Obesity is a major cause of poor health worldwide[19]. Since this problem is made worse by the fact that the success rate of weight loss is low, we constructed a nomogram to predict successful weight loss. Validation of the nomogram demonstrated its good effect discrimination and calibration capabilities. Furthermore, the weight loss success prediction model constructed in this study can be applied before weight loss attempts begin, thereby providing more individualized weight loss guidance for people at different risks and possibly improving the weight loss success rate.
Obesity is associated with an increased risk of type 2 diabetes, cardiovascular disease, certain cancers, and premature death[20]. In addition to adverse health outcomes, obesity also impacts the healthcare system, creating direct costs related to healthcare as well as indirect costs such as lost productivity[21]. Once weight is gained, it is extremely difficult to lose it again, with only 40% of those who attempt losing weight losing ≥ 5% and 20% losing ≥ 10%. However, most people have difficulty maintaining such weight loss, with reported weight regain of 30–50% within 1 year[22]. Failure to maintain weight loss is usually attributed to lack of adherence to the initial weight loss diet, so we sought to predict the success rate of weight loss before it is even begun, thereby possibly helping people with weight loss difficulties strengthen behavioral, dietary, and other interventions to that can improve their chances of success. There are currently few risk models that predict successful weight loss. In this study, however, 12 variables were selected based on LASSO regression and logistic regression and included in the nomogram together with age. The line segment corresponding to each variable is marked with a scale, which represents the possible value range of the variable, and the total score of the corresponding individual scores after all variables are added up is called the Total Points. The length of the line segment reflects the contribution of the factor to the outcome event. In our model age is the most important predictor, followed by LDL, blood glucose at 0 min, HOMA β, TC, hair loss, F-8:00, hirsutism, Ca, blood glucose at 60min, ALT, BMI, and insulin at 0 min.
Our nomogram also showed that hair loss and hirsutism are important factors, and their effects may even exceed those of BMI and fasting insulin. This shows that the more hair you have, the more likely you are to lose weight successfully. Excessive body hair may be due to the body's sensitivity to androgens, indicators of abdominal obesity in men are negatively correlated with testosterone levels. Unlike men, high androgen levels in women are usually a high risk factor for obesity and are closely related to the occurrence of abdominal obesity[23].This study also identified ALT as a prognostic factor using, with lower ALT being more likely to result in successful weight loss. Finally, the calibration curve showed that the nomogram was well calibrated and the AUC (0.807) showed its statistical accuracy. However, accuracy does not necessarily mean it has clinical application. To this end, we also performed DCA, which showed that the nomogram indeed has good clinical utility.
Although the model's predictions were good, three major limitations of this study are that the follow-up period was too short and did not incorporate the effects of regaining weight after weight loss. Another key limitation is the limited number of people who attended the weight management clinic, thus limiting the sample size for this study and resulted in only internal validation but no external validation. The third limitation is that only diet and exercise interventions were studied without other intervention methods such as drugs and surgery. For many people, although they want to lose weight, they are not actively engaged in weight loss due to the perceived difficulty and low probability of success. Therefore, more and more accurate weight loss success prediction models need to be developed to improve people's perceptions of weight loss success. In addition, this study did not include sleep[24], support from friends and family[25], eating habits[26], reasons for weight loss, or other factors that may affect the success of weight loss.
In summary, based on baseline data from a population that a weight management clinic, we developed a nomogram prediction model to predict successful weight loss following diet and exercise intervention. The nomogram is easy to use, highly accurate, and has excellent effect discrimination and calibration capabilities. Therefore, this nomogram may help clinicians make personalized predictions about the probability of weight loss success for each people with obesity and in doing so provide more individualized weight loss intervention that may improve their chances of success.