When testing genotype–phenotype associations using linear regression, departure of the trait distribution from normality can impact both Type I error rate control and statistical power, with worse consequences for rarer variants. Because genotypes are expected to have small effects (if any) investigators now routinely use a two‐stage method, in which they first regress the trait on covariates, obtain residuals, rank‐normalize them, and then use the rank‐normalized residuals in association analysis with the genotypes. Potential confounding signals are assumed to be removed at the first stage, so in practice, no further adjustment is done in the second stage. Here, we show that this widely used approach can lead to tests with undesirable statistical properties, due to both combination of a mis‐specified mean–variance relationship and remaining covariate associations between the rank‐normalized residuals and genotypes. We demonstrate these properties theoretically, and also in applications to genome‐wide and whole‐genome sequencing association studies. We further propose and evaluate an alternative fully adjusted two‐stage approach that adjusts for covariates both when residuals are obtained and in the subsequent association test. This method can reduce excess Type I errors and improve statistical power.

Prior studies suggest dental caries traits in children and adolescents are partially heritable, but there has been no large-scale consortium genome-wide association study (GWAS) to date. We therefore performed GWAS for caries in participants aged 2.5-18.0 years from nine contributing centres. Phenotype definitions were created for the presence or absence of treated or untreated caries, stratified by primary and permanent dentition. All studies tested for association between caries and genotype dosage and the results were combined using fixed-effects meta-analysis. Analysis included up to 19 003 individuals (7530 affected) for primary teeth and 13 353 individuals (5875 affected) for permanent teeth. Evidence for association with caries status was observed at rs1594318-C for primary teeth [intronic within ALLC, odds ratio (OR) 0.85, effect allele frequency (EAF) 0.60, P 4.13e-8] and rs7738851-A (intronic within NEDD9, OR 1.28, EAF 0.85, P 1.63e-8) for permanent teeth. Consortium-wide estimated heritability of caries was low [h2 of 1% (95% CI: 0%: 7%) and 6% (95% CI 0%: 13%) for primary and permanent dentitions, respectively] compared with corresponding within-study estimates [h2 of 28% (95% CI: 9%: 48%) and 17% (95% CI: 2%: 31%)] or previously published estimates. This study was designed to identify common genetic variants with modest effects which are consistent across different populations. We found few single variants associated with caries status under these assumptions. Phenotypic heterogeneity between cohorts and limited statistical power will have contributed; these findings could also reflect complexity not captured by our study design, such as genetic effects which are conditional on environmental exposure.

OBJECTIVES: Counts of missing teeth or measures of incident tooth loss are gaining attention as a simple way to measure dental status in large population studies. We explore the meaning of these metrics and how missing teeth might influence other measures of dental status.

METHODS: An observational study was performed in 2 contrasting adult populations. In total, 62 522 adult participants were available with clinically assessed caries and periodontal indices from the Swedish arm of the Gene-Lifestyle Interactions and Dental Endpoints Study (GLIDE) and the Korea National Health and Nutrition Examination Survey (KNHANES) in the Republic of Korea. Longitudinal measures of tooth loss were available for 28 244 participants in GLIDE with median follow-up of 10.6 years.

RESULTS: In longitudinal analysis, hazard for tooth loss was associated with baseline dental status (previous tooth loss, periodontal status and caries status) and socio-demographic variables (age, smoking status and highest educational level). Analysis of cross-sectional data suggested that indices of caries exposure were not independent of periodontal status. The strength and direction of association varied between groups, even for measures specifically intended to avoid measuring tooth loss. Individuals with impaired periodontal health (community periodontal index [CPI] 3 or higher in any sextant) had higher standardized decayed and filled surfaces (DFS; number of DFS divided by total number of tooth surfaces) in GLIDE (incidence risk ratio [IRR] 1.05 [95% CI: 1.04, 1.07], but lower standardized DFS in KNHANES (IRR: 0.95 [0.92, 0.98]) than individuals with better periodontal health (CPI <3 in all sextants).

CONCLUSIONS: Incident tooth loss is a complex measure of dental disease, with multiple determinants. The relative importance of dental caries and periodontal disease as drivers of tooth loss differs between age groups. Measures of dental caries exposure are associated with periodontal status in the studied populations, and these associations can be population-specific. Consideration of the study-specific properties of these metrics may be required for valid inference in large population studies.

Background: A positive association between nonfermented milk intake and increased all-cause mortality was recently reported, but overall, the association between dairy intake and mortality is inconclusive. Objective: We studied associations between intake of dairy products and all-cause mortality with an emphasis on nonfermented milk and fat content. Design: A total of 103,256 adult participants (women: 51.0%) from Northern Sweden were included (7121 deaths; mean follow-up: 13.7 y). Associations between all-cause mortality and reported intakes of nonfermented milk (total or by fat content), fermented milk, cheese, and butter were tested with the use of Cox proportional hazards models that were adjusted for age, sex, body mass index, smoking status, education, energy intake, examination year, and physical activity. To circumvent confounding, Mendelian randomization was applied in a subsample via the lactase LCT-13910 C/T single nucleotide polymorphism that is associated with lactose tolerance and milk intake. Results: High consumers of nonfermented milk (>= 2.5 times/d) had a 32% increased hazard (HR: 1.32; 95% CI: 1.18, 1.48) for all-cause mortality compared with that of subjects who consumed milk <= 1 time/wk. The corresponding value for butter was 11% (HR: 1.11; 95% CI: 1.07, 1.21). All nonfermented milk-fat types were independently associated with increased HRs, but compared with full-fat milk, HRs were lower in consumers of medium-and low-fat milk. Fermented milk intake (HR: 0.90; 95% CI: 0.86, 0.94) and cheese intake (HR: 0.93; 95% CI: 0.91, 0.96) were negatively associated with mortality. Results were slightly attenuated by lifestyle adjustments but were robust in sensitivity analyses. Mortality was not significantly associated with the LCT-13910 C/T genotype in the smaller subsample. The amount and type of milk intake was associated with lifestyle variables. Conclusions: In the present Swedish cohort study, intakes of nonfermented milk and butter are associated with higher all-cause mortality, and fermented milk and cheese intakes are associated with lower all-cause mortality. Residual confounding by lifestyle cannot be excluded, and Mendelian randomization needs to be examined in a larger sample.

Background: Cross-sectional genome-wide association studies have identified hundreds of loci associated with blood lipids and related cardiovascular traits, but few genetic association studies have focused on long-term changes in blood lipids. Methods: Participants from the GLACIER Study (N-max = 3492) were genotyped with the MetaboChip array, from which 29 387 SNPs (single nucleotide polymorphisms;replication, fine-mapping regions and wildcard SNPs for lipid traits) were extracted for association tests with 10-year change in total cholesterol (Delta TC) and triglycerides (Delta TG). Four additional prospective cohort studies (MDC, PIVUS, ULSAM, MRC Ely; N-max = 8263 participants) were used for replication. We conducted an in silico look-up for association with coronary artery disease (CAD) in the Coronary ARtery DIsease Genome-wide Replication and Meta-analysis (CARDIoGRAMplusC4D) Consortium (N similar to 190 000) and functional annotation for the top ranking variants. Results: In total, 956 variants were associated (P < 0.01) with either Delta TC or Delta TG in GLACIER. In GLACIER, chr19:50121999 at APOE was associated with Delta TG and multiple SNPs in the APOA1/A4/C3/A5 region at genome-wide significance (P < 5 x 10(-8)), whereas variants in four loci, DOCK7, BRE, SYNE1 and KCNIP1, reached study-wide significance (P < 1.7 x 10(-6)). The rs7412 variant at APOE was associated with DTC in GLACIER (P < 1.7 x 10(-6)). In pooled analyses of all cohorts, 139 SNPs at six and five loci were associated with Delta TC and for Delta TG, respectively (P < 10(-3)). Of these, a variant at CAPN3 (P = 1.2 x 10(-4)), multiple variants at HPR (P-min = 1.5 x 10(-6)) and a variant at SIX5 (P = 1.9 x 10(-4)) showed evidence for association with CAD. Conclusions: We identified seven novel genomic regions associated with long-term changes in blood lipids, of which three also raise CAD risk.

Phenotypic variance heterogeneity across genotypes at a single nucleotide polymorphism (SNP) may reflect underlying gene-environment (GxE) or gene-gene interactions. We modeled variance heterogeneity for blood lipids and BMI in up to 44,211 participants and investigated relationships between variance effects (P-v), GxE interaction effects (with smoking and physical activity), and marginal genetic effects (P-m). Correlations between P-v and P-m were stronger for SNPs with established marginal effects (Spearman's rho = 0.401 for triglycerides, and rho = 0.236 for BMI) compared to all SNPs. When P-v and P-m were compared for all pruned SNPs, only BMI was statistically significant (Spearman's rho = 0.010). Overall, SNPs with established marginal effects were overrepresented in the nominally significant part of the P-v distribution (P-binomial < 0.05). SNPs from the top 1% of the P-m distribution for BMI had more significant P-v values (Pmann-Whitney = 1.46x10(-5)), and the odds ratio of SNPs with nominally significant (< 0.05) P-m and P-v was 1.33 (95% CI: 1.12, 1.57) for BMI. Moreover, BMI SNPs with nominally significant GxE interaction P-values (Pint < 0.05) were enriched with nominally significant P-v values (P-binomial = 8.63x10(-9) and 8.52x10(-7) for SNP x smoking and SNP x physical activity, respectively). We conclude that some loci with strong marginal effects may be good candidates for GxE, and variance-based prioritization can be used to identify them.

BACKGROUND: Obesity is a complex disease caused by the interplay of genetic and lifestyle factors, but identification of gene-lifestyle interactions in obesity has remained challenging. Few large-scale studies have reported use of genome-wide approaches to investigate gene-lifestyle interactions in obesity. METHODS: In the Pakistan Risk of Myocardial Infraction Study, a cross-sectional study based in Pakistan, we calculated body mass index (BMI) variance estimates (square of the residual of inverse-normal transformed BMI z-score) in 14 131 participants and conducted genome-wide heterogeneity of variance analyses (GWHVA) for this outcome. All analyses were adjusted for age, age(2), sex and genetic ancestry. RESULTS: The GWHVA analyses identified an intronic variant, rs140133294, in the FLJ33544 gene in association with BMI variance (P-value = 3.1 x 10(-8)). In explicit tests of gene x lifestyle interaction, smoking was found to significantly modify the effect of rs140133294 on BMI (Pinteraction = 0.0005), whereby the minor allele (T) was associated with lower BMI in current smokers, while positively associated with BMI in never smokers. Analyses of ENCODE data at the FLJ33534 locus revealed features indicative of open chromatin and high confidence DNA-binding motifs for several transcription factors, providing suggestive biological support for a mechanism of interaction. CONCLUSIONS: In summary, we have identified a novel interaction between smoking and variation at the FLJ33534 locus in relation to BMI in people from Pakistan.

BACKGROUND: Recent cross-sectional genome-wide scans have reported associations of 97 independent loci with body mass index (BMI). In 3541 middle-aged adult participants from the GLACIER Study, we tested whether these loci are associated with 10-year changes in BMI and other cardiometabolic traits (fasting and 2-h glucose, triglycerides, total cholesterol, and systolic and diastolic blood pressures).

METHODS: A BMI-specific genetic risk score (GRS) was calculated by summing the BMI-associated effect alleles at each locus. Trait-specific cardiometabolic GRSs comprised only the loci that show nominal association (P⩽0.10) with the respective trait in the original cross-sectional study. In longitudinal genetic association analyses, the second visit trait measure (assessed ~10 years after baseline) was used as the dependent variable and the models were adjusted for the baseline measure of the outcome trait, age, age(2), fasting time (for glucose and lipid traits), sex, follow-up time and population substructure.

RESULTS: The BMI-specific GRS was associated with increased BMI at follow-up (β=0.014 kg m(-2) per allele per 10-year follow-up, s.e.=0.006, P=0.019) as were three loci (PARK2 rs13191362, P=0.005; C6orf106 rs205262, P=0.043; and C9orf93 rs4740619, P=0.01). Although not withstanding Bonferroni correction, a handful of single-nucleotide polymorphisms was nominally associated with changes in blood pressure, glucose and lipid levels.

CONCLUSIONS: Collectively, established BMI-associated loci convey modest but statistically significant time-dependent associations with long-term changes in BMI, suggesting a role for effect modification by factors that change with time in this population.

Sixty genetic loci associated with abdominal obesity, measured by waist circumference (WC) and waist-hip ratio (WHR), have been previously identified, primarily from studies conducted in Europeanancestry populations. We conducted a meta-analysis of associations of abdominal obesity with approximately 2.5 million single nucleotide polymorphisms (SNPs) among 53,052 (for WC) and 48,312 (for WHR) individuals of Asian descent, and replicated 33 selected SNPs among 3,762 to 17,110 additional individuals. We identified four novel loci near the EFEMP1, ADAMTSL3, CNPY2, and GNAS genes that were associated with WC after adjustment for body mass index (BMI); two loci near the NID2 and HLA-DRB5 genes associated with WHR after adjustment for BMI, and three loci near the CEP120, TSC22D2, and SLC22A2 genes associated with WC without adjustment for BMI. Functional enrichment analyses revealed enrichment of corticotropin-releasing hormone signaling, GNRH signaling, and/or CDK5 signaling pathways for those newly-identified loci. Our study provides additional insight on genetic contribution to abdominal obesity.

To increase our understanding of the genetic basis of adiposity and its links to cardiometabolic disease risk, we conducted a genome-wide association meta-analysis of body fat percentage (BF%) in up to 100,716 individuals. Twelve loci reached genome-wide significance (P < 5 x 10(-8)), of which eight were previously associated with increased overall adiposity (BMI, BF%) and four (in or near COBLL1/GRB14, IGF2BP1, PLA2G6, CRTC1) were novel associations with BF%. Seven loci showed a larger effect on BF% than on BMI, suggestive of a primary association with adiposity, while five loci showed larger effects on BMI than on BF%, suggesting association with both fat and lean mass. In particular, the loci more strongly associated with BF% showed distinct cross-phenotype association signatures with a range of cardiometabolic traits revealing new insights in the link between adiposity and disease risk.