BACKGROUND: There is a limited amount of evidence concerning the association between seasonal variation and the level of physical activity in individuals with chronic disease. This longitudinal observational study aimed to explore seasonal, monthly, and holiday variations in self-monitored step counts over two years among adults with prediabetes and type 2 diabetes in Sweden.

METHODS: Participants were recruited at primary care centers from 2013 to 2018 to take part in a physical activity intervention. Inclusion criteria included the following: an age of 40-80 years, having prediabetes or type 2 diabetes (≥1 year), and the ability to communicate in Swedish. Individuals with recent myocardial infarction, impaired renal function, diabetic ulcers, a limited capacity for physical activity, insulin onset (<6 months), recurrent or severe hypoglycemia, a high baseline for physical activity, or no internet access were excluded. In total, 120 participants wore step counters and recorded daily steps for two years. Linear mixed models adjusted for sex, age, and body mass index were applied.

RESULTS: Mean (95% CI) step counts were statistically significantly higher in summer (7825 [7762, 7889]) and spring (7805 [7757, 7853]) compared to winter (7098 [7052, 7145]) and fall (7422 [7349, 7490]). Highest step counts were registered in May (7993 [7904, 8071]), followed by June (7968 [7895, 8063]), and the lowest in January (6944 [6856, 7034]) and November (7208 [7113, 7289]). Step counts during holiday periods were statistically significantly lower than non-holiday periods across all seasons.

CONCLUSION: Self-monitored daily steps varied over the seasons in this sample of individuals with prediabetes and type 2 diabetes. Declined physical activity levels in months with unfavorable weather conditions require attention in consultations and research.