The research, published in BMC Geriatrics, drew its participants from the Neyshabur Longitudinal Study on Ageing, an ongoing population-based project designed to track health and wellbeing in a large cohort of older Iranians. From this framework, the investigators analyzed a representative sample of 647 adults aged 60 and older who had been diagnosed with type 2 diabetes. The average participant was 67.5 years old, and women made up just over half of the group at 55.8 percent. By focusing specifically on older adults with diabetes, the team targeted a population in which sleep disturbances are known to be especially common and particularly damaging to quality of life.

Methodologically, the study relied on two well-established and validated instruments, each with a long track record in geriatric research. Physical activity was measured using the Persian version of the Physical Activity Scale for the Elderly, known as PASE, which captures not only leisure-time exercise but also the everyday movements that dominate life in older age: household chores, caregiving, walking, and occupational activity. Sleep quality was assessed with the Persian version of the Pittsburgh Sleep Quality Index, or PSQI, a questionnaire that evaluates multiple dimensions of sleep, including duration, latency, disturbances, efficiency, and daytime dysfunction. Higher PSQI scores indicate poorer sleep, which is a crucial detail for interpreting the study’s statistical results.

The numbers themselves tell an interesting story about this population. The mean PASE score was 74.4, with a striking standard deviation of 70.8, revealing enormous variability in how active these older adults were; some participants were nearly sedentary while others maintained substantial daily movement. The mean PSQI score was 6.9 with a standard deviation of 4.1, a value that sits near the commonly used threshold of 5 above which sleep quality is generally considered poor. In other words, the average participant in this cohort was already teetering on the edge of clinically meaningful sleep problems, underscoring why identifying modifiable factors like physical activity matters so much for this group.

To isolate the independent contribution of physical activity, the researchers employed a multiple linear regression framework, progressively adjusting for potential confounders across several models. The fully adjusted analysis, designated Model 4, incorporated age, sex, education, marital status, income, and current job status. Because regression estimates in smaller or skewed samples can be sensitive to individual data points, the team also applied bootstrap regression with 1,000 replications, a resampling technique that provides more robust confidence intervals by repeatedly re-estimating the association across thousands of simulated versions of the dataset. This statistical rigor strengthens confidence that the reported effect is not an artifact of a few unusual participants.

The headline finding emerged clearly from this fully adjusted model: physical activity and sleep quality were significantly and inversely associated, with a beta coefficient of -0.0057 and a 95 percent confidence interval spanning -0.0100 to -0.0014, and a p-value of 0.009. Because lower PSQI scores mean better sleep, the negative coefficient translates directly into a positive message: higher physical activity was associated with better sleep quality. The authors themselves are careful to characterize the magnitude of this association as modest, and the confidence interval, while excluding zero, indicates an effect that is statistically reliable but small at the level of the individual score. In practical terms, the association suggests that meaningful shifts in daily activity could nudge sleep quality in the right direction, even if no single change will transform a poor sleeper into a perfect one.

Equally instructive are the secondary findings about who slept worst in this cohort. Female sex was significantly associated with poorer sleep quality, carrying a beta of 1.85 with a 95 percent confidence interval of 1.20 to 2.49 and a p-value below 0.001. Financial hardship also left a measurable mark: participants in the lowest income category, labeled income status 3 in the analysis, showed a beta of 2.13 with a confidence interval of 1.13 to 3.13 and a p-value below 0.001. These results align with a broader literature showing that sleep problems in later life are not distributed evenly, but cluster among women and among those facing economic strain, possibly through channels such as stress, caregiving burdens, less comfortable sleeping environments, or untreated health complaints.

What might explain a biological link between movement and sleep in people with type 2 diabetes? Several mechanisms are plausible, though the study design cannot confirm any of them. Physical activity is known to improve insulin sensitivity and glycemic control, and hyperglycemia itself is associated with nocturnal awakenings, polyuria, and restless legs-like symptoms, so better metabolic regulation could plausibly translate into fewer sleep disruptions. Activity also increases exposure to daylight, raises body temperature in ways that reinforce circadian amplitude, promotes the buildup of sleep-promoting adenosine in the brain, and reduces symptoms of depression and anxiety that frequently sabotage sleep in chronic illness. Each of these pathways offers a testable hypothesis for future longitudinal work.

The authors are appropriately cautious about the limits of their design. Because the study is cross-sectional, capturing a single snapshot in time, it cannot establish causality. People who sleep poorly may simply move less because they are exhausted, rather than inactivity causing poor sleep; reverse causation remains a live possibility. Residual confounding is also possible, since no adjustment set can capture every relevant variable such as medication use, diabetes duration, comorbidities, or body mass index. The modest effect size further counsels against overstating the clinical implications. Still, the use of a representative sample, validated instruments, and bootstrap-robust regression gives the finding a solid evidentiary footing within its scope.

For clinicians and public health planners, the study offers a low-cost, low-risk lever worth pulling. Encouraging older adults with type 2 diabetes to accumulate more daily activity, whether through structured exercise, walking programs, or simply more household and occupational movement, is already recommended for metabolic and cardiovascular benefits, and this research suggests sleep health may be an additional dividend. The finding that women and those with financial difficulties carry a heavier sleep burden also points toward the need for targeted screening in these subgroups. As the population ages and diabetes prevalence climbs worldwide, understanding how everyday behavior shapes one of the most fundamental pillars of health, sleep, becomes not an academic curiosity but a practical necessity, and this Iranian cohort study provides a carefully measured step in that direction.

Subject of Research: The association between physical activity and sleep quality in older adults with type 2 diabetes

Article Title: Higher physical activity is independently associated with better sleep quality in older Iranian adults with type 2 diabetes

Article References: Darrudi, A., Ameli, E., Moghadam, S. G., & Abbasi, A. (2026). Higher physical activity is independently associated with better sleep quality in older Iranian adults with type 2 diabetes. BMC Geriatrics. https://doi.org/10.1186/s12877-026-08364-z

Image Credits: AI Generated

DOI: 10.1186/s12877-026-08364-z

Keywords: physical activity, sleep quality, type 2 diabetes, older adults, PSQI, PASE, cross-sectional study, Iran, geriatrics, regression analysis, sleep disorders, aging