Likely causal effects of insulin resistance and IGF-1 bioaction on childhood and adult adiposity: a Mendelian randomization study
Olwi, Duaa I.; Kaisinger, Lena R.; Kentistou, Katherine A.; Vaudel, Marc; Stankovic, Stasa; Njølstad, Pål Rasmus; Johansson, Stefan; Perry, John R. B.; Day, Felix R.; Ong, Ken K.
Journal article, Peer reviewed
Published version

View/ Open
Date
2024Metadata
Show full item recordCollections
- Department of Clinical Science [2536]
- Registrations from Cristin [11745]
Abstract
Background
Circulating insulin and insulin-like growth factor-1 (IGF-1) concentrations are positively correlated with adiposity. However, the causal effects of insulin and IGF-1 on adiposity are unclear.
Methods
We performed two-sample Mendelian randomization analyses to estimate the likely causal effects of fasting insulin and IGF-1 on relative childhood adiposity and adult body mass index (BMI). To improve accuracy and biological interpretation, we applied Steiger filtering (to avoid reverse causality) and ‘biological effect’ filtering of fasting insulin and IGF-1 associated variants.
Results
Fasting insulin-increasing alleles (35 variants also associated with higher fasting glucose, indicative of insulin resistance) were associated with lower relative childhood adiposity (P = 3.8 × 10−3) and lower adult BMI (P = 1.4 × 10−5). IGF-1-increasing alleles also associated with taller childhood height (351 variants indicative of greater IGF-1 bioaction) showed no association with relative childhood adiposity (P = 0.077) or adult BMI (P = 0.562). Conversely, IGF-1-increasing alleles also associated with shorter childhood height (306 variants indicative of IGF-1 resistance) were associated with lower relative childhood adiposity (P = 6.7 × 10−3), but effects on adult BMI were inconclusive.
Conclusions
Genetic causal modelling indicates negative effects of insulin resistance on childhood and adult adiposity, and negative effects of IGF-1 resistance on childhood adiposity. Our findings demonstrate the need to distinguish between bioaction and resistance when modelling variants associated with biomarker concentrations.