How You Move, Sit and Sleep Together May Shape Your Aging Brain

What if the secret to protecting an aging brain is not found in any single habit, but in the way daily behaviors combine into a pattern? A new study published in the Journal of Activity, Sedentary and Sleep Behaviors suggests exactly that. Researchers led by Yuzi Zhang of the University of Texas Health Science Center at Houston analyzed data from 1,240 adults aged 55 and older enrolled in the Cooper Center Longitudinal Study, a long-running preventive medicine program based at the Cooper Clinic in Dallas, Texas. Rather than treating physical activity, sitting time and sleep as separate risk factors, the team used a statistical technique called latent class analysis to discover natural groupings of these behaviors across a full day, and then asked whether those groupings were linked to differences in cognitive performance.

The science behind the approach is worth unpacking. Latent class analysis is a person-centered method that searches a population for hidden subgroups, or classes, defined by the probability of responding in particular ways across several indicators. In this study, the indicators were four self-reported behaviors: leisure-time aerobic activity measured in metabolic equivalent of task minutes per week, muscle-strengthening activity measured in sessions per week, the proportion of waking time spent sitting, and nightly sleep duration. Each behavior was converted into categories aligned with the 2018 Physical Activity Guidelines for Americans and current sleep recommendations. Information criteria such as the Akaike and Bayesian information criteria, along with a measure of classification quality called scaled relative entropy and the interpretability of the resulting groups, guided the researchers to a three-class solution as the best balance of statistical fit and meaning.

The three profiles that emerged paint a vivid portrait of how differently people arrange the same 24 hours. The largest group, containing roughly 62 percent of participants, was labeled very active short sleepers. These individuals had a high probability of exceeding aerobic activity guidelines, a substantial share performed two or more muscle-strengthening sessions per week, but two-thirds slept fewer than seven hours a night and many sat for three-quarters or more of their waking time. The smallest group, about 11 percent of the sample, were the active long sleepers, characterized by the highest probability of exceeding aerobic guidelines, the least sitting time, and sleep of at least seven hours, though most did little muscle-strengthening work. The final group, about 27 percent, were the moderately active short sleepers: the least aerobically active of all, with 85 percent failing to meet aerobic guidelines, high sitting time, short sleep, and yet paradoxically the highest probability of meeting muscle-strengthening recommendations.

Who ended up in which class was far from random. Compared with the very active short sleepers, members of the active long sleeper group were more likely to be aged 65 or older and to be women. The moderately active short sleepers told a more concerning story: they were more likely to be women, to lack a bachelor’s degree, and to carry a heavier burden of cardiometabolic risk. Statistically, they had significantly higher odds of diabetes, high triglycerides, joint or muscle conditions, hearing loss, low high-density lipoprotein cholesterol, and being overweight or obese. The authors point to prior evidence that prolonged sitting can blunt some metabolic benefits of exercise and is associated with unfavorable blood lipid profiles, which makes the health profile of this sedentary, under-active subgroup unsurprising, even though its members lifted weights as often as anyone else in the study.

Cognitive function was assessed with the Montreal Cognitive Assessment, or MoCA, a ten-minute screening tool scored from 0 to 30, with scores below 26 flagging possible mild cognitive impairment. The average score in the sample was 26.85, and about a quarter of participants met the threshold for mild cognitive impairment. When the researchers compared the moderately active short sleepers with the very active short sleepers, the former scored significantly lower on the MoCA in unadjusted models and in models adjusted for sociodemographic factors such as age, sex, race and education. The difference was modest in absolute terms, but it persisted in a sensitivity analysis that used the Bolck, Croon and Hagenaars method, a bias-adjusted statistical approach that corrects for the misclassification errors inherent in assigning people to latent classes.

Here is where the story acquires an important twist. When the researchers further adjusted for health behaviors and clinical indicators, including smoking status, stroke, hypertension, hearing loss and body mass index, the association between class membership and MoCA score was no longer statistically significant. The authors interpret this pattern through the lens of mediation: obesity and hypertension may sit further along the mechanistic pathway between movement behavior and cognitive impairment, acting as intermediaries through which an inactive, sedentary, sleep-short lifestyle exerts its effects on the brain. In other words, the behavioral pattern may damage cognition partly by promoting the very cardiometabolic conditions long known to harm the brain. No significant associations were found between the movement patterns and mild cognitive impairment status, and the active long sleepers did not differ cognitively from the very active short sleepers.

That last finding raises a fascinating question about sleep. Previous research has described a U-shaped relationship between sleep duration and health, with seven to eight hours per night appearing optimal, and both very short and very long sleep linked to higher risks of decline. Yet in this study, the shorter sleep of the largest group did not appear to cost them cognitively relative to the long-sleeping active group. The authors note that sleep durations just under seven hours may fall within acceptable ranges for some individuals under National Sleep Foundation guidance, and they suggest that future work should probe more extreme durations, below five or above nine hours, where harms may become clearer. They also raise the possibility of an inverted U-shaped dose-response between physical activity and cognition, in which very high activity volumes add no benefit beyond high volumes, which could explain why the very active and merely active groups performed similarly.

The study’s strengths lie in its holistic framing and its methodological rigor. By including muscle-strengthening activity as a distinct indicator rather than collapsing all exercise into an aerobic metric, the analysis captured a dimension of fitness often ignored in cognitive aging research. The three-step latent class approach, combined with a bias-adjusted sensitivity analysis, gives the central finding, that the least active, most sedentary and sleep-short group had the poorest global cognition, a degree of robustness that many cross-sectional analyses lack. The Cooper Center sample also came with unusually rich clinical data, including fasting glucose, lipid panels, C-reactive protein and physician-measured blood pressure, allowing careful adjustment for confounding.

Limitations nonetheless temper the conclusions. The cross-sectional design cannot establish whether the behavioral patterns caused the cognitive differences or vice versa, since early cognitive change may itself reduce activity. All movement behaviors were self-reported, leaving room for recall and social desirability bias, and the scaled relative entropy of the chosen model was below 0.6, indicating some overlap between classes. The sample was also strikingly homogeneous: 95 percent White, 71 percent male, more than 80 percent college-educated, and drawn from preventive medicine patients, which limits generalizability to the broader, more diverse aging population.

Even so, the message for the public is compelling and timely. With no cure available for dementia, identifying modifiable risk factors remains one of the most powerful tools in preventive medicine, and this study suggests that interventions should target the whole day rather than a single behavior. A large majority of participants were sedentary for much of their waking hours and slept less than recommended, meaning that even a fitness-conscious population has room to improve. The clearest cognitive risk concentrated among those who moved too little, sat too much and slept too little, a combination that also clustered with obesity, diabetes and hypertension. For anyone hoping to keep an aging brain sharp, the research implies that a brisk morning run cannot fully compensate for a day chained to a chair and nights cut short, and that the most protective lifestyle may be one in which activity, rest and reduced sitting reinforce one another around the clock.

Subject of Research: Combined patterns of physical activity, sedentary behavior and sleep in relation to cognitive function in middle-aged and older adults

Article Title: Associations of physical activity, sedentary behavior, and sleep patterns with cognitive function among middle-aged and older adults

Article References: Zhang, Y., Chen, B., Hébert, E. T., DeFina, L. F., Leonard, D., Barlow, C. E., Pavlovic, A., & Kohl, H. W., III (2025). Associations of physical activity, sedentary behavior, and sleep patterns with cognitive function among middle-aged and older adults. Journal of Activity, Sedentary and Sleep Behaviors, 4(1), Article 9. https://doi.org/10.1186/s44167-025-00079-7

Image Credits: AI Generated

DOI: 10.1186/s44167-025-00079-7

Keywords: physical activity, sedentary behavior, sleep duration, cognitive function, latent class analysis, Montreal Cognitive Assessment, mild cognitive impairment, dementia prevention, aging, Cooper Center Longitudinal Study, cardiometabolic risk, 24-hour movement behaviors