Heavy Metal Index
Ask the IndexAboutIndex Pro

Associations of metal mixtures with insulin resistance and glucose homeostasis in U.S. adults from the NHANES 2011-2018

Cite this pageVersion history

Overview

Cross-sectional study of 3,110 US adults from NHANES 2011-2018 using generalized linear models, logistic regression, restricted cubic splines, BKMR, and Bayesian WQS (BWQS) to examine associations of blood and urine metals with insulin resistance (HOMA-IR) and glucose-insulin homeostasis. Five plasma metals (cadmium, lead, mercury, manganese, selenium) and three urinary analytes (cobalt, molybdenum, arsenic) were included. The HOMA-IR threshold for insulin resistance was defined as HOMA-IR above 2.5 in adults.

Key numbers

n=3,110 US adults, NHANES 2011-2018. HOMA-IR = (fasting plasma glucose in mmol/L times fasting plasma insulin in uU/mL) divided by 22.5.

Single-metal findings:

  • Manganese: positively correlated with insulin resistance.
  • Cadmium, lead, mercury, arsenic: negatively correlated with HOMA-beta (a marker of beta-cell function).
  • Plasma cadmium and manganese concentrations were significantly lower in males than in females; mercury showed a tendency toward lower concentrations in males (p=0.07, not significant).

Mixture findings:

  • Both BWQS and BKMR models consistently showed that overall metal co-exposure had a positive effect on insulin resistance in the general population.
  • Manganese was the most heavily weighted metal across all subgroups; this association was more pronounced in males and in individuals over 60 years of age.
  • A negative association of metal mixtures with HOMA-beta was observed in BWQS models.
  • BKMR identified possible interactions between insulin resistance and some mixture components in glucose homeostasis.
  • Nonlinear dose-response relationships: urinary molybdenum and HOMA-beta (RCS), urinary cobalt and both fasting plasma glucose and HbA1c (RCS).

Evidence Fitness

Context only. Cross-sectional design; temporality not established. Plasma metals reflect recent exposure; urinary metals (creatinine-adjusted) reflect recent exposure. No food concentration data. The consistent cross-method finding that the overall mixture has a positive effect on insulin resistance, with manganese as the dominant contributor, is evidence for mixture-level metabolic risk beyond what single-metal analyses capture.

Methods (brief)

NHANES 2011-2018 (four cycles). Plasma metals by ICP-DRC-MS; urine metals creatinine-adjusted by same method. HOMA-IR and HOMA-beta from fasting glucose and insulin. Missing covariate data filled by multiple imputation. GLM, logistic regression, RCS, BWQS, BKMR. All metal concentrations log10-transformed prior to analysis.

Implications

Certification: Manganese’s dominant role in the insulin-resistance metal mixture is relevant background for the HMT&C manganese monitoring rationale.

Courses: Useful for demonstrating how mixture methods distinguish dominant contributors from co-exposures and how BKMR can identify interactions among mixture components.

App: No food concentration data.

Microbiome: Potential gut-insulin-axis crosswalk; not primary focus.

Update history

No substantive edit history is available in this build. The full commit record is available in git.

Referenced by