Where can obesity management policy make the largest impact? Evaluating sub-populations through a microsimulation approach

Figures & data

Figure 1: Schematic Overview of the Obesity Component of the Microsimulation Model

Figure 2: BMI's four routes of influence on IHD

Figure 3. Subpopulations Analyzed

Figure 4. Per capita savings in direct medical cost from different subpopulations

Table 1. Multivariate analysis on per capita medical savings.

Parameter	Coefficient	Standard error	Significance probability	Relative risk
Intercept	6.924	0.007	<0.0001
Time span^a
10 years	0.619	0.005	<0.0001	1.857
15 years	1.004	0.005	<0.0001	2.729
Obesity class^a
Class II	−0.27	0.005	<0.0001	0.763
Class III	−0.049	0.006	<0.0001	0.952
Age group^a
35–44	0.004	0.007	0.5237	1.004
45–54	−0.195	0.007	<0.0001	0.823
55–64	−0.228	0.008	<0.0001	0.796
65–74	−0.116	0.009	<0.0001	0.890
Glucose levels^a
Pre-diabetes	0.229	0.005	<0.0001	1.257
Diabetes	−0.036	0.007	0.0001	0.965
Hypercholesterolemia	0.033	0.005	<0.0001	1.034
Hypertension	0.323	0.005	<0.0001	1.381
Male	−0.495	0.004	<0.0001	0.610
Weight loss of 12% (vs 5%)	1.022	0.004	<0.0001	2.779

Generalized linear regression with random effect on log transformed outcome variable.

^aReference categories are aged 25–34; obesity class I; time span of 5 years; normal blood glucose levels (HbA1c < 5.7%); normal cholesterol (total cholesterol <240 mg/dl or LDL cholesterol <160 mg/dl) and blood pressure levels (SBP <130 or DBP <80 mmHg); female; and weight loss of 5%.

Figure 5. Length of sustained weight loss on per capita medical savings over 15 years