Urinary metal mixtures and longitudinal changes in glucose homeostasis: The Study of Women’s Health Across the Nation (SWAN)

https://doi.org/10.1016/j.envint.2020.106109Get rights and content
Under a Creative Commons license
open access

Highlights

  • Urinary zinc was positively and molybdenum was inversely associated with HOMA-IR at baseline.

  • Urinary zinc was inversely associated with HOMA-β at baseline.

  • Urinary arsenic was associated with a faster rate of decline in HOMA-β.

  • Metal mixtures may play a role in insulin resistance and β-cell dysfunction.

Abstract

Background

Epidemiologic studies on associations between metals and insulin resistance and β-cell dysfunction have been cross-sectional and focused on individual metals.

Objective

We assessed the association of exposure to metal mixtures, based on assessment of 15 urinary metals, with both baseline levels and longitudinal changes in homeostatic model assessments for insulin resistance (HOMA-IR) and β-cell function (HOMA-β).

Methods

We examined 1262 women, aged 45–56 years at baseline (1999–2000), who were followed through 2015–2016, from the Study of Women’s Health Across the Nation. Urinary concentrations of 15 metals (arsenic, barium, cadmium, cobalt, cesium, copper, mercury, manganese, molybdenum, nickel, lead, antimony, tin, thallium, and zinc) were determined at baseline. HOMA-IR and HOMA-β were repeatedly measured over 16 years of follow-up. A two-stage modeling was used to account for correlations in dependent and independent variables: In stage-1, linear mixed effects models were used to estimate the participant-specific baseline HOMA levels from random intercepts and participant-specific rates of changes from random slopes. In stage-2, adaptive elastic-net (AENET) models were fit to identify components of metal mixtures associated with participant-specific baseline levels and rates of changes in HOMA-IR and HOMA-β, respectively. An environmental risk score (ERS) was used to integrate metal mixture effects from AENET results.

Results

In multivariable adjusted AENET models, urinary zinc was associated with higher HOMA-IR at baseline, whereas molybdenum was associated with lower HOMA-IR at baseline. The estimated changes in baseline HOMA-IR for one standard deviation increase in log-transformed urinary metal concentrations were 5.76% (3.05%, 8.55%) for zinc and −3.25% (−5.45%, −1.00%) for molybdenum, respectively. Urinary zinc was also associated with lower HOMA- β at baseline. Arsenic was associated with a slightly faster rate of decline in HOMA-β in the AENET model evaluating associations between metals and rate of changes. Significant associations of ERS with both HOMA-IR and HOMA-β at baseline were observed. ERS for the rate of changes was not calculated and examined in relation to rates of changes in HOMA-IR and HOMA-β because only a single metal was selected by AENET.

Conclusion

Exposure to metal mixtures may be exerting effects on insulin resistance and β-cell dysfunction, which might be mechanisms by which metal exposures lead to elevated diabetes risks.

Keywords

Metals
Mixtures
Insulin resistance
β-cell dysfunction
Women

Abbreviations

AENET
adaptive elastic-net
BKMR
Bayesian Kernel Machine Regression
BMI
body mass index
ERS
environmental risk score
ENET
elastic-net
FFQ
food frequency questionnaire
HOMA-IR
homeostatic model assessments for insulin resistance
HOMA-β
homeostatic model assessments for β-cell function
ICP-MS
inductively coupled plasma-mass spectrometry
LASSO
least absolute shrinkage and selection operator
LOD
limit of detection
SWAN
Study of Women’s Health Across the Nation
SWAN-MPS
Study of Women’s Health Across the Nation Multi-Pollutant Substudy
T2DM
type 2 diabetes mellitus

Cited by (0)