There’s a well-known relationship between obesity and low vitamin D status. People who weigh more need more vitamin D (because of volumetric dilution, not fat sequestration), thus obese people often have low vitamin D levels. There’s also been some research that suggests vitamin D deficiency may be one cause of obesity. However, it’s never been clear which one impacts the other more. Is it more likely that people with low vitamin D will gain weight or that people with larger bodies have a lower concentration of vitamin D because they are bigger?
Earlier this year a research team in the UK used genetic mutations to try to answer this question.
Vimaleswaran KS, Berry DJ, Lu C, et al. Causal relationship between obesity and vitamin D status: bi-directional Mendelian randomization analysis of multiple cohorts. PLoS medicine. 2013;10(2):e1001383.
While the Vitamin D Council covered this study earlier this year, this time I want to talk about the methodology the researchers used, which we’ll see used more and more in future research.
Typically when cells divide, for example, to heal a cut finger, they use a process called mitosis. Both of the resulting cells have exactly the same DNA, which is organized into pairs of chromosomes (23 pairs in humans). Half of the pair comes from the organism’s mother, the other half from the father.
However, there’s another process, called meiosis, that occurs in cells involved in the reproductive system. With meiosis, segments from each chromosome pair are randomly mixed and single chromosomes appear in the resulting cells. These become paired in the new organism during the reproductive process. This randomization involves DNA segments that are larger than a single gene but smaller than a chromosome. As a result, genetic mutations are, generally speaking, randomly distributed through a population.
This randomization provides a way to determine the direction of causality without doing a randomized controlled trial (RCT). RCTs are useful because they can demonstrate the direction of causation and because the randomization process controls for confounding variables. A confounding variable is some unknown factor that is also involved in the causal relationship. Randomization is assumed to evenly distribute that unknown factor into both the treatment and control groups.
The genetic technique, known as Mendelian randomization, is based on the statistical method of instrumental variables. In the approach used in this study, researchers look for simple mutations involving just a single change in a gene’s DNA sequence. These are called single nucleotide polymorphisms, or SNPs. They result in a different amino acid at that point in the resulting protein. To be of interest in this kind of study, the SNP has to be widely distributed and have a known effect of some aspect of health. In this study, for example, the researchers used 12 SNPs related to body-mass index (BMI) and four SNPs related to serum 25(OH)D levels.
They combined the 12 SNPs related to BMI into a single score and found that the score was related to both higher BMIs and lower vitamin D status. The vitamin D SNPs were divided into two scores, one for two SNPs related to vitamin D synthesis and the other for two SNPs related to vitamin D metabolism. The vitamin D scores were strongly related to vitamin D status, but not to BMI.
The interpretation of these results is that the relationship between BMI and vitamin D status is mostly caused by BMI. Bigger people have lower concentrations of vitamin D because they get the same amount of vitamin D as everyone else, but being bigger, their concentration is lower
It appears there are a number of useful SNPs and other types of mutations related to the genes involved in the vitamin D system, so we can expect to see more studies like this one. If you’d like to learn more about Mendelian randomization, the most readable article on it I’ve found so far is this one: