Purpose: Summary measures of sedentary behavior (SB), e.g., daily sedentary minutes, are linked to negative health  in older adults. By leveraging minute-level accelerometry, can we derive clusters of diurnal timing of SB that further elucidate SB-health associations?

Methods: 6204 OPACH participants (79±7 years) wore accelerometers for 4-7 days, yielding 41,356 days with >600 minutes/day of data. Participants were followed for upto 6 years with annual physical functioning (PF) assessments.  We conducted novel 2-phase clustering: phase I longitudinal k-means clustering classified 41,356 person-days into 4 diurnal SB patterns; phase II hierarchical clustering classified participants using phase I patterns. We used mixed models to evaluate SB cluster and longitudinal PF, adjusting for demographics, BMI, moderate-vigorous-physical activity (MVPA), comorbidities, and sedentary minutes. Effect modification by MVPA was evaluated via likelihood ratio tests.

Results: At baseline, sample means (SD) were 28.1 (5.7) kg/m² for BMI, 597 (103) minutes/day for SB, 50 (34) minutes/day for MVPA, and 69 (26) score for PF. Mean PF decline was 2.2 units/year (SE=0.07). Four diurnal SB clusters were identified: high-SB-throughout-the-day (c1, n=2239); moderate-SB-with-lower-morning-SB (c2, n= 1536); moderate SB-with-higher-morning-SB (c3, n= 1137); low-SB-throughout-the-day (c4, n=1292).  C1 was less healthy (high BMI, low MVPA, more comorbidities) than c2-c4. SB and PF associations differed by baseline MVPA (p < 0.001); across all clusters, lower vs higher MVPA had worse baseline PF (c1-c4 means 70  -75 low MVPA vs 78.5 -81.6 high MVPA) and higher decline (PF slopes 2.2  – 3 low MVPA vs 1.6  to 2.1 high MVPA).  Paradoxically, the c2-low-MVPA group, had higher baseline PF, yet the steepest PF decline  (p < 0.04, vs c1,c4). These effects were mitigated by higher baseline MVPA.

Conclusions: In older women, diurnal SB pattern was associated with differential rates of PF decline.  Diurnal timing of SB, independent of total sedentary time, may offer new insights into SB-associated healthy aging.

Purpose:

No studies have examined associations between change in diet quality over time and healthcare costs. Our aims were to investigate the relationships between baseline diet quality, and change in diet quality over time, with 15-year cumulative healthcare service use and costs.

Methods:

The Australian Recommended Food Score (ARFS) was measured in 2001 (n=9377) and 2013 (n=7415 both time-points) for the 1946-51 cohort of the nationally-representative Australian Longitudinal Study on Women’s Health. Fifteen-year (2001-2015) data on healthcare services/costs were obtained from Medicare (Australia’s universal healthcare cover).

Part i: To investigate relationships between baseline diet quality and 15-year cumulative services/costs were reported by baseline ARFS quintiles within BMI category.

Part ii: To investigate the relationship between change in diet quality and healthcare costs, by categories of diet quality ‘worsened’ (ARFS decrease ≤-4 points), ‘remained stable’ (-3≤change in ARFS≤3 points) or ‘improved’ (ARFS increase ≥4 points). Analyses were conducted using linear regression modelling, adjusting for geographical area and socioeconomic status.

Results/findings:

Part i: Healthy weight women with the highest diet quality at baseline had fewer healthcare services over 15 years compared to women with the lowest diet quality, although overweight/obese women incurred higher gap (out-of-pocket) costs (p<0.05). Among all mid-aged women, having a greater variety of vegetables was inversely associated with total services and costs over 15 years. For every 1-point increase in ARFS vegetable subscale (equivalent to one extra type of vegetable per week), healthy weight women had 3.4 (95%CI 1.2-5.6) fewer services and incurred $267 ($103-430) fewer charges, while overweight/obese women had 3.3 (1.0-5.5) fewer services and incurred $241 ($80-403) less charges (p<0.05). Higher ARFS dairy and fruit subscales were associated with higher services and costs (p<0.05).

Part ii: Women whose diet quality remained stable or improved over 12 years had fewer services compared to those whose diet quality worsened (p<0.05).

Conclusions:

Higher diet quality is associated with fewer Medicare healthcare services among healthy weight mid-aged women. Higher vegetable variety is associated with fewer healthcare services and costs, although results are reversed for fruits and dairy. Women who maintained or improved their diet quality over time had fewer healthcare services.

Purpose: Reducing sitting time may confer important health benefits for inactive people with chronic disease.  The global prevalence of diabetes in adults has almost doubled from 1980 to 2014. There is scant evidence investigating ways to decrease sitting time in people with type 2 diabetes (T2DM). This study aimed to evaluate the feasibility and preliminary effectiveness of an incremental goal-setting intervention on sitting time and blood glucose levels (BGLs) in people with T2DM.

Methods: This study used a pre-post design. Non-working adults (≥60 years) with non-insulin dependent T2DM were recruited. Participants attended a one-hour face-to-face session and were: guided through a review of their sitting time; provided with normative feedback on sitting time; and assisted with setting goals to reduce sitting time and bouts of prolonged sitting. Participants chose six goals and integrated one per week incrementally for six weeks. Sitting time and bouts (activPAL3c) and BGLs (Continuous Glucose Monitoring System) were measured for 7-days pre- and post-intervention. During these periods, a 24-hour time recall instrument was administered by computer-assisted telephone interview. Paired t-tests (2-tailed) were completed (α=0.05).

Results: Of the 28 participants enrolled (69.8±5.4 years, 57% female, BMI 31.3±5.2), 22 have completed, two withdrawn, and four are pending. Data collection will cease March 2020. Post-intervention (n=22), there was no significant change in total daily sitting time (4min/d), and small non-significant increases in standing (18min/d) and stepping (6min/d) time. The time participants spent in short sitting bouts (<30min) increased significantly post-intervention by 36min/d (p=0.03), and the time spent in prolonged sitting bouts (≥30min) decreased by 30min/d, however this was not significant. There was no significant change in average daily BGLs from pre- (7.65mmol/L) to post-intervention (7.88mmol/L). Participant retention was high (93%), all participants reported achieving their goals, were highly satisfied (mean 9.1, 10 was ‘extremely satisfied’) and would recommend the program to others (9.3, 10 ‘definitely recommend’).

Conclusions: The program is feasible to be implemented with older adults with T2DM. Only small changes in sitting time behaviour were found, with participants spending more time in shorter bouts of sitting but no change to overall daily sitting time.

Purpose: Fundamental to supporting healthy ageing are policies to support global population changes; at their core are food and nutrition requirements at population and individual levels. Accurate estimates of energy requirements are crucial to guide nutritional recommendations for healthy ageing. We aimed to determine the optimal method of estimating total energy expenditure (TEE) in adults aged ≥ 65 years through 1) determining which resting metabolic rate (RMR) equations best reflect measured RMR and 2) using these equations to calculate TEE and determining the agreement with TEE measured by doubly labelled water (DLW).

Methods:  An international database of 31 studies of healthy, ambulatory adults aged ≥ 65 years, (988 participant level RMR data and 1488 participant level TEE data) was developed to enable comparison with 17 commonly used predictive equations. Bland-Altman plots assessed the agreement between measured RMR and TEE with predicted RMR and TEE in adults aged ≥ 65 years, 65-79 years and ≥ 80 years. Proportional bias was assessed using linear regression.

Results/findings: The Ikeda, Livingston and Mifflin equations agreed most closely with measured RMR for the population aged ≥ 65 years, as well as the sub-group analyses of 65-79 years and ≥ 80 years. In adults aged ≥ 65 years, the Ikeda and Livingston equations overestimated TEE by a mean ± SD of 175 ± 1362 kJ/d and 86 ± 1344 kJ/d respectively. The Mifflin equation underestimated TEE by 24 ± 1401 kJ/d. Commonly cited equations (Schofield and Harris–Benedict) showed proportional bias and overestimated RMR. Proportional bias was seen with increasing energy expenditure.

Conclusions: The Ikeda, Livingston and Mifflin equations are recommended for the estimates of energy requirements for older adults aged ≥ 65 years, including both 65-79 years and ≥80 years. Predictive energy equations currently used in practice rely on chronological rather than biological age, itself a potential limitation on accuracy. Future research should focus on the development of equations to meet the requirements of older adults with considerations given to body composition and functional state.

Purpose: Physical activity is an essential component of health. Objective measurement of physical activity trajectories is sparse in older adults. We aimed to analyze the association of objectively measured trajectories of physical activity during older age on mortality in community-dwelling older adults.

Methods: Walking duration as a measure for physical activity was recorded from 1,406 participants (≥65 years, 56.2% men) of the ActiFE-Ulm study in 2009/10 and followed-up two times (2012/13 and 2017/18), each measurement period was over one week (16,761 measurement days in total) using a thigh-worn uniaxial accelerometer (activPAL; PAL Technologies, Glasgow, Scotland). Mortality was assessed last in February 2019.

A latent class joint survival model was used to identify trajectory classes over age and estimate its association with mortality. The longitudinal sub-model included attained age as fixed, random and mixture effects, weekday and daily maximum temperature were included as additional fixed effects (as weather conditions and Sundays have previously been found to affect activity levels). The survival sub-model was adjusted for sex. Both sub-models were joined by attained age as the shared component.

Results/findings: The analysis identified three walking duration trajectory classes over age. A slowly declining class one (n=1,224, 87.1%), a moderately declining class two (n=158, 11.2%), and a rapidly declining class three (n=24, 1.7%). There were 390 deaths during follow-up. Median life expectancy for individuals following trajectory class one was estimated as 92.4 years. Individuals following trajectory class two showed a hazard ratio (HR) of 4.96 (95%-confidence interval (CI): 3.37; 7.32), and those following class three a HR of 68.7 (95%-CI: 29.0; 162.6) compared to class one; equivalent to 7.2 and 18.6 years shorter remaining median life expectancy.

Conclusions: In summary, trajectories of walking duration were clearly associated with overall mortality in community-dwelling older adults. A rapidly declining walking duration trajectory was associated with the highest risk of subsequent mortality.

Purpose: Immobility is major contributor to poor outcomes for older people during hospitalisation with an acute medical illness. As there is no specific mobility guidance for this population, this study aimed to generate draft physical activity and sedentary behaviour recommendations for older adults’ during hospitalisation for an acute medical illness.

Methods: A 4-Round online Delphi consensus process was conducted. International researchers, multidisciplinary clinicians, academics from national activity guideline development teams, and patients were invited to participate. Responses to Round 1 open-ended questions were collated before participants rated the importance of items using a Likert scale (1-9) over Rounds 2-3. Consensus was defined a priori as: ≥70% of respondents rating an item as “critical” (score ≥7) and ≤15% of respondents rating an item as “not important” (score ≤3). Round 4 provided the opportunity for comments on draft recommendations.

Results: 49 participants from nine countries were invited to each Round; response rates were 94%, 88%, 81% and 71% from Rounds 1-4 respectively. 43 items from Rounds 2 and 3 were incorporated into 29 draft statements under the themes of PA (e.g. “some PA is better than no activity”), SB (e.g. “older adults should aim to minimise long periods of SB while hospitalised”), people (e.g. “when encouraging PA and minimising SB, people should be culturally responsive and mindful of older adults' physical and mental capabilities”) and organisational factors (e.g. “opportunities for PA and minimising SB should be incorporated into the daily care of older adults with a focus on function, independence and activities of daily living”). There was consensus that three of the six WHO guidelines for older adults remain applicable when hospitalised with an acute medial illness, two are “not applicable” and consensus was not reached on activity bout duration.

Conclusions: These world-first consensus-based statements from expert and stakeholder consultation provide the starting point for recommendations to address PA and SB for older adults hospitalised with an acute medical illness. Further consultation and evidence review will enable final recommendations to be developed with examples to improve their specificity and translation to clinical practice.

Purpose: Urbanisation and urban densification are global trends. A recent review found insufficient evidence of associations between dwelling density and either total or within-neighbourhood leisure-time walking. There also little evidence regarding associations between density and walking outside the neighbourhood in comparison to within-neighbourhood leisure-time walking. High density may encourage displacement of within-neighbourhood walking with outside-neighbourhood leisure-time walking. To better understand the influence of neighbourhood dwelling density on leisure-time walking, this study examined the pathways through which dwelling density and related environmental attributes may impact on within- and outside-neighbourhood leisure-time walking in Hong Kong Chinese older adults.

Methods: Data are from the Active Lifestyle and the Environment study in Chinese Senior epidemiological study on environmental correlates of 65+ year old Hong Kong community dwellers’ walking and mental health. Participants (n=909) were recruited from administrative units stratified by area-level income and walkability. Within and outside-neighbourhood leisure-time walking, socio-demographics and health status were measured using validated, interviewer-administered questionnaires. Environmental attributes were assessed using Geographic Information Systems. Generalised Additive Mixed Models were used to estimate associations.

Results: Total effects indicated that neighbourhood dwelling density was positively associated with frequency and weekly minutes of outside-neighbourhood leisure-time walking. The former association was mediated by neighbourhood intersection density and suppressed by recreation facilities density via its effect on frequency of within-neighbourhood leisure-time walking. The association of dwelling density with weekly minutes of outside-neighbourhood leisure-time walking was mediated and suppressed by the pathways affecting outside-neighbourhood frequency of leisure-time walking.  It was also mediated by transportation density and car ownership and suppressed by ill-health. The effects of dwelling density on within-neighbourhood leisure-time walking were mediated by recreational density and suppressed by street intersection density.  

Conclusions: Multiple antagonistic pathways are responsible for the potential effects of urban densification on leisure-time walking in Hong Kong older adults. Disentangling these mediating and suppressing factors is important for the planning of environmental interventions aimed at creating healthy and age-friendly communities.