Randomization was not observed and the intervention was not blinded between participant and researcher due to the nature of the intervention. No post-intervention follow-up was made, which made it infeasible to assess the long-term effect on behavioral change. Measurement bias may occur as the IG were required to self-monitor their BW, WC and BP at home without any standardized methods. Sample may not represent to the community as the recruitment was carried out only among patients at the hospital, which may lead to the potential selection bias. In general, this intervention shows prominence in significantly reducing WC, BW, BP and serum lipid except FPG after 12 weeks. Diet and exercise result were not reported in this study, which makes it impossible to evaluate the behavioral change in diet and physical habits. Long-term effect on behavioral change cannot be assessed properly and the sample size is quite modest and may not be representative to the community. A standardized method...