Associations among diabetes medication use, serum magnesium, and insulin resistance in a cohort of older Puerto Rican adults

Fan Chen; Kelsey M Mangano; Mahdi Garelnabi; Kellee Cardaleen; Katherine L Tucker

doi:10.1016/j.ajcnut.2024.04.005

Associations among diabetes medication use, serum magnesium, and insulin resistance in a cohort of older Puerto Rican adults

Am J Clin Nutr. 2024 Apr 8:S0002-9165(24)00395-2. doi: 10.1016/j.ajcnut.2024.04.005. Online ahead of print.

Authors

Fan Chen¹, Kelsey M Mangano², Mahdi Garelnabi², Kellee Cardaleen², Katherine L Tucker³

Affiliations

¹ Department of Public Health, Zuckerberg College of Health Sciences, University of Massachusetts Lowell, Lowell, MA, United States; Department of Biomedical and Nutritional Sciences, Zuckerberg College of Health Sciences, University of Massachusetts Lowell, Lowell, MA, United States.
² Department of Biomedical and Nutritional Sciences, Zuckerberg College of Health Sciences, University of Massachusetts Lowell, Lowell, MA, United States.
³ Department of Biomedical and Nutritional Sciences, Zuckerberg College of Health Sciences, University of Massachusetts Lowell, Lowell, MA, United States. Electronic address: katherine_tucker@uml.edu.

PMID: 38599521
DOI: 10.1016/j.ajcnut.2024.04.005

Abstract

Background: Hypomagnesemia is commonly observed in individuals with diabetes, but how diabetes medications alter magnesium (Mg) status remains unclear.

Objectives: We aimed to examine the association between diabetes medication and hypomagnesemia and evaluate whether serum Mg mediates the association between diabetes medication and Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) in a prospective cohort.

Methods: Adults from the Boston Puerto Rican Health Study were included (n = 1106). Multivariable logistic regression models were used to estimate odds ratio (OR) and 95% confidence interval (CI) for cross-sectional association between diabetes medication and hypomagnesemia (serum Mg <0.75 mmol/L). Longitudinal mediation analysis was performed to evaluate the direct and indirect (via serum Mg) associations between diabetes medication and 4-y HOMA-IR in 341 participants with baseline hemoglobin A1c (HbA1c) of ≥6.5%.

Results: Mean age at baseline was 59.0 ± 7.6 y, with 28.0% male and 45.8% with hypomagnesemia. Use of metformin [OR (95% CI) = 3.72 (2.53, 5.48)], sulfonylureas [OR (95% CI) = 1.68 (1.00, 2.83)], and glitazones [OR (95% CI) = 2.09 (1.10, 3.95)], but not insulin, was associated with higher odds of hypomagnesemia. Use of multiple diabetes medications and longer duration of use were associated with higher odds of hypomagnesemia. Serum Mg partially mediated the association between metformin and HOMA-IR [indirect association: β (95% CI) = 1.11 (0.15, 2.07)], which weakened the direct association [β (95% CI) = -5.16 (-9.02, -1.30)] by 22% [total association: β (95% CI) = -4.05 (-7.59, -0.51)]. Similarly, serum Mg mediated 17% of the association between sulfonylureas and elevated HOMA-IR. However, the mediation by serum Mg was weak for insulin and glitazones.

Conclusions: Diabetes medication, especially metformin, was associated with elevated odds of hypomagnesemia, which may weaken the association between metformin and lowering of HOMA-IR. The causal inference needs to be confirmed in further studies.

Keywords: diabetes medication; glitazones; hypomagnesemia; insulin; insulin resistance; metformin; serum magnesium; sulfonylureas; type 2 diabetes.