Genetic risk score for adult body mass index associations with childhood and adolescent weight gain in an African population

Richard J Munthali; Venesa Sahibdeen; Juliana Kagura; Liesl M Hendry; Shane A Norris; Ken K Ong; Felix R Day; Zané Lombard

doi:10.1186/s12263-018-0613-7

Genetic risk score for adult body mass index associations with childhood and adolescent weight gain in an African population

Genes Nutr. 2018 Aug 1:13:24. doi: 10.1186/s12263-018-0613-7. eCollection 2018.

Authors

Richard J Munthali^{1

2

3}, Venesa Sahibdeen^{2

4}, Juliana Kagura³, Liesl M Hendry^{1

2}, Shane A Norris³, Ken K Ong^{3

5}, Felix R Day⁵, Zané Lombard^{1

2

4}

Affiliations

¹ 1Faculty of Science, School of Molecular and Cell Biology, University of the Witwatersrand, Johannesburg, South Africa.
² 2Sydney Brenner Institute for Molecular Bioscience (SBIMB), University of the Witwatersrand, The Mount, 9 Jubilee Road, Parktown, Johannesburg, Gauteng 2193 South Africa.
³ 3MRC/Wits Developmental Pathways for Health Research Unit (DPHRU), University of the Witwatersrand, Johannesburg, South Africa.
⁴ 4Faculty of Health Sciences, Division of Human Genetics, School of Pathology, University of the Witwatersrand and National Health Laboratory Service, Johannesburg, South Africa.
⁵ 5MRC Epidemiology Unit, Institute of Metabolic Science, University of Cambridge, Cambridge, UK.

Abstract

Background: Ninety-seven independent single nucleotide polymorphisms (SNPs) are robustly associated with adult body mass index (BMI kg/m²) in Caucasian populations. The relevance of such variants in African populations at different stages of the life course (such as childhood) is unclear. We tested whether a genetic risk score composed of the aforementioned SNPs was associated with BMI from infancy to early adulthood. We further tested whether this genetic effect was mediated by conditional weight gain at different growth periods. We used data from the Birth to Twenty Plus Cohort (Bt20+), for 971 urban South African black children from birth to 18 years. DNA was collected at 13 years old and was genotyped using the Metabochip (Illumina) array. The weighted genetic risk score (wGRS) for BMI was constructed based on 71 of the 97 previously reported SNPs.

Results: The cross-sectional association between the wGRS and BMI strengthened with age from 5 to 18 years. The significant associations were observed from 11 to 18 years, and peak effect sizes were observed at 13 and 14 years of age. Results from the linear mixed effects models showed significant interactions between the wGRS and age on longitudinal BMI but no such interactions were observed in sex and the wGRS. A higher wGRS was associated with an increased relative risk of belonging to the early onset obese longitudinal BMI trajectory (relative risk = 1.88; 95%CI 1.28 to 2.76) compared to belonging to a normal longitudinal BMI trajectory. Adolescent conditional relative weight gain had a suggestive mediation effect of 56% on the association between wGRS and obesity risk at 18 years.

Conclusions: The results suggest that genetic susceptibility to higher adult BMI can be tracked from childhood in this African population. This supports the notion that prevention of adult obesity should begin early in life. The genetic risk score combined with other non-genetic risk factors, such as BMI trajectory membership in our case, has the potential to be used to screen for early identification of individuals at increased risk of obesity and other related NCD risk factors in order to reduce the adverse health risk outcomes later.

Keywords: Body mass index; Childhood adiposity; Genetic risk score; Mediation analysis; Obesity; Weight gain.

Abstract

Grants and funding