<p>We tested the hypothesis
that low plasma adiponectin is observationally and genetic, causally associated
with increased risk of type 2 diabetes. Observational analyses are prone to
confounding and reverse causation, while genetic Mendelian randomization analyses
are much less influenced by these biases. We examined 30,045 Copenhagen individuals observationally (1,751 type 2 diabetes;
plasma adiponectin), 96,903 Copenhagen individuals using <a>one-sample
Mendelian randomization </a>(5,012 type 2 diabetes; five
genetic variants),
and 659,316 Europeans (ADIPOGen, GERA, DIAGRAM, UK
Biobank) using two-sample Mendelian randomization (62,892 type 2 diabetes; ten
genetic variants). Observationally,
and compared to individuals with median plasma adiponectin of 28.9µg/mL(4th
quartile), multivariable adjusted hazard ratios for type 2 diabetes were
1.42(95% CI:1.18-1.72) for 19.2µg/mL(3rd quartile), 2.21(1.84-2.66) for
13.9µg/mL(2nd quartile), and 4.05(3.38-4.86) for 9.2µg/mL(1st quartile).
Corresponding cumulative incidences for type 2 diabetes at age 70 were 3%, 7%, 11%, and
20%, respectively.<b> </b>A 1µg/mL lower
plasma adiponectin conferred a hazard
ratio for type 2 diabetes of 1.07(1.06-1.09), while genetic, causal
risk ratios per 1 unit log-transformed lower plasma adiponectin were
1.13(0.83-1.53) in one-sample
Mendelian randomization and 1.26(1.01-1.57) in two-sample Mendelian
randomization. <a>In conclusion, </a>low plasma
adiponectin is associated with increased risk of type 2 diabetes, an
association that could represent a causal relationship.</p>