Randomized trial of a "talking computer" to improve adults' eating habits

Purpose. To assess efficacy of an intervention delivered by an interactive, computer-controlled telephone system to improve individuals' diets. Design. Randomized controlled trial. Setting. Large multispecialty group practice. Subjects. Two hundred ninety-eight adults who were both sedentary and had suboptimal diet quality. Intervention. Weekly communication for 6 months via a totally automated, computer-based voice system. Among intervention group subjects, the system monitored dietary habits and provided educational feedback, advice, and behavioral counseling. Control group subjects received physical activity promotion counseling. Measures. Daily intake of fruits, vegetables, red and processed meats, whole fat dairy foods, and whole grain foods estimated from a food frequency questionnaire. Results. Mean age 45.9 years, 72% women, 45% white, and 45% African-American, Among participants who completed diet assessments, compared with the control group, the intervention raised fruit intake a mean of 1.1 servings per day (95% confidence interval [CI] .4, 1.7). On a 0 to 100 global diet quality score combining all five food groups, intervention participants improved their mean score 9 (95% CI 4, 13) points more than In the control group. The intervention also raised dietary fiber intake 4.0 g/d (95% CI .1, 7.8) and decreased saturated fat, as a proportion of energy intake, by 1.7% (95% CI -2.7, -.7). Conclusions. This computer-based telecommunications dietary behavior intervention helped improve participants' overall diet.