Association between Blood Manganese Levels and Visceral Adipose Tissue in the United States: A Population-Based Study

Abstract

Background: Manganese (Mn) is an essential trace element with a narrow toxic margin for human health. The association between Mn exposure and adverse visceral adipose tissue (VAT) accumulation is unclear. Objective: This study aimed to estimate the associations of blood Mn levels with VAT mass or visceral obesity in the general population in the United States. Method: This cross-sectional study included data of 7297 individuals released by National Health and Nutrition Examination Survey (NHANES). VAT was quantified with dual-energy X-ray absorptiometry, and blood Mn was measured using inductively coupled plasma mass spectrometry. The generalized linear model and generalized additive model (GAM) were applied to estimate the linear and non-linear associations between Mn levels and VAT mass, respectively. Logistic regression was used to estimate the associations between blood Mn levels and the risk of visceral obesity. Results: Fully adjusted generalized linear regression revealed that individuals in the higher quantile of Mn had increased VAT mass compared with those in the lower quantile (β per quantile change = 0.025; 95% CI of 0.017, 0.033; p < 0.001). Positive associations were also observed in males and females (males: β per quantile change = 0.012, 95% CI of 0.002, 0.022 (p = 0.020); female: β per quantile change = 0.036; 95% CI of 0.023, 0.048 (p < 0.001)). The GAM illustrated that the non-linear associations between blood Mn levels and VAT mass were in U-shape patterns (effective degree of freedom >1 in total participants, males, and females). A stratified analysis found significant interactions between Mn and the family income-to-poverty ratio (PIR) in males, with stronger associations in males with a PIR < 1.3 (β = 0.109; 95% CI of 0.048, 0.170). Additional analyses revealed that individuals in the highest quantile of Mn had a 39% higher risk of visceral obesity (OR = 1.39; 95% CI of 1.15−1.69; p < 0.001). Conclusions: Higher blood Mn levels were positively associated with increased VAT mass and visceral obesity risk. The adverse VAT phenotype associated with excessive blood Mn levels should be further investigated.

Keywords: NHANES; epidemiology; manganese; visceral adipose tissue; visceral obesity.

Figure 1

Flow chart of the study.

Figure 2

Non-linear associations between blood Mn levels and VAT mass using generalized additive model (GAM): (A) Dose–response association between blood Mn levels and VAT mass in the total dataset. (B) Dose–response association between blood Mn levels and VAT mass in the male subpopulation. (C) Dose–response association between blood Mn levels and VAT mass in the female subpopulation. Models were adjusted for gender (only in total), age, race, educational level, BMI, marital status, PIR, smoking status, 24 h alcohol consumption, physical activity, 24 h energy intake, and 24 h fat intake.

Figure 3

Stratified analyses on male and female subpopulations: (A) Stratified analysis of Mn and VAT mass in males. (B) Stratified analysis of Mn and VAT mass in females. Models were all adjusted for age, race, educational level, BMI, marital status, PIR, smoking status, 24 h alcohol consumption, physical activity, 24 h energy intake, and 24 h fat intake. The p-value for cross products of ln-transformed Mn levels and related stratified variables was calculated to estimate the potential interaction. *: p value < 0.05; **: p value < 0.01.

Figure 4

Associations between blood Mn levels and visceral obesity risk. (A). Associations between blood Mn levels and visceral obesity in total subjects. Minimally adjusted models were adjusted for gender (only in total), age and race. Fully adjusted models were adjusted for gender, age, race, educational level, BMI, marital status, PIR, smoking status, 24-h alcohol consumption, physical activities, 24-h energy intake, 24-h fat intake. (B). Stratified associations between blood Mn levels and visceral obesity. **: p value < 0.01, ***: p value < 0.001.