Skip to Accessibility Tools Skip to Content Skip to Footer

Bilirubin and Type 2 Diabetes: New Evidence for a Causal Relationship

Bilirubin is a yellow pigment produced when red blood cells break down. It is known for causing jaundice, a condition in which the skin and eyes turn yellow because bilirubin builds up.1 Although high levels of bilirubin can be a problem, it is normal to find some amount of bilirubin in the blood. In fact, there is evidence that bilirubin may reduce the risk of developing type 2 diabetes mellitus (T2DM).2

Bilirubin acts like an antioxidant. Antioxidants help to reduce tissue damage and inflammation. Initial studies have shown that mice and humans with higher levels of bilirubin are less likely to have T2DM. However, these studies had limitations, and they could not prove that high bilirubin prevents T2DM. Therefore, researchers in the Netherlands used a special study design called “Mendelian randomization” (Italics) to learn more about this relationship.2 Their goal was to see if people with a gene known to cause higher bilirubin levels (called “rs6742078”) were less likely to have T2DM.

What is Mendelian randomization?

Mendelian randomization is a fairly new way of studying how an “exposure” is related to an outcome.3 An “exposure” is defined as the source of a disease. For example, a study could be done to answer the question: Does high blood pressure (the exposure) cause heart disease (the outcome)?

In this case, the question is: “Does high bilirubin (the exposure) prevent T2DM (the outcome)?”2 This is a hard question to study. Traditional observational studies may tell us that people with high bilirubin are less likely to have T2DM, which is called an association. But these studies do not prove that one factor causes another. For example, could there be a third factor that causes high bilirubin and also prevents T2DM (called “confounding”). Alternatively, T2DM might cause lower bilirubin levels (called “reverse causation”).

One way to sort this out is to look for small variations in the genes that change the exposure. The Dutch researchers knew that a gene variation called “rs6742078” causes people to have higher bilirubin levels.2 There is no known link between this gene and any other factors that to cause T2DM. This means that there are no additional factors confounding the results. Having T2DM cannot change a person’s genes, so researchers also know that their results are not due to reverse causation. Therefore, showing that there is a link between rs6742078 and T2DM would be evidence that there is a causal relationship between bilirubin and T2DM.

More information about Mendelian randomization studies is available here.

Study Population

This study took place in Groningen, the Netherlands.2 It is part of a larger study of that has followed more than 8500 for nearly 20 years to study how kidney and heart disease develop. For the bilirubin study, genetic testing was done on 4016 people for variations in rs6742078.


Over 7.7 years, 6.2% of the people in the bilirubin substudy developed T2DM.2 Just like in earlier studies, people who developed T2DM had lower bilirubin levels than the people who did not develop T2DM.

Most importantly, the researchers found that one variation of rs6742078 was linked with higher levels of bilirubin and lower risk of T2DM. In fact, people with this gene had a 31% lower risk of having T2DM.


While earlier studies could only show that there was an association—or link—between high bilirubin and lower risk of T2DM, this study provides some evidence of cause.2 This information may be useful in many ways. First, it helps researchers to understand more about T2DM and its complications. Second, it might mean that medications that increase bilirubin levels could be helpful for people with T2DM. The authors concluded that more studies should be done to understand exactly how bilirubin might protect people from developing diabetes.

  1. Bilirubin - blood. MedlinePlus.
  2. Abbasi A, Deetman PE, Corpeleijn E, et al. Bilirubin as a potential causal factor in type 2 diabetes risk: a Mendelian randomization study. Diabetes: 2015;64:1459-1469.
  3. Smith GD, Ebrahim S. Chapter 16: Mendelian Randomization: Genetic Variants as Instruments for Strengthening Causal Inference in Observational Studies. In: Weinstein M, Vaupel JW, Wachter KW, eds. National Research Council (US) Committee on Advances in Collecting and Utilizing Biological Indicators and Genetic Information in Social Science Surveys. National Academies Press: 2008.
  4. Introduction to Epidemiology. University of Illinois at Chicago.