Relationship between Intervention Dose and Outcomes in Living Well with Diabetes—A Randomized Trial of a Telephone-Delivered Lifestyle-Based Weight Loss Intervention

Purpose. To examine associations of intervention dose with weight, physical activity, glycemic control, and diet outcomes in a randomized trial of a telephone counseling intervention. Design. Study design was a secondary analysis of intervention group. Setting. Study setting was primary care practices in a disadvantaged community in Australia. Subjects. Participants were adult patients with type 2 diabetes (n = 151). Intervention. Up to 27 telephone counseling calls were made during 18 months. Measures. Intervention dose was assessed as the number of calls completed (in tertile categories). Primary outcomes were weight and HbA1c, and moderate to vigorous intensity physical activity via accelerometer. Secondary outcomes were dietary energy intake and diet quality via a validated questionnaire. Analysis. Analyses employed were bivariate associations of call completion with sociodemographics, and confounder-adjusted linear mixed models for associations of call completion with outcomes (multiple imputation of missing data). Results. Only previous diagnosis of depression/anxiety had a statistically significant (p= .008) association with call completion. Call completion was significantly associated with weight loss (p < .001) but not the other outcomes (p > .05). Relative to low call completion, mean weight loss (as a percentage of baseline weight) was greater in the high–call completion group by −3.3% (95% confidence interval, −5.0% to −1.5%). Conclusion. Increased dose of intervention was associated with greater weight loss. More needs to be done to retain patients for the duration of weight loss and behavior change interventions, particularly those with diabetes and comorbid depression, who were the most difficult to engage.