Circulating serum metabolites as predictors of dementia: a machine learning approach in a 21-year follow-up of the Whitehall II cohort study.

BACKGROUND: Age is the strongest risk factor for dementia and there is considerable interest in identifying scalable, blood-based biomarkers in predicting dementia. We examined the role of midlife serum metabolites using a machine learning approach and determined whether the selected metabolites improved prediction accuracy beyond the effect of age. METHODS: Five thousand three hundred seventy-four participants from the Whitehall II study, mean age 55.8 (standard deviation (SD) 6.0) years in 1997-1999 when 233 metabolites were quantified using nuclear magnetic resonance metabolomics. Participants were followed for a median 21.0 (IQR 20.4, 21.7) years for clinically-diagnosed dementia (N=329). Elastic net penalized Cox regression with 100 repetitions of nested cross-validation was used to select models that improved prediction accuracy for incident dementia compared to an age-only model. Risk scores reflecting the frequency with which predictors appeared in the selected models were constructed, and their predictive accuracy was examined using Royston's R2, Akaike's information criterion, sensitivity, specificity, C-statistic and calibration. RESULTS: Sixteen of the 100 models had a better c-statistic compared to an age-only model and 15 metabolites were selected at least once in all 16 models with glucose present in all models. Five risk scores, reflecting the frequency of selection of metabolites, and a 1-SD increment in all five risk scores was associated with higher dementia risk (HR between 3.13 and 3.26). Three of these, constituted of 4, 5 and 15 metabolites, had better prediction accuracy (c-statistic from 0.788 to 0.796) compared to an age-only model (c-statistic 0.780), all p<0.05. CONCLUSIONS: Although there was robust evidence for the role of glucose in dementia, metabolites measured in midlife made only a modest contribution to dementia prediction once age was taken into account.

Association of physical activity, sedentary behaviour, and daylight exposure with sleep in an ageing population: findings from the Whitehall accelerometer sub-study.

BACKGROUND: Ageing is accompanied by changes in sleep, while poor sleep is suggested as a risk factor for several health outcomes. Non-pharmacological approaches have been proposed to improve sleep in elderly; their impact remains to be investigated. The aim of this study was to examine the independent day-to-day associations of physical behaviours and daylight exposure with sleep characteristics among older adults. METHODS: Data were drawn from 3942 participants (age range: 60-83 years; 27% women) from the Whitehall II accelerometer sub-study. Day-to-day associations of objectively-assessed daytime physical behaviours (sedentary behaviour, light-intensity physical activity (LIPA), moderate-to-vigorous physical activity (MVPA), mean acceleration, physical activity chronotype) and daylight exposure (proportion of waking window with light exposure > 1000 lx and light chronotype) with sleep characteristics were examined using mixed models. RESULTS: A 10%-increase in proportion of the waking period spent sedentary was associated with 5.12-minute (4.31, 5.92) later sleep onset and 1.76-minute shorter sleep duration (95%confidence interval: 0.86, 2.66). Similar increases in LIPA and MVPA were associated with 6.69 (5.67, 7.71) and 4.15 (2.49, 5.81) earlier sleep onset respectively and around 2-minute longer sleep duration (2.02 (0.87, 3.17) and 2.23 (0.36, 4.11), respectively), although the association was attenuated for MVPA after adjustment for daylight exposure (1.11 (- 0.84, 3.06)). A 3-hour later physical activity chronotype was associated with a 4.79-minute later sleep onset (4.15, 5.43) and 2.73-minute shorter sleep duration (1.99, 3.47). A 10%-increase in proportion of waking period exposed to light> 1000 lx was associated with 1.36-minute longer sleep (0.69, 2.03), independently from mean acceleration. Associations found for sleep duration were also evident for duration of the sleep windows with slightly larger effect size (for example, 3.60 (2.37, 4.82) minutes for 10%-increase in LIPA), resulting in associations with sleep efficiency in the opposite direction (for example, - 0.29% (- 0.42, - 0.16) for 10%-increase in LIPA). Overall, associations were stronger for women than for men. CONCLUSIONS: In this study, higher levels of physical activity and daylight exposure were associated with slightly longer sleep in older adults. Given the small effect sizes of the associations, increased physical activity and daylight exposure might not be enough to improve sleep.

Association of daily composition of physical activity and sedentary behaviour with incidence of cardiovascular disease in older adults.

BACKGROUND: Moderate-to-vigorous physical activity (MVPA) is proposed as key for cardiovascular diseases (CVD) prevention. At older ages, the role of sedentary behaviour (SB) and light intensity physical activity (LIPA) remains unclear. Evidence so far is based on studies examining movement behaviours as independent entities ignoring their co-dependency. This study examines the association between daily composition of objectively-assessed movement behaviours (MVPA, LIPA, SB) and incident CVD in older adults. METHODS: Whitehall II accelerometer sub-study participants free of CVD at baseline (N = 3319, 26.7% women, mean age = 68.9 years in 2012-2013) wore a wrist-accelerometer from which times in SB, LIPA, and MVPA during waking period were extracted over 7 days. Compositional Cox regression was used to estimate the hazard ratio (HR) for incident CVD for daily compositions of movement behaviours characterized by 10 (20 or 30) minutes greater duration in one movement behaviour accompanied by decrease in another behaviour, while keeping the third behaviour constant, compared to reference composition. Analyses were adjusted for sociodemographic, lifestyle, cardiometabolic risk factors and multimorbidity index. RESULTS: Of the 3319 participants, 299 had an incident CVD over a mean (SD) follow-up of 6.2 (1.3) years. Compared to daily movement behaviour composition with MVPA at recommended 21 min per day (150 min/week), composition with additional 10 min of MVPA and 10 min less SB was associated with smaller risk reduction - 8% (HR, 0.92; 95% CI, 0.87-0.99) - than the 14% increase in risk associated with a composition of similarly reduced time in MVPA and more time in SB (HR, 1.14; 95% CI, 1.02-1.27). For a given MVPA duration, the CVD risk did not differ as a function of LIPA and SB durations. CONCLUSIONS: Among older adults, an increase in MVPA duration at the expense of time in either SB or LIPA was found associated with lower incidence of CVD. This study lends support to public health guidelines encouraging increase in MVPA or at least maintain MVPA at current duration.

Timing of Grape Downy Mildew Onset in Bordeaux Vineyards.

Grapevine downy mildew (GDM) is a severe disease of grapevines. Because of the lack of reliable information about the dates of GDM symptom onset, many vine growers begin fungicide treatments early in the season. We evaluate the extent to which such preventive treatments are justified. Observational data for 266 untreated sites for the years between 2010 and 2017 were used to estimate the timing of GDM onset on vines and bunches of grapes in South West France (Bordeaux region) through survival analyses. The onset of GDM was not apparent on vines and bunches before early to mid-May, and the rate of GDM symptom appearance was highly variable across years. Depending on the year, 50% of the plots displayed symptoms between mid-May and late June for vines. For several years, our statistical analysis revealed that the proportion of plots with no symptoms was high in early August on vines (27.5 and 43.7% in 2013 and 2016) and on bunches (between 23 and 79% in 2011, 2013, and 2016). We found a significant effect of the amount of rainfall in spring on the date of symptom appearance. These results indicate that preventive fungicide application is unjustified in many vineyards, and that regional disease surveys should be used to adjust fungicide treatment dates according to local characteristics, in particular according to rainfall conditions in spring.

Association between profiles of accelerometer-measured daily movement behaviour and mortality risk: a prospective cohort study of British older adults.

Objectives: We identified profiles of wake-time movement behaviours (sedentary behaviours, light intensity physical activity and moderate-to-vigorous physical activity) based on accelerometer-derived features among older adults and then examined their association with all-cause mortality. Methods: Data were drawn from a prospective cohort of 3991 Whitehall II accelerometer substudy participants aged 60-83 years in 2012-2013. Daily movement behaviour profiles were identified using k-means cluster analysis based on 13 accelerometer-assessed features characterising total duration, frequency, bout duration, timing and activity intensity distribution of movement behaviour. Cox regression models were used to assess the association between derived profiles and mortality risk. Results: Over a mean follow-up of 8.1 (SD 1.3) years, a total of 410 deaths were recorded. Five distinct profiles were identified and labelled as 'active' (healthiest), 'active sitters', 'light movers', 'prolonged sitters', and 'most sedentary' (most deleterious). In model adjusted for sociodemographic, lifestyle, and health-related factors, compared with the 'active' profile, 'active sitters' (HR 1.57, 95% CI 1.01 to 2.44), 'light movers' (HR 1.75, 95% CI 1.17 to 2.63), 'prolonged sitters' (HR 1.67, 95% CI 1.11 to 2.51), 'most sedentary' (HR 3.25, 95% CI 2.10 to 5.02) profiles were all associated with a higher risk of mortality. Conclusion: Given the threefold higher mortality risk among those with a 'most sedentary' profile, public health interventions may target this group wherein any improvement in physical activity and sedentary behaviour might be beneficial.

Identification of physical activity and sedentary behaviour dimensions that predict mortality risk in older adults: Development of a machine learning model in the Whitehall II accelerometer sub-study and external validation in the CoLaus study.

Background: Identification of new physical activity (PA) and sedentary behaviour (SB) features relevant for health at older age is important to diversify PA targets in guidelines, as older adults rarely adhere to current recommendations focusing on total duration. We aimed to identify accelerometer-derived dimensions of movement behaviours that predict mortality risk in older populations. Methods: We used data on 21 accelerometer-derived features of daily movement behaviours in 3991 participants of the UK-based Whitehall II accelerometer sub-study (25.8% women, 60-83 years, follow-up: 2012-2013 to 2021, mean = 8.3 years). A machine-learning procedure was used to identify core PA and SB features predicting mortality risk and derive a composite score. We estimated the added predictive value of the score compared to traditional sociodemographic, behavioural, and health-related risk factors. External validation in the Switzerland-based CoLaus study (N = 1329, 56.7% women, 60-86 years, follow-up: 2014-2017 to 2021, mean = 3.8 years) was conducted. Findings: In total, 11 features related to overall activity level, intensity distribution, bouts duration, frequency, and total duration of PA and SB, were identified as predictors of mortality in older adults and included in a composite score. Both in the derivation and validation cohorts, the score was associated with mortality (hazard ratio = 1.10 (95% confidence interval = 1.05-1.15) and 1.18 (1.10-1.26), respectively) and improved the predictive value of a model including traditional risk factors (increase in C-index = 0.007 (0.002-0.014) and 0.029 (0.002-0.055), respectively). Interpretation: The identified accelerometer-derived PA and SB features, beyond the currently recommended total duration, might be useful for screening of older adults at higher mortality risk and for diversifying PA and SB targets in older populations whose adherence to current guidelines is low. Funding: National Institute on Aging; UK Medical Research Council; British Heart Foundation; Wellcome Trust; French National Research Agency; GlaxoSmithKline; Lausanne Faculty of Biology and Medicine; Swiss National Science Foundation.

Individual Barriers to an Active Lifestyle at Older Ages Among Whitehall II Study Participants After 20 Years of Follow-up.

Importance: Identification of individual-level barriers associated with decreased activity in older age is essential to inform effective strategies for preventing the health outcomes associated with high sedentary behavior and lack of physical activity during aging. Objective: To assess cross-sectional and prospective associations of a large set of factors with objectively assessed sedentary time and physical activity at older age. Design, Setting, and Participants: This population-based cohort study was conducted among participants in the Whitehall II accelerometer substudy with accelerometer data assessed in 2012 to 2013. Among 4880 participants invited to the accelerometer substudy, 4006 individuals had valid accelerometer data. Among them, 3808 participants also had factors assessed in 1991 to 1993 (mean [SD] follow-up time, 20.3 [0.5] years), 3782 participants had factors assessed in 2002 to 2004 (mean [SD] follow-up time, 9.1 [0.3] years), and 3896 participants had factors assessed in 2012 to 2013 (mean follow up time, 0 years). Data were analyzed from May 2020 through July 2021. Exposures: Sociodemographic factors (ie, age, sex, race and ethnicity, occupational position, and marital status), behavioral factors (ie, smoking, alcohol intake, and fruit and vegetable intake), and health-related factors (ie, body mass index, 36-Item Short Form Health Survey (SF-36) physical and mental component summary scores [PCS and MCS], and number of chronic conditions) were assessed among 3808 individuals in 1991 to 1993; 3782 individuals in 2002 to 2004; and 3896 individuals in 2012 to 2013. High alcohol intake was defined as more than 14 units of alcohol per week, and high fruit and vegetable intake was defined as twice daily or more. Main Outcomes and Measures: Accelerometer-assessed time spent in sedentary behavior, light-intensity physical activity (LIPA), and moderate to vigorous physical activity (MVPA) in 2012 to 2013 were analyzed in 2021 using multivariate linear regressions. Results: A total of 3896 participants (986 [25.3%] women; age range, 60-83 years; mean [SD] age, 69.4 [5.7] years) had accelerometer data and exposure factors available in 2012 to 2013. Older age, not being married or cohabiting, having overweight, having obesity, more chronic conditions, and poorer SF-36 PCS, assessed in midlife or later life, were associated with increased sedentary time at the expense of time in physical activity. Mean time differences ranged from 9.8 min/d (95% CI, 4.1 to 15.6 min/d) of sedentary behavior per 10-point decrease in SF-36 PCS to 51.4 min/d (95% CI, 37.2 to65.7 min/d) of sedentary behavior for obesity vs reference range weight, from -6.2 min/d (95% CI, -8.4 to -4.1 min/d) of LIPA per 5 years of age to -28.0 min/d (95% CI, -38.6 to -17.4 min/d) of LIPA for obesity vs reference range weight, and from -5.3 min/d (95% CI, -8.2 to -2.4 min/d) of MVPA per new chronic condition to -23.4 min/d (95% CI, -29.2 to -17.6 min/d) of MVPA for obesity vs reference range weight in 20-year prospective analyses for men. There was also evidence of clustering of behavioral factors: high alcohol intake, high fruit and vegetable consumption, and no current smoking were associated with decreased sedentary time (mean time difference in cross-sectional analysis in men: -12.7 min/d [95% CI, -19.8 to -5.5 min/d]; -6.0 min/d [95% CI, -12.3 to -0.2]; and -37.4 min/d [95% CI, - 56.0 to -18.8 min/d], respectively) and more physical activity. Conclusions and Relevance: This study found a large range of individual-level barriers associated with a less active lifestyle in older age, including sociodemographic, behavioral, and health-related factors. These barriers were already evident in midlife, suggesting the importance of early implementation of targeted interventions to promote physical activity and reduce sedentary time.

Objectively Measured Total Sedentary Time and Pattern of Sedentary Accumulation in Older Adults: Associations With Incident Cardiovascular Disease and All-Cause Mortality.

BACKGROUND: We examined associations of total duration and pattern of accumulation of objectively measured sedentary behavior (SB) with incident cardiovascular disease (CVD) and all-cause mortality among older adults. METHODS: Total sedentary time and 8 sedentary accumulation pattern metrics were extracted from accelerometer data of 3 991 Whitehall II study participants aged 60-83 years in 2012-2013. Incident CVD and all-cause mortality were ascertained up to March 2019. RESULTS: Two hundred and ninety-nine CVD cases and 260 deaths were recorded over a mean (standard deviation [SD]) follow-up of 6.2 (1.3) and 6.4 (0.8) years, respectively. Adjusting for sociodemographic and behavioral factors, 1-SD (100.2 minutes) increase in total sedentary time was associated with 20% higher CVD risk (hazard ratio [95% confidence interval]: 1.20 [1.05-1.37]). More fragmented SB was associated with reduced CVD risk (eg, 0.86 [0.76-0.97] for 1-SD [6.2] increase in breaks per sedentary hour). Associations were not evident once health-related factors and moderate-to-vigorous physical activity (MVPA) were considered. For all-cause mortality, associations with more fragmented SB (eg, 0.73 [0.59-0.91] for breaks per sedentary hour) were found only among the youngest older group (<74 years; p for interaction with age < .01) independently from all covariates. CONCLUSIONS: In this study, no associations of total sedentary time and sedentary accumulation patterns with incident CVD and all-cause mortality were found in the total sample once MVPA was considered. Our findings of reduced mortality risk with less total and more fragmented SB independent from MVPA among individuals <74 years need to be replicated to support the recent recommendations to reduce and fragment SB.

Delaying the first grapevine fungicide application reduces exposure on operators by half.

Downy mildew is a severe disease of grapevines treated by repeated fungicide applications during the growing season. The impact of these treatments on human health is currently under scrutiny. Fungicide application long before disease onset is not thought to be greatly beneficial for grape production, but the first fungicide treatment is applied at least six weeks before disease onset in more than 50% of the vineyards in the Bordeaux region, a major French vine-growing area. We estimate that applying one fungicide every two weeks at disease onset would reduce fungicide applications against downy mildew by 56% (95%IC = [51.0%, 61.3%]), on average, relative to current levels. This decrease is slightly greater than the level of exposure reduction resulting from the random suppression of one out of every two fungicide treatments (i.e. 50%). The reduction is lower when treatments are sprayed weekly but still reaches at least 12.4% (95%IC = [4.3%, 20.8%]) in this case. We show that this and other strategies reducing the number of treatments would decrease operator exposure to pesticides as effectively as the use of various types of personal protective equipments in the Bordeaux region. The implementation of this strategy would significantly decrease fungicide use, health risks, and adverse environmental impacts of vineyards.

Forecasting severe grape downy mildew attacks using machine learning.

Grape downy mildew (GDM) is a major disease of grapevine that has an impact on both the yields of the vines and the quality of the harvested fruits. The disease is currently controlled by repetitive fungicide treatments throughout the season, especially in the Bordeaux vineyards where the average number of fungicide treatments against GDM was equal to 10.1 in 2013. Reducing the number of treatments is a major issue from both an environmental and a public health point of view. One solution would be to identify vineyards that are likely to be heavily attacked in spring and then apply fungicidal treatments only to these situations. In this perspective, we use here a dataset including 9 years of GDM observations to develop and compare several generalized linear models and machine learning algorithms predicting the probability of high incidence and severity in the Bordeaux region. The algorithms tested use the date of disease onset and/or average monthly temperatures and precipitation as input variables. The accuracy of the tested models and algorithms is assessed by year-by-year cross validation. LASSO, random forest and gradient boosting algorithms show better performance than generalized linear models. The date of onset of the disease has a greater influence on the accuracy of forecasts than weather inputs and, among weather inputs, precipitation has a greater influence than temperature. The best performing algorithm was selected to evaluate the impact of contrasted climate scenarios on GDM risk levels. Results show that risk of GDM at bunch closure decreases with reduced rainfall and increased temperatures in April-May. Our results also show that the use of fungicide treatment decision rules that take into account local characteristics would reduce the number of treatments against GDM in the Bordeaux vineyards compared to current practices by at least 50%.