-

Stable levels of physical activity (PA) have been linked to better health and well-being. Reasons for physical activity fluctuations remain poorly understood. We present data on the psychological predictors of PA decreases based on prospective data from the Healthy Aging in Industrial Environment project. Possible weekly PA decreases, defined as a decrease of the past week’s PA level below the 90% confidence interval of the weekly average for the previous 4 weeks, were evaluated on daily basis based on Fitbit fitness monitor data. Participants with detected decreases received a short survey through a custom smartphone application and provided details about the decrease. Data from 600 participants who have completed 6 months of the 12-month ongoing study have been analyzed. A total number of 4229 surveys was received; of these 3494 (82.6 %) contained at least one answered question. In 2009 (57.7 %) of the surveys, the participants confirmed the PA decrease; in 1264 (36.3 %) surveys reported similar PA levels; in 206 (5.9 %) surveys reported being more active. The most common reported reasons for perceived PA decreases were: busy at work/school (n=680), feeling tired (n=250), not in the mood to exercise (n=183), bad weather (n=160), or other reasons (n=410). Demographic and stable psychological predictors of i) the number of received PA decrease surveys (indicating the total number of detected decreases) and ii) the number of perceived PA decreases (as confirmed in the survey) were analyzed. The number of received surveys was predicted only by social support for exercise (Β=.116, p=.020). The number of perceived decreases was predicted by habit (Β=-.137, p=.021), exercise motivation: identified regulation (Β=.163, p=.032), social support for exercise (Β=.116, p=.021), and PA status (Β=-.116, p=.044). No other variables (self-regulation, exercise identity, barriers self-efficacy, other aspects of motivation, education) directly affected the number of decreases.