Psychological predictors of physical activity in the diabetes prevention program

To identify the baseline psychological variables before receiving a Diabetes Prevention Program (DPP) lifestyle intervention that predict physical activity levels (PALs) at baseline, 1 year, and end of study (2 to 3 years after randomization).

Of the final 293 DPP lifestyle participants randomized, 274 (94%) completed validated questionnaires at baseline assessing stage of change for PAL, exercise self-efficacy, perceived stress, depression, and anxiety.

Correlations and stepwise multiple regression analyses.

At baseline, this subset was similar to the entire DPP lifestyle cohort: mean age was 52.5 years, 65% were women, and mean PAL was 15.7 metabolic equivalent hours per week. Higher levels of baseline leisure PAL correlated with greater readiness to change PAL (r=0.44, P<0.0001), higher exercise self-efficacy (r=0.18, P=0.002), and lower levels of perceived stress (r=-0.16, P=0.009), depression (r=-0.18, P=0.003), and anxiety (r=-0.14, P=0.03), with similar correlations at 1 year and end of study. In multivariate models, being a man, lower levels of depression, and lower body mass index were independent correlates of higher baseline leisure PAL; being a man, greater baseline exercise self-efficacy, and activity level were independent correlates of greater leisure PAL levels at 1 year and end of study. Greater readiness to change PAL at baseline was also an independent correlate of greater PAL at end of study.

In this representative sample of DPP lifestyle participants, being a man, lower body mass index, greater readiness for change in PAL, higher exercise self-efficacy, and lower perceived stress, depression, and anxiety scores correlated with higher levels of baseline PAL with similar patterns at 1 year and end of study. These findings may help determine which patients are most likely to increase PAL in lifestyle intervention programs.