2015 Volume 27 Issue 1 Pages 2-39
This study uses program evaluation approaches to analyze the impact of the specific health checkups and specific health guidance that began in FY2008 in Japan. The specific health checkups target individuals aged 40 to 74 to prevent lifestyle-related diseases for those at risk through focusing on metabolic syndrome. The specific health guidance provides the following two types of support, based on the checkup results: (i) one-off practical advice and motivation for improving lifestyle, and (ii) consecutive practical advice and support. We use micro data on the Japanese National Health Insurance members in Fukui Prefecture who had specific health checkups from FY2008 to FY2010. This data set was made available by collaboration between Fukui Prefecture and the Institute of Gerontology, at the University of Tokyo.
We estimated the effect of the specific health guidance on people whose abdominal circumference (AC) and body mass index (BMI) measurements exceed their thresholds. Methodologically, we adopted fixed effect model (FE), matching estimates, and difference-in-differences (DD) estimates to distinguish between the true effect of the specific health guidance and the so-called regression to the mean effect. Our major findings are as follows;
1) Although the results of both FE and Maharanobis matching estimates show no effect of eligibility for the guidance on AC, propensity score matching estimates showed that the effect is only around a 0.3-0.4 decline over a year.
2) We found a small effect of the guidance on BMI. The estimate results of Maharanobis matching, propensity score matching, and FE estimates show that eligibility for the guidance decreases their BMI by only around 0.4%, around 0.5%, and around 0.5% over a year, respectively.
3) FE estimates for the members who participated in specific health guidance show that the guidance decreased its participants' AC by around 0.4% and BMI by around 0.6% over a year, compared to non-participants.
4) The empirical results of FE and the DD estimates show that the effect of continuous support on their AC is no higher than that of one-off support.
5) We also found from FE estimates that the effect of continuous support on their AC is no higher than that of one-off support, while DD estimates show that eligibility for the continuous support has approximately 0.5 percent (annual rate) larger significant effects to decrease their BMI than that of the one-off support.
