The Influence of Metabolic Syndrome in Predicting Mortality Risk Among US Adults

Importance of Metabolic Syndrome Even in Adults With Normal Weight

Ting Huai Shi, BA; Binhuan Wang, PhD; Sundar Natarajan, MD, MSc

Disclosures

Prev Chronic Dis. 2020;17(5):E36 

In This Article

Results

The prevalence of MetS was 61.6% in the obese group, 33.2% in the overweight group, and 8.6% in the normal-weight group. We found significant differences in the prevalence of MetS and weight groups for all demographic variables. Groups with MetS were generally older, less educated, and less physically active and had a lower income and a higher prevalence of smoking than their no-MetS counterparts (Table 1). Non-Hispanic white adults and other/multiracial adults had a lower prevalence of overweight and obesity but a higher prevalence of MetS compared with non-Hispanic black or Mexican American adults, whereas the inverse was true for Mexican American, other Hispanic, and non-Hispanic black adults.

According to the product-limit method from the LIFETEST procedure (Figure 1), the normal-weight–MetS group had the highest mortality rate. The log rank test showed significant differences between the normal-weight–MetS group and the overweight–MetS group (P <.001) and between the overweight–MetS group and the obese–MetS group (P <.001). Each no-MetS group had significantly lower mortality than their MetS counterparts, but we found no significant differences among no-MetS groups.

Figure 1.

Unadjusted mortality curve during 150 person-month follow-up for each MetS–BMI category, National Health and Nutrition Examination Survey, 1999–2010, and National Death Index, 2011. Abbreviation: BMI, body mass index; MetS, metabolic syndrome.

Follow-up by group ranged from 29,270 person-months (normal-weight–MetS group) to 221,490 person-months (normal-weight–no-MetS group) (Table 2). The number of deaths was lowest in the obese–no-MetS group (n = 61) and highest in the obese–MetS group (n = 255); the mortality rate per 1,000 person-month was lowest in the obese–no-MetS group (0.52) and highest in the normal-weight–MetS group (2.94). Unadjusted mortality rates showed that the normal-weight–MetS group had the highest mortality per person-month, followed by the overweight–MetS group and the obese–MetS group. Cox regression models that adjusted for age, sex, race/ethnicity, education, poverty-income ratio, smoking status, and physical activity showed significantly higher hazard ratios (HRs) for only the normal-weight–MetS group (HR, 1.70, 95% confidence interval [CI], 1.16–2.51) and obese–MetS group (HR, 1.30; 95% CI, 1.07–1.60) compared with the normal-weight–no-MetS group. The HR was higher in the normal-weight–MetS group than in the obese–MetS group, although the difference was not significant (HR, 1.31; 95% CI, 0.92–1.84). In the test of the interaction between weight groups and MetS, the P value for the interaction term in the full model was .16; however, tests for interactions typically have low power, and the recommended approach in causal inference is to base this on a priori hypotheses rather than the data. When we directly compared the normal-weight–MetS group with other groups, we found an HR of 0.64 (95% CI, 0.46–0.90; P = .01) in the overweight–MetS group and an HR of 0.76 (95% CI, 0.54–1.08; P = .13) in the obese–MetS group.

In analyses of cause-specific mortality (Figure 2), we found that among 985 total deaths, 184 (weighted, 16.2%) were due to cardiovascular disease, 233 (weighted, 25.9%) were due to cancer, and the rest were due to causes that each accounted for less than 7% of total mortality. The adjusted Cox regression model for cardiovascular mortality showed a significant hazard ratio only for the normal-weight–MetS group (HR, 2.12; 95% CI, 1.16–2.51). The model for cancer mortality showed a significant hazard ratio for the overweight–MetS group (HR, 1.86; 95% CI, 1.09–3.19) and the obese–MetS group (HR, 1.91; 95% CI, 1.15–3.17).

Figure 2.

Weight–MetS categories and all-cause and selected cause-specific mortality, National Health and Nutrition Examination Survey, 1999–2010, and National Death Index, 2011. The normal-weight–no-MetS group was used as the reference group. Models were adjusted for age, sex, race/ethnicity, education, poverty-income ratio, smoking history, and physical activity. Error bars indicate 95% confidence intervals. Abbreviations: MetS, metabolic syndrome; NWMS; normal-weight–MetS; OWNMS, overweight–no MetS; OWMS, overweight–MetS; OBNMS, obese–no MetS; OBMS, obese–MetS.

When we compared adults with MetS in each weight group with the no-MetS group (Table 3), we found that only the normal-weight–MetS group had a significant hazard ratio (HR, 1.70; 95% CI, 1.16–2.51). Although the obese–MetS group had a significantly higher HR when compared with the normal-weight–no-MetS group (HR, 1.30; 95% CI, 1.07–1.60), it did not have a significantly higher HR when compared with the obese–no-MetS group (HR, 1.20; 95% CI, 0.85–1.70).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....