Background: To examine the impact of lifestyle intervention by telephone (TSC) and face-to-face (FFA) according to the stages of lifestyle change (SLC) on the drop-out from regular medical care of diabetes

Methods: Secondary analysis was performed on data from the Japan Diabetes Outcome Intervention Trial-2 large trial (J-DOIT2-LT), which was a 1-year, prospective, cluster randomized, intervention trial. 11 district medical associations were divided into two subregions (clusters), in which patients were registered, and then were randomized to an intervention group (TSC or FFA) or control group (CG). Lifestyle advisors reminded patients of medical visits to the primary care physician (PCP) and gave lifestyle advice by TSC or FFA. The CG received usual medical care. The outcome of this analysis was the drop-out rate from regular medical care. We estimated the HR and 95% CI of either TSC or FFA relative to CG according to the SLC (stage 1/2/3/4/5: pre-contemplation/contemplation/preparation/action/maintenance).

Results: A total of 2,071 participants (mean age: 57 ± 6 years; 38% female) were studied. 1170, 616, and 285 patients were divided into CG, TSC, and FFA. In the stage 1-3 of diet, TSC (HR:0.47; 95% CI:0.27-0.82) and FFA (HR:0.20; 95% CI:0.07-0.55) were significantly associated with a low risk of the drop-out after adjusting for covariates. The similar association was observed in the stage 1-3 of exercise. In contrast, neither TSC nor FFA reduced the risk for drop-out in the stage 4-5 of diet. TSC (HR:0.12; 95% CI:0.02-0.95) but not FFA (HR:0.37; 95% CI:0.09-1.62) was associated with a low risk for the drop-out in the stage 4-5 of exercise.

Conclusions: FFA could be more effective than TSC to reduce the risk for drop-out from regular medical care of diabetes at the stage 1-3. In contrast, TSC may be enough to reduce the risk for the drop-out in the stage 4-5.


R. Bouchi: None. M. Noda: None. Y. Hayashino: None. K. Yamazaki: None. H. Suzuki: None. A. Goto: None. T.A. Furukawa: None. K. Izumi: None. M. Kobayashi: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at