UC San Diego

Background: Sedentary behavior has been associated with increased risk of diabetes and cardiovascular disease (CVD), but most studies rely on self-reported sedentary behavior measures, which are inaccurate. Furthermore, there is little evidence on these associations in adults over 80 years. The way sedentary time is accumulated (i.e., sedentary accumulation patterns) has shown acute effects on glucose and lipid metabolism in laboratory studies, but little evidence exists on how accumulation patterns relate to diabetes risk factors among adults and there are no previous studies on how accumulation patterns are related to diabetes risk in older adults.

Methods: This dissertation is composed of four separate studies. Data for Chapters 2 through 4 were from the Objective Physical Activity and Cardiovascular Health Study (OPACH), a cohort of older women (n=6489; average age = 79±7) that wore ActiGraph GT3X+ accelerometers around their hip for up to 7 days between 2012-2014 and were followed for incident CVD and diabetes through September 30, 2016. Data for Chapter 5 were from the 2011-2012 wave of the Australian Diabetes, Obesity, and Lifestyle Study (AusDiab), a national, population-based cohort (n=739; average age 58±10 years) established to study the distribution and determinants of diabetes.

Results: Chapter 2 results suggest that women in the highest (≥ ~11 hr/day) quartile of sedentary time had higher risk for CVD events (hazard ratio (HR)=1.44; 95% confidence intervals (CI)=1.05-1.98) and coronary heart disease (CHD) events (HR=2.19; 95% CI=1.09-4.40) than women in the lowest quartile of sedentary time (≤ ~9 hr/day) with a linear dose-response relation (P-linear<.05; P-nonlinear>.05 | all). Results from Chapters 3 and 4 reveal that women in the highest quartile of sedentary time had higher odds of prevalent diabetes (odds ratio (OR) = 1.96; CI, 1.59-2.42) than women in the lowest quartile, after adjustment for covariates. Those that accumulated sedentary time with the most prolonged accumulation patterns (i.e., many long bouts of sedentary time with few short bouts and few interruptions) had higher odds of prevalent diabetes than women with the most interrupted patterns, though the ORs were weaker than for total sedentary time. Due to non-proportional hazards by family history of diabetes (FH+/-), models of diabetes incidence were stratified by FH. FH+ women in the highest quartile of sedentary time had higher risk for diabetes (HR=2.38; 95% CI=1.05-1.98) than women in the lowest quartile. Women with the most prolonged sedentary accumulation patterns had higher risk for new-onset diabetes (HR=2.32; 95% CI=1.15-4.71) than women with the most interrupted patterns. No significant associations were observed for FH- women. Results from Chapter 5 indicate that accumulation patterns of frequently interrupted sitting (compared to patterns with relatively more prolonged sitting) were significantly beneficially associated with BMI, waist circumference, HDL cholesterol, triglycerides, 2-hour post-load plasma glucose levels (PLG), and fasting plasma glucose levels. Significant interactions (p<0.05) showed that associations of sitting time with HDL, triglycerides and PLG became more deleterious with longer usual bout durations indicating a joint relationship between sedentary behavior and sedentary accumulation patterns.

Conclusion: High levels of sedentary time were associated with increased risk for CVD and diabetes in older women. Furthermore, prolonged sedentary accumulation patterns were associated with increased diabetes risk in older FH+ women and were deleteriously associated with several cardio-metabolic biomarkers in Australian adults. Accumulation patterns interacting with total sitting time in relations with key diabetes-related biomarkers (PLG, HDL, and triglycerides) provides further evidence that the way in which sedentary time is accumulated may be a relevant factor in diabetes etiology, in addition to the total amount of time sedentary.