Causal Effect of Chondroitin, Glucosamine, Vitamin, and Mineral Intake on Kidney Function: A Mendelian Randomization Study

Abstract

The causal effects of chondroitin, glucosamine, and vitamin/mineral supplement intake on kidney function remain unknown, despite being commonly used. We conducted a two-sample summary-level Mendelian randomization (MR) analysis to test for causal associations between regular dietary supplement intake and kidney function. Genetic instruments for chondroitin, glucosamine, and vitamin/mineral supplement intake were obtained from a genome-wide association study of European ancestry. Summary statistics for the log-transformed estimated glomerular filtration rate (log-eGFR) were provided by the CKDGen consortium. The multiplicative random-effects inverse-variance weighted method showed that genetically predicted chondroitin and glucosamine intake was causally associated with a lower eGFR (chondroitin, eGFR change beta = -0.113%, standard error (SE) = 0.03%, p-value = 2 × 10^-4; glucosamine, eGFR change beta = -0.240%, SE = 0.035%, p-value = 6 × 10^-12). However, a genetically predicted vitamin/mineral supplement intake was associated with a higher eGFR (eGFR change beta = 1.426%, SE = 0.136%, p-value = 1 × 10^-25). Validation analyses and pleiotropy-robust MR results for chondroitin and vitamin/mineral supplement intake supported the main results. Our MR study suggests a potential causal effect of chondroitin and glucosamine intake on kidney function. Therefore, clinicians should carefully monitor their long-term effects.

Keywords: chondroitin; glomerular filtration rate; glucosamine; minerals; vitamins.

Figure 1

Study flow diagram. The genetic instruments for genetically proxied chondroitin, glucosamine, and vitamin/mineral intake were obtained from GWAS based on white British participants from UKB (n = 361,194). The numbers of lead SNPs for chondroitin, glucosamine, and vitamin/mineral supplement intake are described in brackets. Two-sample summary-level MR was conducted on three datasets for kidney function traits, including main analysis implemented on CKDGen phase 4 GWAS meta-analysis for log-eGFRcr (n = 765,348). CKDGen is a genetic database for kidney function traits which is publicly available. GWAS = genome-wide association study; UKB = UK Biobank, SNP = single nucleotide polymorphism; MR = Mendelian randomization; eGFR = estimated glomerular filtration rate; log-eGFRcr = creatinine-based log-transformed eGFR.

Figure 2

Mendelian randomization plots for causal association between (A) chondroitin, (B) glucosamine, and (C) vitamin/mineral supplement intake and kidney function. For each exposure, left and right plots show scatter and leave-one-out plots, respectively. Scatter plot of the effect sizes of the association between SNP-genetically predicted intake of each dietary supplement (x-axis) against the SNP-log-eGFRcr (y-axis) with standard error bars. The slopes of the lines correspond to the causal estimates per method. Each black point of the leave-one-out analysis represents the MR-IVW method applied to estimate the causal effect of genetically predicted intake of each dietary supplement on kidney function, excluding that certain variant from the analysis. The red point indicates the IVW estimate using all SNPs. There were no instances where the exclusion of a single SNP led to dramatic changes in the overall result.