Can vitamins really fight cancer? New study uncovers surprising links between micronutrients and cancer risk

By Tarun Sai LomteAug 27 2023Reviewed by Benedette Cuffari, M.Sc.

In a recent study published in the journal BMC Medicine, researchers evaluate the associations between micronutrients and cancer outcomes.

Study: An atlas of associations between 14 micronutrients and 22 cancer outcomes: Mendelian randomization analyses. Image Credit: monticello / Shutterstock.com

Can micronutrients mitigate cancer outcomes?

Previous studies have examined the effects of micronutrients on different health outcomes. Whereas observational studies support the benefit of supplementing micronutrients on cancer risks, randomized trials have not reported these effects. Importantly, observational studies are susceptible to reverse causation and confounding factors, whereas trials are also associated with various limitations, including their high related costs and the lack of powering their analyses adequately for cancer outcomes.

Mendelian randomization (MR) analyses address these challenges by incorporating genetic variants as instrumental variables to evaluate the causal relationship between disease and risk factors. The health benefits of micronutrients described in observational studies over the past decade are not causal associations in MR analyses.

MR methodologies are inconsistent across studies, thus making it challenging to compare and evaluate their robustness. Recently, a systematic review revealed that many MR studies lacked sensitivity analyses. Several studies also used instrumental variables with linkage disequilibrium thresholds that were less strict than conventional thresholds, thereby leading to biased estimates.

About the study

In the present study, researchers perform a two-sample MR using single-nucleotide polymorphisms (SNPs) associated with micronutrient levels as instrumental variables to test the causal relationship between micronutrients and cancer outcomes. To this end, data were obtained from extensive genome-wide association studies (GWASs) for seven minerals, including zinc, selenium, phosphorous, magnesium, iron, copper, and calcium, as well as vitamins A1, B6, B9, B12, C, D, and E.

Likewise, data were used from GWASs from the FinnGen study, the United Kingdom (UK) Biobank study, and other cancer consortia for cancer outcomes. GWASs were available for 22 cancer outcomes from the FinnGen and U.K. Biobank cohorts. Summary statistics from these cohorts were meta-analyzed for each cancer outcome.

Six additional cancer outcomes were identified from updated cancer consortia data. Subsequently, MR was performed to compare findings with those from the MR analyses on U.K. Biobank and FinnGen cohorts. Data for 20 additional cancer subsets were also available for ovarian, lung, and breast cancers from the consortia.

SNPs with a minor allele frequency above 0.01 and not associated with disequilibrium were selected. MR analysis was performed for each exposure-outcome pair.

The inverse variance-weighted method was used to obtain the summary of the association. Bonferroni correction was applied for multiple testing.

Several sensitivity analyses were performed to assess horizontal pleiotropy. Cochran’s Q test, I-squared statistic, and scatter plots were used to analyze heterogeneity.

Study findings

Eighteen out of 308 associations in the FinnGen and U.K. Biobank meta-analyses were statistically significant, two of which showed Bonferroni-corrected significance. An increased risk of breast and colorectal cancers was observed with higher magnesium and vitamin B12 levels, respectively. There was no evidence of pleiotropy, and heterogeneity was low.

The remaining 16 associations showed an increased or decreased risk of cancer outcomes with increased micronutrient levels. Micronutrient levels were not associated when the overall cancer risk was considered. Further, 10 out of 84 associations were significant in the MR analyses of six additional cancer outcomes.

Increased colorectal and breast cancer risk with higher vitamin B12 and magnesium levels, respectively, showed similar effect sizes in UK Biobank-FinnGen and cancer consortia analyses. There was evidence of horizontal pleiotropy for the association between colorectal cancer and vitamin B12, albeit it retained significance after excluding the outlier SNP.

In the MR analyses of 20 cancer subsets, magnesium was associated with increased lumina A-like breast cancer risk. The associations between magnesium levels and lower risks of invasive and endometroid ovarian cancers, and vitamin B12 levels and higher risks of clear cell, serous, invasive, and non-invasive ovarian cancers were concordant between overall and subset cancer analyses.

Conclusions

The current study provides a comprehensive atlas of the associations between cancer and micronutrients through an extensive MR analysis. The associations of vitamin B12 and magnesium levels with colorectal and breast cancers, respectively, were robust, sensitive, and reproducible across cohorts.

Subset cancer analyses showed an association between magnesium levels and luminal A-like breast cancer. However, no specific micronutrient was beneficial against overall cancer risk.

Taken together, these findings will help inform clinicians on regulating micronutrient intake, especially in high-risk groups, as well as future trials.