Physical Function and Survival in Older Adults: A longitudinal study accounting for time-varying effects.

PURPOSE OF THE STUDY: Variation in physical function in older adults over time raises several methodological challenges in the study of its association with survival, many of which have largely been overlooked in previous studies. The objective of this study is to examine the relationship between time-varying measures of physical function and survival in men and women aged 70 years and over, while accounting for the time-varying effects of health and lifestyle characteristics. METHODS: 1,846 women and 1,245 men in the Cardiovascular Health Study followed annually for up to 10 years beginning at age 70-74 years were included. We estimated the effect of gait speed and grip strength on survival over the subsequent year, using age as the timescale. RESULTS: A 0.1m/s higher gait speed was associated with a 12% decrease in the likelihood of death in the subsequent year among women (HR 0.88, 95% CI 0.82-0.94). There was no statistically significant effect of gait speed on survival among men (HR 0.97, 95% CI 0.91 to 1.03), or of grip strength on survival among women (HR 0.97, 95% CI 0.95-1.00) or men (HR 0.99, 95% CI 0.97-1.01), over one year. CONCLUSIONS: Upon using time-varying measures of physical function while accounting for time-varying effects of health and lifestyle characteristics, higher gait speed was associated with increased survival among the women in our study. We found no evidence of an association between gait speed and one-year survival in men, or between grip strength and one-year survival in women or men.

The association between physical function and proximity to death in older adults: a multilevel analysis of 4,150 decedents from the Cardiovascular Health Study.

PURPOSE: When examining whether poor physical function is a risk factor for imminent death in older adults, one challenge is the lack of a meaningful time origin, a time point on which the estimate of time-to-death is anchored. In this study, we overcame this challenge by discarding the traditional-and flawed-approach of survival analysis with "time since beginning of follow up" as the time variable, and instead used a novel analytic approach that uses time-to-death as a covariate to examine its association with physical function. METHODS: Physical function and other covariates were measured annually in the Cardiovascular Health Study on 4150 individuals followed up to their time of death. Using multilevel models, we estimated gait speed and grip strength in relation to two time axes: age and proximity to death. RESULTS: As individuals approached death, both gait speed and grip strength decreased significantly. However, after adjustment for health and lifestyle covariates, there was significant variation in the level of physical function between individuals. CONCLUSION: Although physical function was significantly associated with time-to-death, there was significant variation in level of physical function between individuals at comparable proximity to death. A better understanding of these variations is needed before measures of physical function are recommended as a clinical tool for identifying individuals at high risk of death.

Profiling obesity phenotypes and trajectories in older adults of the Quebec NuAge cohort on nutrition and successful aging: A cluster analysis.

Obesity in older adults results from several interacting factors. Consequently, interventions have shown mitigated effects. We determined (a) the different subgroups of older adults with obesity based on clusters of associated comorbidities and (b) the trajectory of these clusters to assess their stability over 3 years and factors contributing to transitions. Obese men (n = 193; body mass index [BMI] = 33.15 ± 2.69 kg/m2 ) and women (n = 220; BMI = 33.71 ± 3.71 kg/m2 ) aged between 68 and 82 years were studied. Outcome variables were body composition, strength, physical capacity (PC), nutrition, psychological and physical health and social participation. Cluster analyses, stratified by sex, were used to identify obesity profiles at baseline and follow-up. Three profiles were identified, based on general health (GH), psychological health (PH) and PC: Cluster 1: healthy obese (GH+, PH+, PC+); Cluster 2: obese with low PC (GH+/-, PH+/-, PC-); Cluster 3: unhealthy obese (GH-, PH-, PC-). After 3 years, 61.2% and 70.2% of men and women remained in their initial cluster, compared to 20.4% and 13.7% who transitioned towards a worse health cluster and 18.3% and 16.0% who transitioned towards a more favourable cluster, partly explained by changes in physical health for men and physical health and PH for women. The results of this study show that targeting physical function in men and physical health and PH functions in women could prevent further health decline in older adults with obesity. Further studies are needed to investigate the role of these clusters in the prediction of cardiometabolic complications and mortality.

Validity and Reliability of a Short Diet Questionnaire to Estimate Dietary Intake in Older Adults in a Subsample of the Canadian Longitudinal Study on Aging.

This study assessed test-retest reliability and relative validity of the Short Diet Questionnaire (SDQ) and usability of an online 24 h recall among 232 participants (62 years ± 9.1; 49.6% female) from the Canadian Longitudinal Study on Aging (CLSA). Participants were asked to complete four 24 h dietary recalls (24HRs) using the Automated Self-Administered 24-h Dietary Assessment Tool (ASA24-Canada-2014), two SDQ administrations (prior to recalls one and four), and the System Usability Scale (SUS) for ASA24. For the SDQ administrations, Intraclass Correlation Coefficients ranged from 0.49 to 0.57 for nutrients and 0.35 to 0.72 for food groups. Mean intakes estimated from the SDQ were lower compared than those from the 24HRs. For nutrients, correlation coefficients were highest for fiber, calcium, and vitamin D (45⁻64 years: 0.59, 0.50, 0.51; >65 years: 0.29, 0.38, 0.49, p < 0.01); Kappas ranged from 0.14 to 0.37 in those 45⁻64 years and 0.17 to 0.32 in participants >65 years. Among the 70% who completed all recalls independently, the SUS indicated poor usability, though the majority reported feeling confident using ASA24. Overall, the SDQ captures intake with varying test-retest reliability and accuracy by nutrient and age. Further research is needed to inform use of a more comprehensive dietary measure in the CLSA.

Adaptation transculturelle en français du Edinburgh Feeding Evaluation in Dementia (EdFED) Scale : un questionnaire pour évaluer les difficultés à s'alimenter de personnes âgées présentant des troubles cognitifs en centre d'hébergement.

ABSTRACTThe Edinburgh Feeding Evaluation in Dementia (EdFED) Scale was designed to identify feeding difficulties for people with moderate to severe dementia. Its Canadian-French cross-cultural adaptation was carried out, as part of an experimental study, whose secondary objective was to document its construct validity. A back-translation method was followed. The EdFED-f was used during a meal with 26 elderly residents who had cognitive disorders. There was a significant correlation between EdFED-f scores and energy intakes calculated using a visual estimation of plate wash method (r=-0,50, p=0,009). EdFED-f scores also showed a significant difference according to the percentage of food consumed at meals (p=0,015). These results support the validity of the EdFED-f to assess feeding difficulties among elderly French-speaking Canadians living in residential and long-term care centers.

Letter to the Editor: RE: Costa et al. (''Nutritional Risk is Associated with Chronic Musculoskeletal Pain in Community-Dwelling Older Persons: The PAINEL Study'').

We read Costa and colleagues' report1 in 2016 in the Journal of Nutrition in Gerontology and Geriatrics with great interest. In this article, the authors reported a statistically significant association between chronic musculoskeletal pain and nutritional risk after controlling for age, gender, body mass index (BMI), depression (GDS, geriatric depression scale), and diabetes in a Brazilian community-dwelling older adult's population (the PAINEL study, see Table 2 in Costa et al. 1 . However, in their study, Costa and colleagues used the DETERMINE 2 Questionnaire, a questionnaire in which 3 out of the 10 questions are directly or indirectly related to pain (i.e., presence of an illness, tooth or mouth problems, and physical disability). The purpose of this letter is to discuss the potential drawbacks of including pain-related questions in nutrition risk screening tool when these tools are used to investigate the relationship between pain and nutritional risk.

Association Between Anticholinergic Drug Use and Health-Related Quality of Life in Community-Dwelling Older Adults.

BACKGROUND: The use of drugs with anticholinergic properties (AC drugs) has been associated with decreased functioning and impaired cognition in older adults. Studies assessing the association between AC-drug use and health-related quality of life (HRQoL) show conflicting results. OBJECTIVE: The aim was to evaluate the association between AC-drug use and HRQoL in community-dwelling older adults. METHODS: The NuAge cohort study enrolled 1793 men and women aged 68-82 years. The participants were free of disabilities in activities of daily living, not cognitively impaired at recruitment and followed annually for 3 years (December 2003-May 2005). AC-drug exposure was assessed using the Anticholinergic Cognitive Burden Scale (ACBS). HRQoL was assessed using the physical (PCS) and mental (MCS) component summaries of the 36-item Short Form Survey (SF-36) questionnaire. The association between AC drug and HRQoL was determined by a mixed model analysis using four annual time points. RESULTS: At recruitment the mean age was 74.4 ± 4.2 years, 52% were female and 33% of participants were prescribed at least one AC drug. The mean PCS and MCS (/100) scores were 49.0 ± 8.2 and 54.9 ± 8.1, respectively. In the mixed model analysis, an increase of 1 on the ACBS was associated with a decrease of -0.50 (95% CI -0.68 to -0.31) in the PCS and an increase of 0.19 (95% CI 0.01-0.37) in the MCS. CONCLUSIONS: In a cohort of generally healthy community-dwelling older adults, AC-drug exposure was associated with a statistically significant decrease in the PCS and increase in the MCS throughout the entire follow-up period. However, the effects on the PCS and MCS were small and likely not clinically relevant.

Reduction in targeted potentially inappropriate medication use in elderly inpatients: a pragmatic randomized controlled trial.

PURPOSE: The use of potentially inappropriate medications (PIMs) in hospitalized older adults is a complex problem, but the use of computerized alert systems (CAS) has shown some potential. The study's objective is to assess the change in PIM use with a CAS-based pharmacist-physician intervention model compared to usual clinical care. METHODS: Pragmatic single-site randomized controlled trial was conducted at a university teaching hospital. Hospitalizations identified with selected Beers or STOPP criteria were randomized to usual clinical care or to the CAS-based pharmacist-physician intervention. The primary outcome was PIM drug cessation or dosage decrease. Clinical relevance of the CAS alerts was assessed. RESULTS: Analyses included 231 patients who had 128 and 126 hospitalizations in the control and intervention groups, respectively. Patients had a mean age of 81, and 60% were female. In the intervention compared to the control group, drug cessation or dosage decrease were more frequent at 48 h post-alert (45.8 vs 15.9%; absolute difference 30.0%; 95%CI 13.8 to 46.1%) and at discharge from the hospital (48.1 vs 27.3%; absolute difference 20.8%; 95%CI 4.6 to 37.0%). In a post hoc analysis of all alerts, regardless of their clinical relevance, the absolute difference in drug cessation or dosage decrease between the intervention and control groups was 16.2% (95%CI 2.9 to 29.6%) at 48 h and 8.0% (95%CI -4.0 to 20.0%) at discharge from the hospital. CONCLUSIONS: In hospitalized older adults, a CAS-based pharmacist-physician intervention, compared to usual clinical care, resulted in significant higher number of drug cessation and dosage reductions for targeted PIMs.

A prospective evaluation of the depression-nutrient intake reverse causality hypothesis in a cohort of community-dwelling older Canadians.

Studies have investigated the potential protective effects that diet may have on late-life depression incidence. This disorder can, however, affect the person's food intake, widely known as the reverse causality hypothesis of depression. To test this hypothesis, we compared mean nutrient intakes from three 24-h recalls during the year depression was detected (Geriatric Depression Scale ≥11 or antidepressant medication) with intakes from 1 year earlier among community-dwelling older adults (67-83 years) followed up annually in the 4-year Québec Longitudinal Study on Nutrition and Aging, who were free of depression and cognitive impairment at baseline. Participants (n 158, 64·4 % female) who became depressed and had data available for all follow-up years were matched by age group and sex with non-depressed participants. General linear mixed models were adjusted for percentage changes in physical activity, functional autonomy and stressful life events reported at the time of positive screening. A significant group effect for the dietary intake of all three B-vitamins was observed, as depression cases had consistently lower dietary intakes than controls (P<0·01). Over time, intakes of dietary vitamin B12 declined within depressed participants in bivariate analysis, but there was no time×group effect for any nutrient tested in the multivariate analyses. Intakes of energy, protein, saturated fat and total dietary fibre did not change in cases v. CONTROLS: Among community-dwelling older adults, declines in dietary vitamins B6, B12 and folate may precede depression incidence. To help preventative efforts by programmes and practitioners, longitudinal cohorts of longer duration should investigate the extent of the decline in dietary intakes relative to the time of depression.

Even mealtime distribution of protein intake is associated with greater muscle strength, but not with 3-y physical function decline, in free-living older adults: the Quebec longitudinal study on Nutrition as a Determinant of Successful Aging (NuAge study).

Background: Functional status declines with aging, thus impeding autonomy. Recently, a more even mealtime distribution of dietary protein was positively associated with muscle mass, but the relation of this distribution to physical performance remains unknown.Objective: We examined the relation between mealtime protein-intake distribution and physical performance and its 3-y decline in community-dwelling older adults.Design: Three-year follow-up data from 827 men and 914 women (67-84 y) in the longitudinal study on nutrition and aging [Quebec longitudinal study on Nutrition as a Determinant of Successful Aging (NuAge study); Quebec, Canada] were analyzed. Physical performance, which was measured yearly, was grouped into the following 2 functional composite scores: muscle strength (handgrip, arm, and leg) and mobility (timed-up-and-go, chair stand, and walking speed). Dietary data were collected in 2 sets of three 24-h food recalls at baseline and year 2. The individual mealtime protein distribution was calculated as the CV (i.e., SD divided by the mean) of grams of protein per meal. A mixed model analysis was used to examine trajectories of muscle strength and mobility across time by sex as conditioned by the protein distribution and adjusted for potential covariates.Results: Physical performance deteriorated over 3 y with muscle strength declining more than the mobility score in men (-1.51 ± 1.68 compared with -0.66 ± 2.81) and women (-1.35 ± 1.77 compared with -0.78 ± 2.63) (means ± SD, P < 0.001). More-evenly distributed protein intake, independent of the total quantity, was associated with a higher muscle-strength score in both sexes throughout follow-up. It was also associated with a greater mobility score, but only in men and only before adjustment for covariates. Strength and mobility rates of decline were not affected by protein-intake distribution in either sex.Conclusions: In addition to the previously observed association with lean mass, an even distribution of daily protein intake across meals is independently associated with greater muscle strength, but not with the mobility score, in older adults. A longer-term investigation of the role of protein intake and its distribution on physical performance is warranted, as are intervention studies, to support future recommendations.

Functional disability and social participation restriction associated with chronic conditions in middle-aged and older adults.

BACKGROUND: We examine the population impact on functional disability and social participation of physical and mental chronic conditions individually and in combination. METHODS: Cross-sectional, population-based data from community-dwelling people aged 45 years and over living in the 10 Canadian provinces in 2008-2009 were used to estimate the population attributable risk (PAR) for functional disability in basic (ADL) and instrumental (IADL) activities of daily living and social participation restrictions for individual and combinations of chronic conditions, stratified by age and gender, after adjusting for confounding variables. RESULTS: Five chronic conditions (arthritis, depression, diabetes, heart disease and eye disease) made the largest contributions to ADL-related and IADL-related functional disability and social participation restrictions, with variation in magnitude and ranking by age and gender. While arthritis was consistently associated with higher PARs across gender and most age groups, depression, alone and in combination with the physical chronic conditions, was associated with ADL and IADL disability as well as social participation restrictions in the younger age groups, especially among women. Compared to women, the combinations of conditions associated with higher PARs in men more often included heart disease and diabetes. CONCLUSIONS: Our findings suggest that in community-dwelling middle-aged and older adults, the impact of combinations of mental and physical chronic conditions on functional disability and social participation restriction is substantial and differed by gender and age. Recognising the differences in the drivers of PAR by gender and age group will ultimately increase the efficiency of clinical and public health interventions.

Risk and protective factors for heat-related events among older adults of Southern Quebec (Canada): The NuAge study.

OBJECTIVES: Extreme heat is known to increase heat-related health outcomes (HRHO). Incidence and predictors of HRHO were examined among older adults living in Quebec (Canada). METHOD: This prospective five-year study used data from the first follow-up of community-dwelling older adults from the NuAge cohort (2005-2006), located in three health regions of Southern Quebec. Medical, social and environmental factors, identified in Health Canada guidelines (2011), were used to develop the Older Adult Health Vulnerability Index (OAHVI). HRHO, obtained from a medico-administrative database, were defined as events occurring on a hot day (maximal temperature ≥30°C) between 2006 and 2010. Two outcomes were examined: heat-related 1) emergency department presentations (EDPs) and 2) health events (i.e., EDP, hospitalizations or deaths). Multivariate logistic regressions were performed to assess the associations between risk and protective factors, including OAHVI, and both outcomes. RESULTS: EDP and hospitalizations were, respectively, 2.6 (95% CI: 2.0-3.5) and 1.7 (95% CI: 1.1-2.6) times more frequent on hot days compared to normal summer days. Low household income and disability increased risk of heat-related EDP (AOR = 3.20; 95% CI: 1.16-8.81 and AOR = 2.66; 95% CI: 1.15-6.14 respectively) and health events (AOR = 2.84; 95% CI: 1.06-7.64 and AOR = 2.51; 95% CI: 1.13-5.61 respectively). High social participation was a protective factor of heat-related EDP (AOR = 0.05; 95% CI: 0.01-0.20) and health events (AOR = 0.04; 95% CI: 0.01-0.18). Older adults presenting ≥6 OAHVI factors out of 9 were 7-8 times more at risk of heat-related EDP (OR = 7.40; 95% CI: 1.51-36.19) and health events (OR = 7.77; 95% CI: 1.63-37.20) compared to participants having 0-1 factor. CONCLUSION: Social participation, reduced autonomy and low income were predictors of HRHO. The OAHVI, also a strong predictor, should help clinicians identify high-risk elderly patients.

Inflammation and Fat Mass as Determinants of Changes in Physical Capacity and Mobility in Older Adults Displaying A Large Variability in Body Composition: The NuAge Study.

Background/Study context: Determining whether C-reactive protein (CRP), blood lipids, total and trunk fat mass (FM), and waist circumference (WC) are associated with changes in physical capacity over 3 years (Δ) in elderly. METHODS: One hundred twenty-two men and women 68-83 years of age participated in a 3-year follow-up study. Physical capacity was measured using five objective tests: (1) Timed Up and Go (TUG), (2) chair stand (CS), (3) normal walking speed (NWS), (4) fast walking speed (FWS), and (5) one-leg stand (LS), along with physical performance score (PPS) at baseline (T1) and 3 years later (T4). Total and trunk FM, WC, blood lipids, and CRP measured at baseline, were considered as potential predictors. RESULTS: At baseline, CRP and total FM were significantly correlated with all physical capacity tests, whereas trunk FM was correlated with CS and LS, and blood lipids only with FSW. No significant correlation was observed for WC. Total and trunk FM measured at baseline were correlated with ΔTUG and ΔPPS, whereas trunk FM and WC measured at baseline were correlated with ΔNWS. CRP and blood lipids, measured at baseline, were not associated with any changes over 3 years. At the end, WC measured at baseline was the strongest independent predictor for all physical capacity measures at baseline (T1), and ΔPPS measured over 3 years could be predicted by baseline WC. CONCLUSION: FM distribution seems more useful to determine physical capacity than inflammation. Interestingly, over a short follow-up of 3 years, WC significantly predicted changes in a composite score of physical activity. More studies are needed to elucidate factors that may influence physical capacity decline over time.

Comparison of Standardization Methods for the Harmonization of Phenotype Data: An Application to Cognitive Measures.

Standardization procedures are commonly used to combine phenotype data that were measured using different instruments, but there is little information on how the choice of standardization method influences pooled estimates and heterogeneity. Heterogeneity is of key importance in meta-analyses of observational studies because it affects the statistical models used and the decision of whether or not it is appropriate to calculate a pooled estimate of effect. Using 2-stage individual participant data analyses, we compared 2 common methods of standardization, T-scores and category-centered scores, to create combinable memory scores using cross-sectional data from 3 Canadian population-based studies (the Canadian Study on Health and Aging (1991-1992), the Canadian Community Health Survey on Healthy Aging (2008-2009), and the Quebec Longitudinal Study on Nutrition and Aging (2004-2005)). A simulation was then conducted to assess the influence of varying the following items across population-based studies: 1) effect size, 2) distribution of confounders, and 3) the relationship between confounders and the outcome. We found that pooled estimates based on the unadjusted category-centered scores tended to be larger than those based on the T-scores, although the differences were negligible when adjusted scores were used, and that most individual participant data meta-analyses identified significant heterogeneity. The results of the simulation suggested that in terms of heterogeneity, the method of standardization played a smaller role than did different effect sizes across populations and differential confounding of the outcome measure across studies. Although there was general consistency between the 2 types of standardization methods, the simulations identified a number of sources of heterogeneity, some of which are not the usual sources considered by researchers.

Knowledge Translation Strategy to Reduce the Use of Potentially Inappropriate Medications in Hospitalized Elderly Adults.

OBJECTIVES: To evaluate the effect of a knowledge translation (KT) strategy to reduce potentially inappropriate medication (PIM) use in hospitalized elderly adults. DESIGN: Segmented regression analysis of an interrupted time series. SETTING: Teaching hospital. PARTICIPANTS: Individuals aged 75 and older discharged from the hospital in 2013/14 (mean age 83.3, 54.5% female). INTERVENTION: The KT strategy comprises the distribution of educational materials, presentations by geriatricians, pharmacist-physician interventions based on alerts from a computerized alert system, and comprehensive geriatric assessments. MEASUREMENTS: Rate of PIM use (number of patient-days with use of at least one PIM/number of patient-days of hospitalization for individuals aged ≥75). RESULTS: For 8,622 patients with 14,071 admissions, a total of 145,061 patient-days were analyzed. One or more PIMs were prescribed on 28,776 (19.8%) patient-days; a higher rate was found for individuals aged 75 to 84 (24.0%) than for those aged 85 and older (14.4%) (P < .001), and in women (20.8%) than in men (18.6%) (P < .001). The drug classes most frequently accounting for the PIM were gastrointestinal agents (21%), antihistamines (18%), and antidepressants (17%). An absolute decrease of 3.5% (P < .001) of patient-days with at least one PIM was observed immediately after the intervention. CONCLUSION: A KT strategy resulted in decreased use of PIM in elderly adults in the hospital. Additional interventions will be implemented to maintain or further reduce PIM use.

Relation between mealtime distribution of protein intake and lean mass loss in free-living older adults of the NuAge study.

BACKGROUND: Studies have shown that an even protein intake distribution across meals increased 24-h muscle protein synthesis in young adults compared with a skewed intake. Whether this short-term result translates into long-term preservation of lean mass (LM) in older adults remains unknown. OBJECTIVE: The aim was to examine the extent to which protein quantity and distribution are associated with LM and appendicular LM (aLM), and their 2-y decline, in community-dwelling older adults. DESIGN: Baseline and 2-y follow-up data from 351 men and 361 women (aged 67-84 y) in the NuAge study (Quebec Longitudinal Study on Nutrition as a Determinant of Successful Aging) with available body-composition data (by dual-energy X-ray absorptiometry) were used. Food intake was assessed with the use of three 24-h food recalls collected at baseline and 3 collected at the 2-y follow-up. Protein distribution across meals was calculated as the CV of protein ingested per meal, with lower values reflecting evenness of protein intake. Linear mixed-model analysis was performed to examine changes in LM and aLM across time, by sex, as conditioned by the quantity and distribution of protein intake, adjusted for potential covariates. RESULTS: Over 2 y, LM declined in both men (-2.5% ± 4.0%) and women (-2.0% ± 3.4%) (P < 0.05), whereas aLM loss was not significant (men: -1.5% ± 4.8%; women: -1.2% ± 5.3%; P > 0.05). The decline in LM was not independently affected by the quantity and distribution of protein intake. Yet men and women with evenly distributed protein intakes and men with high protein intakes showed higher LM or aLM throughout the entire follow-up period, even after potential confounders were controlled for (P < 0.05). CONCLUSIONS: Our results suggest that greater protein intakes and a more even distribution across meals are modifiable factors associated with higher muscle mass in older adults but not with losses over 2 y. Interventional studies should determine longer-term effects on preserving LM with aging.

Relationships between physical activity across lifetime and health outcomes in older adults: Results from the NuAge cohort.

This study aims to (1) describe participation in four physical activity (PA) domains across life and (2) examine the influence of PA during adolescence, early, middle, and later adulthood on health variables at older age. This observational study was conducted in 1378 generally healthy older adults (age 67-84 at baseline in 2003-2005; 52% women) in Quebec, Canada. Using a modified version of the interviewer-administered Lifetime Total Physical Activity Questionnaire (LTPAQ) and life events calendar to facilitate the recall, participants reported the frequency, duration, and intensity of occupational (OPA), commuting (CPA), household (HPA), and leisure time (LTPA) they participated in at age 15, 25, 45, and 65, and at the first follow-up (age 68-85 in 2005-2006). Fat mass, lean body mass, body mass index, waist-to-hip ratio, fasting glucose, systolic and diastolic blood pressures, self-reported chronic diseases, and socio-demographic data were assessed at baseline. Changes in PA over time differed by sex in each domain. However, there was a general decline in all PA domains in both sexes after age 65. In multiple regression analyses, LTPA at first-follow-up was associated with more favourable waist-to-hip ratio in both sexes, fat mass in women and fat mass percentage in men, whereas CPA, OPA, and HPA across life were not consistently associated with health variables. Older adults' LTPA at first follow-up was related to health variables, but PA recalled during adolescence, early adulthood, and mid-life was not. Results support the idea that current PA is positively related to better health outcomes.