Sugar-sweetened beverage intake associations with fasting glucose and insulin concentrations are not modified by selected genetic variants in a ChREBP-FGF21 pathway: a meta-analysis

AIMS/HYPOTHESIS

Sugar-sweetened beverages (SSBs) are a major dietary contributor to fructose intake. A molecular pathway involving the carbohydrate responsive element-binding protein (ChREBP) and the metabolic hormone fibroblast growth factor 21 (FGF21) may influence sugar metabolism and, thereby, contribute to fructose-induced metabolic disease. We hypothesise that common variants in 11 genes involved in fructose metabolism and the ChREBP-FGF21 pathway may interact with SSB intake to exacerbate positive associations between higher SSB intake and glycaemic traits.

METHODS

Data from 11 cohorts (six discovery and five replication) in the CHARGE (Cohorts for Heart and Aging Research in Genomic Epidemiology) Consortium provided association and interaction results from 34,748 adults of European descent. SSB intake (soft drinks, fruit punches, lemonades or other fruit drinks) was derived from food-frequency questionnaires and food diaries. In fixed-effects meta-analyses, we quantified: (1) the associations between SSBs and glycaemic traits (fasting glucose and fasting insulin); and (2) the interactions between SSBs and 18 independent SNPs related to the ChREBP-FGF21 pathway.

RESULTS

In our combined meta-analyses of discovery and replication cohorts, after adjustment for age, sex, energy intake, BMI and other dietary covariates, each additional serving of SSB intake was associated with higher fasting glucose (β ± SE 0.014 ± 0.004 [mmol/l], p = 1.5 × 10^-3) and higher fasting insulin (0.030 ± 0.005 [log _e pmol/l], p = 2.0 × 10^-10). No significant interactions on glycaemic traits were observed between SSB intake and selected SNPs. While a suggestive interaction was observed in the discovery cohorts with a SNP (rs1542423) in the β-Klotho (KLB) locus on fasting insulin (0.030 ± 0.011 log _e pmol/l, uncorrected p = 0.006), results in the replication cohorts and combined meta-analyses were non-significant.

CONCLUSIONS/INTERPRETATION

In this large meta-analysis, we observed that SSB intake was associated with higher fasting glucose and insulin. Although a suggestive interaction with a genetic variant in the ChREBP-FGF21 pathway was observed in the discovery cohorts, this observation was not confirmed in the replication analysis.

TRIAL REGISTRATION

Trials related to this study were registered at clinicaltrials.gov as NCT00005131 (Atherosclerosis Risk in Communities), NCT00005133 (Cardiovascular Health Study), NCT00005121 (Framingham Offspring Study), NCT00005487 (Multi-Ethnic Study of Atherosclerosis) and NCT00005152 (Nurses' Health Study).

Ethnicity and skin autofluorescence-based risk-engines for cardiovascular disease and diabetes mellitus

Skin auto fluorescence (SAF) is used as a proxy for the accumulation of advanced glycation end products (AGEs) and has been proposed to stratify patients into cardiovascular disease (CVD) and diabetes mellitus (DM) risk groups. This study evaluates the effects of seven different ethnicities (Arab, Central-East African, Eastern Mediterranean, European, North African, South Asian and Southeast Asian) and gender on SAF as well as validating SAF assessment as a risk estimation tool for CVD and DM in an Arabian cohort. SAF data from self-reported healthy 2,780 individuals, collated from three independent studies, has been linear modelled using age and gender as a covariate. A cross-study harmonized effect size (Cohens’d) is provided for each ethnicity. Furthermore, new data has been collected from a clinically well-defined patient group of 235 individuals, to evaluate SAF as a clinical tool for DM and CVD-risk estimation in an Arab cohort. In an Arab population, SAF-based CVD and/or DM risk-estimation can be improved by referencing to ethnicity and gender-specific SAF values. Highest SAF values were observed for the North African population, followed by East Mediterranean, Arab, South Asian and European populations. The South Asian population had a slightly steeper slope in SAF values with age compared to other ethnic groups. All ethnic groups except Europeans showed a significant gender effect. When compared with a European group, effect size was highest for Eastern Mediterranean group and lowest for South Asian group. The Central-East African and Southeast Asian ethnicity matched closest to the Arab and Eastern Mediterranean ethnicities, respectively. Ethnic and gender-specific data improves performance in SAF-based CVD and DM risk estimation. The provided harmonized effect size allows a direct comparison of SAF in different ethnicities. For the first time, gender differences in SAF are described for North African and East Mediterranean populations.