Abstract
Lifestyle interventions in type 2 diabetes (DM2) prevention implementation studies can be effective and lasting. Long-term weight loss maintenance enhances the intervention effect through a significant decrease in diabetes incidence over time. Our objective was to identify factors predicting long-term successful weight reduction maintenance achieved during a DM2 prevention program in patients with high DM2 risk in primary health care. Study participants (n = 263), middle-aged, slightly obese with baseline increased DM2 risk (Finnish Diabetes Risk Score (FINDRISC)>14), but no diabetes were invited to receive 11 lifestyle counselling sessions, guided physical activity sessions and motivational support during 10-months. The study participants had three clinical examinations during the study (baseline, one and three years). Stepwise regression analysis was used to determine demographic, clinical, and lifestyle predictors of weight reduction maintenance two years after the discontinuation of the intervention. Out of 105 patients who completed all three examinations (baseline age 56.6 (standard deviation (SD) = 10.7), body mass index 31.1 kg/m2 (SD = 4.9), FINDRISC 18.6 (SD = 3.1)), 73 patients (70%) showed weight loss during the intervention (mean weight loss 4.2 kg, SD = 5.1). The total weight loss achieved in the maintainers (27 of 73 study participants) two years after the intervention had finished was 6.54 kg (4.47 kg+2.0 kg). The non-maintainers, on the other hand, returned to their initial weight at the start of the intervention (+0.21 kg). In multivariable analysis baseline history of increased glucose (odds ratio (OR) = 3.7; 95% confidence interval (CI) 1.0-13.6) and reduction of total fat in diet during follow-up (OR = 4.3; 95% CI 1.5-12.2) were independent predictors of successful weight loss. Further studies exploring predictors of weight loss maintenance in diabetes prevention are needed to help health care providers to redesign interventions and improve long-term outcomes of real life interventions.