S-\\u003c/b\\u003e (145)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S+\\u003c/b\\u003e (26)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S-\\u003c/b\\u003e (73)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S+\\u003c/b\\u003e (48)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eHRV SDANN (bpm)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e13.0 (11.7, 14.3)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e12.7 (11.2, 14.7)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e12.7 (11.6, 13.9)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e12.0 (10.9, 13.5)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eHRV SDSD (bpm)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e9.25 (8.65, 9.77)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e8.78 (8.29, 9.26)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e9.38 (8.77, 10.09)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e8.79 (8.21, 9.87)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eValidity\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S-\\u003c/b\\u003e (145)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC- S+\\u003c/b\\u003e (26)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S-\\u003c/b\\u003e (73)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003eLC\\u0026thinsp;+\\u0026thinsp;S+\\u003c/b\\u003e (48)\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eWear. T Day (%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e92.9 (85.5, 95.5)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e93.2 (88.7, 95.1)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e91.2 (81.2, 95.3)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e90.1 (83.7, 92.7)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eWear. T Night (%)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e89.8 (76.5, 97.1)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e89.9 (74.6, 97.2)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e90.3 (71.2, 96.6)\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003e\\u003cb\\u003e89.1 (77.4, 95.0)\\u003c/b\\u003e\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003c/tbody\\u003e\\u003c/colgroup\\u003e\\u003c/table\\u003e\\u003c/div\\u003e\\u003c/p\\u003e\\u003cp\\u003eCorrelation heatmaps between outcome measures during the initial assessment revealed distinct patterns between cases (LC\\u0026thinsp;+\\u0026thinsp;S+) and controls (LC-S-) (Supplementary Fig.\\u0026nbsp;1). For example, in cases, the response time to the cognitive tests was moderately inversely correlated to heart rate reserve, walking speed and heart rate variability, suggesting a possible association between cognitive function, exercise capacity and autonomic regulation.\\u003c/p\\u003e\\u003cp\\u003eAssociations between outcome measures and participant status were assessed using linear regressions. After statistically adjusting for age, sex, BMI, education (cognitive function only) and mean heart rate (cardiac autonomic nervous system only), cases (LC\\u0026thinsp;+\\u0026thinsp;S+) presented deficits with respect to controls (LC-S-) in all health systems (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003e, Supplementary Table\\u0026nbsp;3). Aggregate mental health outcomes combining anxiety, depression and satisfaction were the most reduced (β: -0.53 (-0.82, -0.24), p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.01), but deficits were also present in aggregate outcomes for cognitive function (β: -0.39 (-0.60, -0.18), p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.01), exercise capacity (β: -0.31 (-0.55, -0.08), p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.01), cardiorespiratory function (β: -0.39 (-0.60, -0.18), p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.01), sleep (β: -0.34 (-0.63, -0.04), p\\u0026thinsp;=\\u0026thinsp;0.02), physical activity (β: -0.34 (-0.63, -0.05), p\\u0026thinsp;=\\u0026thinsp;0.02), and autonomic function (β: -0.24 (-0.49, 0.01), p\\u0026thinsp;=\\u0026thinsp;0.06). Similarly, LC-S\\u0026thinsp;+\\u0026thinsp;participants tended to show deficits across all aggregate outcome measures, although with smaller effect size and wider confidence intervals, whereas LC\\u0026thinsp;+\\u0026thinsp;S- participants were similar to controls (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003e).\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003cp\\u003eThese system deficits were underpinned by differences in most single outcome measures (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003e, Supplementary Table\\u0026nbsp;4). Compared to controls, all components of mental health (anxiety, depression and reduced satisfaction) were more adverse, while all components of cognitive function (aggregate response time and accuracy), and exercise capacity (heart rate reserve, walking speed and number of steps during walk tests) were decreased. Most of the individual cognitive tasks were performed with lower accuracy, or longer response time, or both (Supplementary Table\\u0026nbsp;5). Differences were detected in motor control, immediate and delayed memory, processing speed and visuo-spatial abilities, but not in mental manipulation and verbal reasoning. Deficits in cardiorespiratory function, autonomic nervous system, physical activity and sleep were mainly driven by a higher resting heart rate, smaller long-term heart rate variability (SDANN), reduced active time, and increased sleep duration variability, respectively.\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003c/div\\u003e\\n\\u003ch3\\u003eTrajectories of health in long COVID\\u003c/h3\\u003e\\n\\u003cp\\u003eOn average, participants were continuously monitored using the smartwatch for 12 (9, 12) months, during which they conducted 9 (3, 16) walk tests, 11 (2, 26) surveys for mental health and 12 (5, 22) cognitive tests (Supplementary Table\\u0026nbsp;6). No difference was found in smartwatch wearing time or in the number of tests/surveys between the four groups.\\u003c/p\\u003e\\u003cp\\u003eAdjusted linear regressions using data collected at 12 months showed that, with respect to LC-S- controls, LC\\u0026thinsp;+\\u0026thinsp;S\\u0026thinsp;+\\u0026thinsp;cases still presented deficits across most of the health systems (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003e, supplementary Table\\u0026nbsp;3). Aggregate outcome measures remained significantly reduced for mental health (β: -0.49 (-0.92, -0.05), p\\u0026thinsp;=\\u0026thinsp;0.03), exercise capacity (β: -0.49 (-0.82, -0.15), p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.01), and cognitive function (β: -0.43 (-0.83, -0.03), p\\u0026thinsp;=\\u0026thinsp;0.04). Aggregate outcome measures for sleep (β: -0.36 (-0.77, 0.04), p\\u0026thinsp;=\\u0026thinsp;0.08), autonomic function (β: -0.25 (-0.58, 0.08), p\\u0026thinsp;=\\u0026thinsp;0.13) and cardiorespiratory system (β: -0.17 (-0.49, 0.14), p\\u0026thinsp;=\\u0026thinsp;0.28) were still reduced to similar extents, although differences were no longer statistically significant probably on account of a reduced sample size. No differences were observed in physical activity (β: -0.13 (-0.52, 0.27), p\\u0026thinsp;=\\u0026thinsp;0.53). These persisting system deficits were mainly driven by increased depressive symptoms, reduced heart rate reserve, increased sleep variability, increased resting heart rate and reduced heart rate variability (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003e, Supplementary Table\\u0026nbsp;4).\\u003c/p\\u003e\\u003cp\\u003eTrajectories of outcome measures were predicted using adjusted linear mixed-effect regressions. Aggregate outcome measures for cases did not converge over time towards control values (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig4\\\" class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003e). The regression coefficients representing the interaction between participant status and time (i.e. the trajectory slope) provided no convincing evidence of recovery for any of the outcome measures monitored during the study. On the contrary, a trend towards a more marked deterioration in sleep, and possibly in exercise capacity and autonomic function was observed in cases as compared to controls (Figs.\\u0026nbsp;\\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003e \\u0026amp; \\u003cspan refid=\\\"Fig4\\\" class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003e).\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\n\\u003ch3\\u003eTemporal variability in outcome measures\\u003c/h3\\u003e\\n\\u003cp\\u003eLong COVID patients often report temporal fluctuations in their symptoms. Temporal variability in psychological, cognitive and physiological outcome measures was assessed by measuring the standard deviation of the linear mixed-effect model residuals. Linear regressions adjusted for age, sex, BMI and education showed that temporal variability in cognitive function was larger in LC\\u0026thinsp;+\\u0026thinsp;S\\u0026thinsp;+\\u0026thinsp;cases compared with LC-S- controls (Supplementary Table\\u0026nbsp;7). These differences were however attenuated after adjusting for the mean level of cognitive function over the monitoring period, which was reduced in cases. Individuals who recovered from long COVID (LC\\u0026thinsp;+\\u0026thinsp;S-) did not show differences in temporal variability of outcome measures, while participants with fatigue-cluster symptoms not attributable to long COVID (LC-S+) showed larger temporal variability in psychological outcome measures. This difference was also attenuated after adjusting for the mean value of psychological outcome measures over the monitoring period.\\u003c/p\\u003e\\u003cp\\u003ePairwise, within-subject, correlation coefficients between outcome measures were estimated to assess the strength of the coupling between the temporal fluctuations in outcome measures (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig5\\\" class=\\\"InternalRef\\\"\\u003e5\\u003c/span\\u003e). Only participants who completed the 12-month monitoring period were included in the analysis to capture consistent patterns. Outcome measures representing physical activity, exercise capacity and autonomic function were positively coupled in both cases and controls. For example, among controls, steps per day correlated with both heart rate variability and heart rate reserve (median [interquartile range] correlation coefficients equal to 0.33 [0.10, 0.56], p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.01, and 0.17 [-0.05, 0.44], p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.01, respectively). Some correlations were statistically significant among LC\\u0026thinsp;+\\u0026thinsp;S\\u0026thinsp;+\\u0026thinsp;cases but not among LC-S- controls (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig5\\\" class=\\\"InternalRef\\\"\\u003e5\\u003c/span\\u003eA). The accuracy of cognitive tests was inversely correlated to depressive symptoms (cc = -0.54 [-0.66, -0.16], p\\u0026thinsp;=\\u0026thinsp;0.02, Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig5\\\" class=\\\"InternalRef\\\"\\u003e5\\u003c/span\\u003eB) and resting heart rate (cc=-0.18 [-0.41,-0.10], p\\u0026thinsp;=\\u0026thinsp;0.02); depression was also negatively correlated with steps per day (cc=-0.28 [-0.45, 0.02, p\\u0026thinsp;=\\u0026thinsp;0.04]), and fluctuations in sleep duration was inversely correlated with heart rate variability (cc=-0.26 [-0.41, 0.04], p\\u0026thinsp;=\\u0026thinsp;0.01). The median correlation coefficient between accuracy of cognitive tests and anxiety (cc=-0.32 [-0.48, 0.11], p\\u0026thinsp;=\\u0026thinsp;0.29) and satisfaction (cc\\u0026thinsp;=\\u0026thinsp;0.33 [-0.10, 0.50], p\\u0026thinsp;=\\u0026thinsp;0.30) was relatively high but did not reach statistical significance, probably on account of small sample size (N\\u0026thinsp;=\\u0026thinsp;11).\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\n\\u003ch3\\u003eSensitivity analyses\\u003c/h3\\u003e\\n\\u003cp\\u003eSeveral sensitivity analyses were conducted.\\u003c/p\\u003e\\u003cp\\u003eDifferences between aggregate outcome measures were assessed after matching cases and controls for BMI (by identifying the control subgroup which minimizes BMI differences). Results were very similar to those obtained in the primary analysis (Supplementary Fig.\\u0026nbsp;2), demonstrating that a slightly larger BMI in cases does not explain health deficits.\\u003c/p\\u003e\\u003cp\\u003eWe assessed the impact of attrition during the monitoring period. Restricting the analysis to study participants who were continuously monitored for 12 months (N\\u0026thinsp;=\\u0026thinsp;202) did not substantially modify the results. In this subgroup, aggregate and single outcome measures at the initial assessment and at 12 months, as well as their predicted trends over time showed patterns similar to those observed in the primary analysis, albeit with larger confidence intervals attributable to smaller sample size (Supplementary Fig.\\u0026nbsp;3\\u0026ndash;5, supplementary table 8\\u0026ndash;9).\\u003c/p\\u003e\\u003cp\\u003eWe then compared trajectories of outcome measures when the starting point was defined as the date of the reported infection, rather than the date of the clinic visit. No evidence of significant recovery in aggregate outcome measures was found in cases (LC\\u0026thinsp;+\\u0026thinsp;S+) as compared to controls (LC-S-) (Supplementary Fig.\\u0026nbsp;6). Qualitatively, in cases, cognitive function slightly improved over the years, while exercise capacity, autonomic function and sleep tended to get worse.\\u003c/p\\u003e\\u003cp\\u003eFinally, we assessed how the definition of cases and controls could have impacted the results. Broadening the spectrum of symptoms used to define cases by including participants that reported confusion or joint pain in addition to fatigue, shortness of breath and muscle pain, did not alter the results and only slightly increased difference in mental health outcomes (Supplementary Fig.\\u0026nbsp;7\\u0026ndash;8, supplementary table 10\\u0026ndash;11). This is probably because 86% of study participants reporting confusion or joint pain also reported fatigue cluster symptoms.\\u003c/p\\u003e\"},{\"header\":\"Discussion\",\"content\":\"\\u003cp\\u003eCompared to controls with no history of long COVID or fatigue, long COVID cases that reported persistent fatigue cluster symptoms two years after infection presented ongoing deficits across all investigated systems (mental health, cognitive function, exercise capacity, physical activity, autonomic nervous system, cardiorespiratory function and sleep). No evidence of recovery was identified in the trajectory of outcome measures over the following year with most of the deficits persisting after 12 months. Participants who had recovered from long COVID or did not report fatigue cluster symptoms at recruitment (LC\\u0026thinsp;+\\u0026thinsp;S-) did not show any discernible health deficits. In contrast, health deficits in cases (LC\\u0026thinsp;+\\u0026thinsp;S+) were similar to those in participants presenting with fatigue cluster symptoms not attributed to long COVID (LC-S+).\\u003c/p\\u003e\\u003cp\\u003eOur data suggest that for long COVID cases that remained symptomatic 2 years post infection, fatigue cluster symptoms coexist not only with worse mental health and cognitive outcomes, as recently reported in hospitalised cases \\u003csup\\u003e\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e\\u003c/sup\\u003e, but also with impaired physiological outcome measures, namely exercise capacity, physical activity, cardiac autonomic and cardiorespiratory function and sleep quality. Importantly, in cases reporting fatigue, the lack of recovery in most outcome measures over 12 months strongly suggests that time may not be a \\u0026lsquo;cure\\u0026rsquo;, and further investigation on possible treatment strategies is urgently needed. Several studies have found persisting cognitive and mental health symptoms at least 1 year after infection in post-hospitalised\\u003csup\\u003e\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR28\\\" class=\\\"CitationRef\\\"\\u003e28\\u003c/span\\u003e\\u003c/sup\\u003e and community-based samples \\u003csup\\u003e\\u003cspan citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR29\\\" class=\\\"CitationRef\\\"\\u003e29\\u003c/span\\u003e\\u003c/sup\\u003e, and our data demonstrate that these deficits may persist for over 3 years. Contrary to a previous report that assessed subjective cognitive function 2 years post infection \\u003csup\\u003e\\u003cspan citationid=\\\"CR20\\\" class=\\\"CitationRef\\\"\\u003e20\\u003c/span\\u003e\\u003c/sup\\u003e, we did not observe cognitive deficits in those who had recovered from long COVID (LC\\u0026thinsp;+\\u0026thinsp;S-). This discrepancy, however, may be partially explained by differences between subjective and objective cognitive assessment. Objectively measured deficits in global cognition in individuals who recovered from long COVID were also identified in a large cross-sectional community sample study, in some cases up to two years post-infection \\u003csup\\u003e\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e\\u003c/sup\\u003e. The magnitude of these deficits was, however, small and their detection may require much larger samples than used in our study.\\u003c/p\\u003e\\u003cp\\u003eReduced exercise capacity\\u003csup\\u003e\\u003cspan citationid=\\\"CR13\\\" class=\\\"CitationRef\\\"\\u003e13\\u003c/span\\u003e,\\u003cspan additionalcitationids=\\\"CR30 CR31\\\" citationid=\\\"CR29\\\" class=\\\"CitationRef\\\"\\u003e29\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR32\\\" class=\\\"CitationRef\\\"\\u003e32\\u003c/span\\u003e\\u003c/sup\\u003e and diminished levels of objectively measured physical activity\\u003csup\\u003e\\u003cspan additionalcitationids=\\\"CR34\\\" citationid=\\\"CR33\\\" class=\\\"CitationRef\\\"\\u003e33\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR35\\\" class=\\\"CitationRef\\\"\\u003e35\\u003c/span\\u003e\\u003c/sup\\u003e in long COVID have been previously reported and are considered hallmarks of long COVID. Our data demonstrate that in those reporting fatigue cluster symptoms two years post infection, reduced exercise capacity, as assessed by maximum walk speed, number of steps in a 6-minute walk test and heart rate reserve, is likely to persist for at least another year. The magnitude of deficits in exercise capacity was however relatively small. Interestingly, after 12 months, cases and controls showed similar aggregate outcome measures of physical activity, but the difference in aggregate outcome measures of exercise capacity remained evident, suggesting that reduced exercise capacity may not be fully explained by reduced physical activity.\\u003c/p\\u003e\\u003cp\\u003eCardiorespiratory function also differed between cases and controls, with both resting heart rate and respiratory rate being modestly higher in cases than controls. This is consistent with the results of a large longitudinal wearable study that demonstrated an increase in resting heart rate following SARS-CoV-2 infection \\u003csup\\u003e\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e\\u003c/sup\\u003e. Heart rate variability, a non-invasive and qualitative measure of cardiac autonomic function, was also reduced in cases as compared to controls, and this difference was maintained during the monitoring period. Differences remained statistically significant after adjusting for mean heart rate, therefore reducing the possibility that these differences were driven by increased resting heart rate. Dysautonomia is a frequent finding in long COVID cases \\u003csup\\u003e\\u003cspan citationid=\\\"CR32\\\" class=\\\"CitationRef\\\"\\u003e32\\u003c/span\\u003e,\\u003cspan additionalcitationids=\\\"CR37\\\" citationid=\\\"CR36\\\" class=\\\"CitationRef\\\"\\u003e36\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR38\\\" class=\\\"CitationRef\\\"\\u003e38\\u003c/span\\u003e\\u003c/sup\\u003e, and our data demonstrate a protracted effect in those individuals with long-term fatigue cluster symptoms.\\u003c/p\\u003e\\u003cp\\u003eCompared to controls, cases also showed greater objectively measured sleep irregularity, which persisted through the entire monitoring period. This is in agreement with data from surveys of individuals with long COVID, which indicate a prevalence of poor sleep quality above 50% \\u003csup\\u003e39\\u003c/sup\\u003e, with irregular sleep being mentioned as the most commonly reported disturbance\\u003csup\\u003e\\u003cspan citationid=\\\"CR40\\\" class=\\\"CitationRef\\\"\\u003e40\\u003c/span\\u003e\\u003c/sup\\u003e and poor sleep quality correlating with symptom severity and reduced quality of life \\u003csup\\u003e\\u003cspan citationid=\\\"CR41\\\" class=\\\"CitationRef\\\"\\u003e41\\u003c/span\\u003e\\u003c/sup\\u003e. Sleep irregularity is an important and perhaps overlooked consequence of long COVID, which has been associated with reduced cardiometabolic health and cognitive function \\u003csup\\u003e\\u003cspan citationid=\\\"CR42\\\" class=\\\"CitationRef\\\"\\u003e42\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR43\\\" class=\\\"CitationRef\\\"\\u003e43\\u003c/span\\u003e\\u003c/sup\\u003e. While in the long COVID cases monitored in this study sleep variability was not associated with reduced cognitive function, sleep duration trends were inversely correlated with heart rate variability, suggesting that periods of increased sleep irregularity coincided with periods of reduced cardiac autonomic function.\\u003c/p\\u003e\\u003cp\\u003eThe multimodal longitudinal design of the study allowed us to analyse outcome measure fluctuations over the 12-month monitoring period and their correlations. Cognitive function was the only outcome measure showing larger fluctuations in long COVID cases as compared to controls. This difference was attenuated after adjusting by the mean cognitive outcome measure suggesting that cognitive function in long COVID cases is not consistently reduced but rather fluctuates around lower levels.\\u003c/p\\u003e\\u003cp\\u003eThe strength of the coupling between fluctuations in outcome measures was estimated using correlation analysis. A small-to-moderate positive correlation between measures of physical activity, exercise capacity, and cardiac autonomic function was identified in both cases and controls. In long COVID cases, cognitive function was inversely coupled with depressive symptoms and resting heart rate, and sleep was inversely coupled with autonomic function. The strength of these correlations was low-to-moderate and based on a small number of cases (only participants with complete outcome measures over the 12-month monitoring period were included). While these correlations do not imply causal links, they may reflect a dynamic impact of long COVID on multiple physiological, cognitive and psychological measures which warrants further investigation.\\u003c/p\\u003e\\u003cp\\u003eAlthough this study focused on physiological metrics captured via remote monitoring, emerging molecular evidence provides complementary insight into the biological mechanisms underlying long COVID. Recent work has identified distinct inflammatory signatures (including neuroinflammation) associated with persistent symptoms such as breathlessness and fatigue \\u003csup\\u003e\\u003cspan citationid=\\\"CR44\\\" class=\\\"CitationRef\\\"\\u003e44\\u003c/span\\u003e\\u003c/sup\\u003e. Additionally, altered microRNA profiles have been implicated in immune dysregulation and cellular stress responses \\u003csup\\u003e\\u003cspan citationid=\\\"CR45\\\" class=\\\"CitationRef\\\"\\u003e45\\u003c/span\\u003e\\u003c/sup\\u003e. While molecular data were not examined in this study, our findings may reflect downstream effects of sustained inflammation or immune dysregulation. Future studies integrating wearable-derived phenotypes with molecular profiling could help delineate pathophysiological subtypes and inform targeted interventions.\\u003c/p\\u003e\\u003cp\\u003eTo the extent of our knowledge, this is the first study to simultaneously and continuously monitor multiple physiological, cognitive and psychological health outcomes in long COVID for up to one year. The study utilised a smartwatch which has been extensively validated for monitoring of physiological parameters including heart rate, steps, distance, physical activity, and sleep \\u003csup\\u003e\\u003cspan additionalcitationids=\\\"CR47\\\" citationid=\\\"CR46\\\" class=\\\"CitationRef\\\"\\u003e46\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR48\\\" class=\\\"CitationRef\\\"\\u003e48\\u003c/span\\u003e\\u003c/sup\\u003e, a battery of state-of-the-art cognitive tests commonly employed in long COVID studies \\u003csup\\u003e\\u003cspan additionalcitationids=\\\"CR8 CR9 CR10\\\" citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e\\u003c/sup\\u003e and an app specifically designed to communicate with study participants and to collect survey data \\u003csup\\u003e\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR49\\\" class=\\\"CitationRef\\\"\\u003e49\\u003c/span\\u003e\\u003c/sup\\u003e. Other strengths include the case-control design and recruitment of community-based participants. The main limitations include a moderate sample size and inclusion of individuals infected mainly by alpha and delta variants circulating before the roll out of vaccination programmes. However, because vaccination reduced both the number of infections and the prevalence of long COVID, our sample may be representative of the majority of long COVID cases with persistent symptoms, who were infected before mass vaccination. Another important limitation is the lack of data characterising health outcome measures before the initial assessment. Although participants were carefully recruited and separated into case and control groups based on recently reported symptoms attributable or not attributable to long COVID (and with serological testing for SARS-Cov-2 infections), we cannot exclude that at least part of the health deficits identified in this study may have already existed before the initial assessment.\\u003c/p\\u003e\"},{\"header\":\"Methods\",\"content\":\"\\u003cdiv id=\\\"Sec8\\\" class=\\\"Section2\\\"\\u003e\\u003cdiv id=\\\"Sec9\\\" class=\\\"Section3\\\"\\u003e\\u003ch2\\u003eStudy participants\\u003c/h2\\u003e\\u003cp\\u003eThe CONVALESCENCE study (REC: 21/SC/0235) is described in detail elsewhere \\u003csup\\u003e\\u003cspan citationid=\\\"CR50\\\" class=\\\"CitationRef\\\"\\u003e50\\u003c/span\\u003e\\u003c/sup\\u003e. In brief, the study was set up in 2021 to establish an evidence base for long COVID and it recruited from two established population-based cohorts, the Avon Longitudinal Study of Parents and Children (ALSPAC; REC: 21/SC/0030; Supplementary text) \\u003csup\\u003e\\u003cspan citationid=\\\"CR25\\\" class=\\\"CitationRef\\\"\\u003e25\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR26\\\" class=\\\"CitationRef\\\"\\u003e26\\u003c/span\\u003e\\u003c/sup\\u003e and Twins UK study (REC: 19/NW/0187) \\u003csup\\u003e\\u003cspan citationid=\\\"CR51\\\" class=\\\"CitationRef\\\"\\u003e51\\u003c/span\\u003e\\u003c/sup\\u003e. 349 study participants from these two UK cohorts completed a survey to establish their health status and were invited to a clinic visit at the Bloomsbury Centre for Clinical Phenotyping (BCCP) in London.\\u003c/p\\u003e\\u003cp\\u003eVisits were held between October 2021 and May 2023. During the visit, participants were offered a smartwatch (Garmin Vivoactive 4) and the possibility of installing two apps on their phones: Garmin Connect, which automatically collects data from the watches, and the Mass Science app \\u003csup\\u003e\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR49\\\" class=\\\"CitationRef\\\"\\u003e49\\u003c/span\\u003e\\u003c/sup\\u003e, which was used to collect survey data and send notifications to study participants as well as to enable the use of the Cognitron test, a platform for quantitative remote assessment of cognitive function \\u003csup\\u003e\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e\\u003c/sup\\u003e. Participants were asked to wear the watch as much as possible, to complete mental health and quality of life surveys and 6-minute walking tests (6MWT) frequently (every fortnight for the first 6 weeks and every 4 weeks afterwards). Participants were included in the study if they gave informed consent and if they provided at least one valid data point, defined as completing at least a survey, or a cognitive test, or wearing the watch for at least 6 hours during the daytime for a minimum of 7 days. Three-hundred and six (N\\u0026thinsp;=\\u0026thinsp;306) participants complied with this requirement and were included in the study (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e).\\u003c/p\\u003e\\u003c/div\\u003e\\u003c/div\\u003e\\n\\u003ch3\\u003eLong COVID case-control definitions\\u003c/h3\\u003e\\n\\u003cp\\u003eBefore the clinic visit, participants were classified into those who experienced and did not experience long COVID (LC\\u0026thinsp;+\\u0026thinsp;and LC+). Long COVID was defined as evidence of SARS-CoV-2 infection (self-reported positive PCR test or self-reported suspected SARS-CoV-2 infection with later evidence of natural infection from consideration of antibody testing and vaccination status) and self-reporting of persistent long COVID symptoms including any of fatigue, respiratory, cognitive, cardiovascular, and gastrointestinal cluster symptoms lasting more than 4 weeks and attributable to the infection (Supplementary Fig.\\u0026nbsp;9). Those who did not experience long COVID (LC-), either did not report long COVID symptoms or, in case of reporting symptoms, they had no evidence of SARS-CoV-2 infection. Among participants reporting SARS-CoV-2 infection, 90% were infected between January 2020 and November 2021, and the last infection was recorded in April 2022. Participants were seen in clinic between October 2021 and May 2023 where they were divided into four groups based on whether they had long COVID (LC\\u0026thinsp;+\\u0026thinsp;and LC-) and whether they presented fatigue cluster symptoms (fatigue, or breathlessness or muscle ache) (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e). Cases were defined as those LC\\u0026thinsp;+\\u0026thinsp;participants still presenting fatigue cluster symptoms (N\\u0026thinsp;=\\u0026thinsp;50, LC\\u0026thinsp;+\\u0026thinsp;S+). Controls were defined as participants who reported neither long COVID nor fatigue cluster symptoms (N\\u0026thinsp;=\\u0026thinsp;152, LC-S-). This included those with no infection or infection and symptoms lasting\\u0026thinsp;\\u0026lt;\\u0026thinsp;4 weeks. Additionally, two other control groups were considered: Participants who had reported long COVID but did not present fatigue cluster symptoms in clinic (N\\u0026thinsp;=\\u0026thinsp;76, LC\\u0026thinsp;+\\u0026thinsp;S-), including those with LC characterised by symptoms unrelated to fatigue and those no longer reporting fatigue cluster symptoms, and Participants who reported not having had long COVID, with a negative antibody test, but who presented fatigue cluster symptoms assumed not to be attributable to SARS-CoV-2 infection (N\\u0026thinsp;=\\u0026thinsp;28, LC-S+).\\u003c/p\\u003e\\u003cdiv id=\\\"Sec11\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003eDemographics and anthropometric measurements\\u003c/h2\\u003e\\u003cp\\u003eParticipant age, sex, long COVID symptomatology, medical history and current medications were collected by questionnaire in a secure electronic Case Report Form (eCRF) using REDCap (Research Electronic Data Capture). Height was measured using a stadiometer (Seca217, Seca, Germany) to the closest centimetre and weight was measured in kilograms using digital bio-impedance scales (BC-418 or MC-780MA, Tanita, USA) to calculate body mass index (BMI).\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec12\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003eSmartwatch measurements\\u003c/h2\\u003e\\u003cp\\u003eData was transferred to a Trusted Research Environment through an IT infrastructure that shares the backend of RADAR BASE \\u003csup\\u003e\\u003cspan citationid=\\\"CR23\\\" class=\\\"CitationRef\\\"\\u003e23\\u003c/span\\u003e\\u003c/sup\\u003e. Supplementary Table\\u0026nbsp;1 provides detailed definitions of outcomes based on smartwatch measurements. The watch recorded heart rate every 15 seconds, respiratory rate, number of steps and intensity of physical activity every 15 minutes, and sleep duration once a day. Smartwatch measurements were used to derive metrics representing four distinct systems: Physical activity, exercise capacity, cardiorespiratory health, cardiac autonomic function, and sleep. Physical activity was quantified by the number of steps per day and the proportion of time spent in active or very active activity as defined by Garmin\\u0026rsquo;s proprietary software. Exercise capacity was assessed using the number of steps covered during the remote 6-minute walk tests, the maximum speed during any walking activity lasting between 4 and 60 minutes, and by heart rate reserve \\u003csup\\u003e\\u003cspan citationid=\\\"CR52\\\" class=\\\"CitationRef\\\"\\u003e52\\u003c/span\\u003e\\u003c/sup\\u003e, defined as the difference between maximum and minimum heart rate between 7 AM and 11 PM (night-time values were not considered because some participants did not use the device overnight). The number of steps instead of distance covered during the remote 6-minute walk test was used because number of steps were found to be measured more accurately than distance using the Garmin watch \\u003csup\\u003e\\u003cspan citationid=\\\"CR46\\\" class=\\\"CitationRef\\\"\\u003e46\\u003c/span\\u003e\\u003c/sup\\u003e. Walk activities for which the maximum speed was \\u0026gt;\\u0026thinsp;10 Km/h, or number of steps was \\u0026lt;\\u0026thinsp;100, or total elevation loss/gain was \\u0026gt;\\u0026thinsp;50 m were not considered. Cardiorespiratory function was captured by resting heart rate and respiratory rate (minimum daily values). Cardiac autonomic function was captured by two measures of heart rate variability: The standard deviation of 5-minute average heart rate (SDANN) and the standard deviation of successive 15-second differences measured daily.\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec13\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003eMental health and wellbeing\\u003c/h2\\u003e\\u003cp\\u003eStudy participants were asked to complete standard questionnaires to assess anxiety (the Generalised Anxiety Disorder, GAD-7), depression (Patient Health Questionnaire, PHQ-8) and life satisfaction (Office National Statistics, ONS-2) through the Mass Science app \\u003csup\\u003e\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR49\\\" class=\\\"CitationRef\\\"\\u003e49\\u003c/span\\u003e\\u003c/sup\\u003e. Questionnaires were presented to participants every 4 weeks, and results were rescaled from 0 to 100.\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec14\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003eCognitive assessment\\u003c/h2\\u003e\\u003cp\\u003eCognitive function was quantified through Cognitron \\u003csup\\u003e\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e,\\u003cspan additionalcitationids=\\\"CR10\\\" citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e\\u003c/sup\\u003e, a battery of web-based tasks which were presented through the Mass Science app. More information is available on the website \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://www.cognitron.co.uk/\\u003c/span\\u003e\\u003cspan address=\\\"https://www.cognitron.co.uk/\\\" targettype=\\\"URL\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e. Eight tasks were presented to assess memory (immediate and delayed), visuospatial abilities/attention, processing speed, verbal reasoning and spatial planning. Information related to performance and compliance was collected automatically and combined in a normalized summary score summarizing accuracy and median response time, which were used to characterize cognitive function.\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec15\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003eData processing\\u003c/h2\\u003e\\u003cp\\u003eThe time series representing the temporal trajectory of each outcome measure over the 12-month period was computed by averaging measurements collected over 30 (for all smartwatch measures, except for those derived from walk tests) or 60 (for mental health and cognitive function) consecutive days. In the case of mental health, cognitive function and walk tests, a longer time frame was necessary because surveys and tests were collected less frequently than smartwatch data.\\u003c/p\\u003e\\u003cp\\u003eTo assess health deficits, an aggregate outcome measure per physiological domain was computed by averaging normalized single outcome measures (formula shown in Supplementary Table\\u0026nbsp;2). Single outcome measures that were inversely associated with the global outcomes (e.g. anxiety, depression, response time) were inverted before averaging.\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec16\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003eStatistical Analysis\\u003c/h2\\u003e\\u003cp\\u003eData were summarized as median (interquartile range) or number (percentage) and differences between the 4 groups and pairwise comparisons were assessed using Kruskal Wallis and Wilcoxon ranksum tests, respectively. Associations between participant status (LC\\u0026thinsp;+\\u0026thinsp;S+, LC\\u0026thinsp;+\\u0026thinsp;S-, LC-S+, and LC-S-) and outcome measures at the beginning of the study (first data point in the time-series) and at 12 months (last data point in the time series) were investigated using linear regressions, adjusted for age, sex, and BMI, which were considered confounders. Cognitive outcome measures were further adjusted for education (having/not having higher education or professional equivalent), while cardiac autonomic outcomes (heart rate variability) were further adjusted for mean heart rate. Associations between outcome measures and status are reported using controls (LC-S-) as reference category.\\u003c/p\\u003e\\u003cp\\u003e Associations between participant status and the trajectory of outcome measures were investigated using linear mixed-effect models, where status, time and status-time interaction were treated as fixed effects and uncorrelated random intercepts and slopes were allowed per participants and time. The model was also adjusted for age, sex, BMI, education (cognitive outcomes only) and mean heart rate (autonomic outcomes only). Trajectory of aggregate outcome measures were predicted based on this model, with estimates and 95% confidence interval plotted against time. To account for a learning effect after repeating the battery of cognitive tests multiple times, the trajectories of cognitive outcome measures were corrected by subtracting the controls\\u0026rsquo; slope (LC-S-). In primary analysis, time zero was defined as the date of the clinic visit, whereas in sensitivity analysis, time zero was defined as the reported date of COVID infection. Participants who did not report a COVID infection were assigned a random time zero from a normal distribution with mean and standard deviation equal to that of the visit date distribution in the latter analysis.\\u003c/p\\u003e\\u003cp\\u003eAssociations between status and the temporal variability in the outcome measures were measured by regressing the standard deviation of the conditional residuals of the mixed-effect model on status and number of valid data points. Participants with less than 3 data points were excluded from this analysis.\\u003c/p\\u003e\\u003cp\\u003eThe strength of the coupling between outcome measure fluctuations was assessed measuring within subject Spearman\\u0026rsquo;s correlation coefficients between outcome measures. In this analysis, data averaged over 30 or 60 days (as described above) was used. Outcome measures averaged over 60 days were resampled every 30 days using nearest value interpolation so that every outcome measure was described by a time-series of at most 12 data-points. To assess whether the distributions of within-subject correlation coefficients were statistically different from zero, the Wilcoxon signrank test was used.\\u003c/p\\u003e\\u003cp\\u003eGiven the exploratory nature of the study and multiple distinct domains being investigated, no statistical correction for multiple comparisons were applied.\\u003c/p\\u003e\\u003c/div\\u003e\"},{\"header\":\"Declarations\",\"content\":\"\\u003cp\\u003eAcknowledgements\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eWe are extremely grateful to all the people who took part in this study, and to the past and present members of the research team who collected and managed the data including Lidia Nigrelli, Fintan McArdle, Chelsea Beckford, Clare Davie, Suzanne Williams, Uhuru Lambert, Felicia Huang and Imran Shah. We are extremely grateful to the all participants and Department of Twin Research\\u0026nbsp;staff. We are extremely grateful to all the families who took part in this study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes data collection staff, data and administrations staff, technical managers and the technical staff with the Bristol Bioresource Laboratory, based within the University of Bristol.\\u003c/p\\u003e\\n\\u003cp\\u003eConflict of Interest\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eNC receives funds from AstraZeneca for serving on data safety and monitoring committees for clinical trials\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eFunding Statement\\u003c/p\\u003e\\n\\u003cp\\u003eThe Characterisation, determinants, mechanisms and consequences of the long-term effects of COVID-19: providing the evidence base for health care services (CONVALESCENCE, COV-LT-0009\\\\MC_PC20051\\\\MC_PC_20059) study was jointly funded by the National Institute for Health and Care Research (NIHR) and UK Research and Innovation (UKRI). The views expressed in this publication are those of the author(s) and not necessarily those of NIHR, The Department of Health and Social Care or UKRI.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eAJ was supported by a British Heart Foundation 4-year PhD studentship (FS/19/63/34902) awarded to UCL. The UK Medical Research Council and Wellcome (Grant ref: 217065/Z/19/Z) and the University of Bristol provide core support for ALSPAC. This publication is the work of the authors and will serve as guarantors for the contents of this paper. A comprehensive list of grants funding is available on the ALSPAC website (http://www.bristol.ac.uk/alspac/external/documents/grant-acknowledgements.pdf). TwinsUK is funded by the Wellcome Trust, Medical Research Council, Versus Arthritis, European Union Horizon 2020, Chronic Disease Research Foundation (CDRF), Zoe Ltd, the National Institute for Health and Care Research (NIHR) Clinical Research Network (CRN) and Biomedical Research Centre based at Guy\\u0026rsquo;s and St Thomas\\u0026rsquo; NHS Foundation Trust in partnership with King\\u0026rsquo;s College London. BCCP received infrastructure support from the NIHR UCLH Biomedical Research Centre and the BHF.\\u003cstrong\\u003e\\u0026nbsp;\\u003c/strong\\u003eNC and ADH work in a unit that receives support from the UK Medical Research Council (grant number MC_UU_12019/1). NJC was supported by the\\u0026nbsp;CONVALESCENCE study [COV-LT-0009]. EJT acknowledges funding from the Wellcome Trust (WT212904/Z/18/Z) and NIHR (CONVALESCENCE grant COV-LT-0009).\\u003c/p\\u003e\\n\\u003cp\\u003eData Availability Declaration\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eDue to the sensitive nature of the data collected for this study, data cannot be made publicly available, but requests to access the dataset from qualified researchers trained in human subject confidentiality protocols may be sent to mrclha.swiftinfo@ucl.ac.uk. \\u0026nbsp; \\u0026nbsp;\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eThe informed consent obtained from ALSPAC (Avon Longitudinal Study of Parents and Children) participants does not allow the data to be made available through any third party maintained public repository. Supporting data are available from ALSPAC on request under the approved proposal number, B3666. Full instructions for applying for data access can be found here: http://www.bristol.ac.uk/alspac/researchers/access/. The ALSPAC study website contains details of all available data (http://www.bristol.ac.uk/alspac/researchers/our-data/).\\u003c/p\\u003e\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\u003cli\\u003e\\u003cspan\\u003eAl-Aly Z et al (2024) Long COVID science, research and policy. Nat Med 30:2148\\u0026ndash;2164\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eAl-Aly Z (2023) Prevention of long COVID: progress and challenges. Lancet Infect Dis 23:776\\u0026ndash;777\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eShen Q et al (2023) COVID-19 illness severity and 2-year prevalence of physical symptoms: an observational study in Iceland, Sweden, Norway and Denmark. Lancet Reg Health - Europe 35\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eWhitaker M et al (2022) Persistent COVID-19 symptoms in a community study of 606,434 people in England. Nat Commun 13\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eServier C, Porcher R, Pane I, Ravaud P, Tran V-T (2023) Trajectories of the evolution of post-COVID-19 condition, up to two years after symptoms onset. Int J Infect Dis 133:67\\u0026ndash;74\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eTran VT, Porcher R, Pane I, Ravaud P (2022) Course of post COVID-19 disease symptoms over time in the ComPaRe long COVID prospective e-cohort. Nat Commun 13\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eTaquet M et al (2024) Cognitive and psychiatric symptom trajectories 2\\u0026ndash;3 years after hospital admission for COVID-19: a longitudinal, prospective cohort study in the UK. Lancet Psychiatry 11:696\\u0026ndash;708\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eHampshire A et al (2021) Cognitive deficits in people who have recovered from COVID-19. \\u003cem\\u003eEClinicalMedicine\\u003c/em\\u003e 39\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eCheetham NJ et al (2023) The effects of COVID-19 on cognitive performance in a community-based cohort: a COVID symptom study biobank prospective cohort study. \\u003cem\\u003eEClinicalMedicine\\u003c/em\\u003e 62\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eHampshire A et al (2024) Cognition and Memory after Covid-19 in a Large Community Sample. N Engl J Med 390:806\\u0026ndash;818\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eWood GK et al (2025) Posthospitalization COVID-19 cognitive deficits at 1 year are global and associated with elevated brain injury markers and gray matter volume reduction. Nat Med 31:245\\u0026ndash;257\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eRouten A, Khunti K (2024) Long-term outcomes in hospitalised COVID-19 survivors and future research priorities. Lancet Respir Med 12:7\\u0026ndash;8\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eHuang L et al (2022) Health outcomes in people 2 years after surviving hospitalisation with COVID-19: a longitudinal cohort study. Lancet Respir Med 10:863\\u0026ndash;876\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eFern\\u0026aacute;ndez-de-las-Pe\\u0026ntilde;as C, Arias-Naval\\u0026oacute;n JA, Mart\\u0026iacute;n-Guerrero JD, Pellicer-Valero OJ, Cigar\\u0026aacute;n-M\\u0026eacute;ndez M (2024) Trajectory of anxiety/depressive symptoms and sleep quality in individuals who had been hospitalized by COVID-19: The LONG-COVID-EXP multicenter study. J Psychosom Res 179\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eFern\\u0026aacute;ndez-De-Las-Pe\\u0026ntilde;as C et al (2023) Trajectory curves of post-COVID anxiety/depressive symptoms and sleep quality in previously hospitalized COVID-19 survivors: The LONG-COVID-EXP-CM multicenter study. Psychol Med 53:4298\\u0026ndash;4299\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eFern\\u0026aacute;ndez-de-las-Pe\\u0026ntilde;as C et al (2023) Trajectory of post-COVID brain fog, memory loss, and concentration loss in previously hospitalized COVID-19 survivors: the LONG-COVID-EXP multicenter study. Front Hum Neurosci 17\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eFern\\u0026aacute;ndez-de-las-Pe\\u0026ntilde;as C et al (2023) Trajectory of Post-COVID Self-Reported Fatigue and Dyspnoea in Individuals Who Had Been Hospitalized by COVID-19: The LONG-COVID-EXP Multicenter Study. \\u003cem\\u003eBiomedicines\\u003c/em\\u003e 11\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eWulf Hanson S et al (2022) Estimated Global Proportions of Individuals With Persistent Fatigue, Cognitive, and Respiratory Symptom Clusters Following Symptomatic COVID-19 in 2020 and 2021. JAMA 328:1604\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eDavis HE, McCorkell L, Vogel JM, Topol EJ (2023) Long COVID: major findings, mechanisms and recommendations. Nat Rev Microbiol 21:133\\u0026ndash;146\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eSzewczyk W et al (2024) Long COVID and recovery from Long COVID: quality of life impairments and subjective cognitive decline at a median of 2 years after initial infection. BMC Infect Dis 24:1241\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eZhang H et al (2024) 3-year outcomes of discharged survivors of COVID-19 following the SARS-CoV-2 omicron (B.1.1.529) wave in 2022 in China: a longitudinal cohort study. Lancet Respir Med 12:55\\u0026ndash;66\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eCai M, Xie Y, Topol EJ, Al-Aly Z (2024) Three-year outcomes of post-acute sequelae of COVID-19. Nat Med 30:1564\\u0026ndash;1573\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eRanjan Y et al (2019) RADAR-Base: Open Source Mobile Health Platform for Collecting, Monitoring, and Analyzing Data Using Sensors, Wearables, and Mobile Devices. JMIR Mhealth Uhealth 7:e11734\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eStewart C et al (2024) Physiological presentation and risk factors of long COVID in the UK using smartphones and wearable devices: a longitudinal, citizen science, case\\u0026ndash;control study. Lancet Digit Health 6:e640\\u0026ndash;e650\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eFraser A et al (2013) Cohort Profile: The Avon Longitudinal Study of Parents and Children: ALSPAC mothers cohort. Int J Epidemiol 42:97\\u0026ndash;110\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eBoyd A et al (2013) Cohort Profile: The \\u0026lsquo;Children of the 90s\\u0026rsquo;\\u0026mdash;the index offspring of the Avon Longitudinal Study of Parents and Children. Int J Epidemiol 42:111\\u0026ndash;127\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eVerdi S et al (2019) TwinsUK: The UK Adult Twin Registry Update. Twin Res Hum Genet 22:523\\u0026ndash;529\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eWahlgren C et al (2023) Two-year follow-up of patients with post-COVID-19 condition in Sweden: a prospective cohort study. Lancet Reg Health - Europe 28:100595\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003ePeter RS et al (2025) Persistent symptoms and clinical findings in adults with post-acute sequelae of COVID-19/post-COVID-19 syndrome in the second year after acute infection: A population-based, nested case-control study. PLoS Med 22:e1004511\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eZheng C et al (2024) Physical exercise-related manifestations of long COVID: A systematic review and meta-analysis. J Exerc Sci Fit 22:341\\u0026ndash;349\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eDurstenfeld MS et al (2022) Use of Cardiopulmonary Exercise Testing to Evaluate Long COVID-19 Symptoms in Adults. JAMA Netw Open 5:e2236057\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eJamieson A et al (2024) Mechanisms underlying exercise intolerance in long\\u0026thinsp;\\u0026lt;\\u0026thinsp;scp\\u0026thinsp;\\u0026gt;\\u0026thinsp;COVID : An accumulation of multisystem dysfunction. Physiol Rep 12\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eDiciolla NS et al (2025) Physical Activity and Sedentary Behaviour in People with Long COVID: A Follow-Up from 12 to 18 Months After Discharge. J Clin Med 14:3641\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eRosa-Souza FJ et al (2024) Association of physical symptoms with accelerometer-measured movement behaviors and functional capacity in individuals with Long COVID. Sci Rep 14:20652\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003ePlekhanova T et al (2022) Device-assessed sleep and physical activity in individuals recovering from a hospital admission for COVID-19: a multicentre study. Int J Behav Nutr Phys Activity 19:94\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eDani M et al (2021) Autonomic dysfunction in \\u0026lsquo;long COVID\\u0026rsquo;: rationale, physiology and management strategies. Clin Med 21:e63\\u0026ndash;e67\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eRaj SR et al (2021) Long-COVID postural tachycardia syndrome: an American Autonomic Society statement. Clin Auton Res 31:365\\u0026ndash;368\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eLadlow P et al (2022) Dysautonomia following COVID-19 is not associated with subjective limitations or symptoms but is associated with objective functional limitations. Heart Rhythm 19:613\\u0026ndash;620\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eChinvararak C, Chalder T (2023) Prevalence of sleep disturbances in patients with long COVID assessed by standardised questionnaires and diagnostic criteria: A systematic review and meta-analysis. J Psychosom Res 175:111535\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eTański W, Tomasiewicz A (2024) Jankowska-Polańska, B. Sleep Disturbances as a Consequence of Long COVID-19: Insights from Actigraphy and Clinimetric Examinations\\u0026mdash;An Uncontrolled Prospective Observational Pilot Study. J Clin Med 13:839\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eAlzueta E et al (2022) An international study of post-COVID sleep health. Sleep Health 8:684\\u0026ndash;690\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eZhu B et al (2022) Associations between sleep variability and cardiometabolic health: A systematic review. Sleep Med Rev 66:101688\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eQin S, Chee M (2024) The Emerging Importance of Sleep Regularity on Cardiovascular Health and Cognitive Impairment in Older Adults: A Review of the Literature. Nat Sci Sleep Volume 16:585\\u0026ndash;597\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eGao Y et al (2025) Identification of soluble biomarkers that associate with distinct manifestations of long COVID. Nat Immunol 26:692\\u0026ndash;705\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003ePaval N-E et al (2025) MicroRNAs in long COVID: roles, diagnostic biomarker potential and detection. Hum Genomics 19:90\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eJamieson A, Jones S, Chaturvedi N, Hughes AD, Orini M (2024) Accuracy of smartwatches for the remote assessment of exercise capacity. Sci Rep 14:22994\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eWilliams K, Jamieson A, Chaturvedi N, Hughes A, Orini M Validation of Wearable Derived Heart Rate Variability and Oxygen Saturation from the Garmin\\u0026rsquo;s Health Snapshot. \\u003cem\\u003eComput Cardiol (\\u003c/em\\u003e(2010)) (2023)) (2023)\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eMiller DJ, Sargent C, Roach GD (2022) A Validation of Six Wearable Devices for Estimating Sleep, Heart Rate and Heart Rate Variability in Healthy Adults. Sensors 22:6317\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eStewart C et al (2021) Investigating the Use of Digital Health Technology to Monitor COVID-19 and Its Effects: Protocol for an Observational Study (Covid Collab Study). JMIR Res Protoc 10:e32587\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eJamieson A et al (2025) Cohort profile: characterisation, determinants, mechanisms and consequences of the long-term effects of COVID-19 \\u0026ndash; providing the evidence base for health care services (CONVALESCENCE) in the UK. BMJ Open 15:e094760\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eMoayyeri A, Hammond CJ, Valdes AM, Spector TD (2013) Cohort profile: Twinsuk and healthy ageing twin study. Int J Epidemiol 42:76\\u0026ndash;85\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eSWAIN, D. P., LEUTHOLTZ, B. C., KING, M. E., HAAS, L. A. \\u0026amp; BRANCH, J. D. Relationship between% heart rate reserve and%??VO2reserve in treadmill exercise. \\u003cem\\u003eMedicine \\u0026amp; Science in Sports \\u0026amp; Exercise\\u003c/em\\u003e 30, 318\\u0026ndash;321 (1998)\\u003c/span\\u003e\\u003c/li\\u003e\\u003c/ol\\u003e\"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":true,\"hideJournal\":false,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"nature-portfolio\",\"isNatureJournal\":true,\"hasQc\":false,\"allowDirectSubmit\":false,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"\",\"title\":\"Nature Portfolio\",\"twitterHandle\":\"\",\"acdcEnabled\":false,\"dfaEnabled\":false,\"editorialSystem\":\"ejp\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":false},\"keywords\":\"Post-acute sequelae of COVID-19, long COVID, wearable, mobile health, cognitive function, physical activity, cardiovascular function, mental health, sleep\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-7876232/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-7876232/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003eLong COVID remains poorly understood and previous studies have mainly focused on describing symptoms or single-system health deficits in severe cases. In the CONVALESCENCE case-control study, 306 community-based participants (aged 57 [IQR 41\\u0026ndash;63], 80% female) were enrolled 2.0 [1.7, 2.5] years post-infection and were monitored for 12 months using a smartwatch and apps to assess mental health, cognitive function, physical activity, exercise capacity, cardiorespiratory function, autonomic function and sleep. Compared to controls (N\\u0026thinsp;=\\u0026thinsp;152), long COVID cases with fatigue-cluster symptoms (N\\u0026thinsp;=\\u0026thinsp;50) presented small persistent deficits across all systems with no evidence of recovery over 12 months. Participants with fatigue-cluster symptoms not attributable to long COVID (N\\u0026thinsp;=\\u0026thinsp;28) showed a similar trend, while participants with prior long COVID but no fatigue-cluster symptoms (N\\u0026thinsp;=\\u0026thinsp;76) were not different from controls. The lack of significant recovery between years two and three post-infection demonstrates protracted multi-system deficits for fatigue-cluster cases and highlights ongoing needs to better understand and manage long COVID.\\u003c/p\\u003e\",\"manuscriptTitle\":\"Persistent Multi-System Impairments Detected by Wearable Monitoring in Long COVID Cases reporting Fatigue up to Three Years Post-Infection\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2025-11-07 06:02:03\",\"doi\":\"10.21203/rs.3.rs-7876232/v1\",\"editorialEvents\":[],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"nature-communications\",\"isNatureJournal\":true,\"hasQc\":false,\"allowDirectSubmit\":false,\"externalIdentity\":\"NCOMMS\",\"sideBox\":\"Learn more about [Nature Communications](http://www.nature.com/ncomms/)\",\"snPcode\":\"\",\"submissionUrl\":\"https://mts-ncomms.nature.com/\",\"title\":\"Nature Communications\",\"twitterHandle\":\"\",\"acdcEnabled\":true,\"dfaEnabled\":true,\"editorialSystem\":\"ejp\",\"reportingPortfolio\":\"Nature Communications\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":false}}],\"origin\":\"\",\"ownerIdentity\":\"d7749e26-afbb-4999-a9f5-5614115a612c\",\"owner\":[],\"postedDate\":\"November 7th, 2025\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"under-review\",\"subjectAreas\":[{\"id\":57042367,\"name\":\"Health sciences/Medical research/Epidemiology\"},{\"id\":57042368,\"name\":\"Health sciences/Diseases/Infectious diseases\"}],\"tags\":[],\"updatedAt\":\"2025-11-07T06:02:03+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2025-11-07 06:02:03\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-7876232\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-7876232\",\"identity\":\"rs-7876232\",\"version\":[\"v1\"]},\"buildId\":\"8U1c8b4HqxoKbykW_rLl7\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}