Abstract
Aims/Hypothesis:
This is the first study to examine the association between variants of the glucagon-like-peptide-1 receptor gene (GLP-1R) and metabolic characteristics among youth. We explored separate associations of three GLP-1R polymorphisms (rs10305420, rs6923761, and rs1042044) with BMI trajectories and markers of glucose-insulin homeostasis.
Methods:
Mixed models examined associations between GLP-1R polymorphisms and trajectories of BMI. Linear models examined associations of GLP-1R polymorphisms with glucose and insulin concentrations across OGTT, insulin sensitivity (HOMA2-IR), insulin secretion (insulinogenic index and HOMA2-%B), and beta cell function (oral disposition index).
Results:
Rs10305420 and rs6923761, but not rs1042044, were associated with growth and metabolic characteristics in early life. Rs6923761 genotype GG was associated with faster BMI growth velocity, when compared to carriers of the minor allele (difference in velocity [95% CI]: 0.16 kg/m2/year [0.07, 0.24] at age 10), which led to significantly higher average BMI by age 16 (average difference [95% CI]: 1.29 kg/m2 [0.22, 2.37]). Rs10305420 CC and rs6923761 GG genotypes had higher HOMA2-IR (β[95% CI]: 1.19%[1.06, 1.32] and 1.13%[1.01, 1.26], respectively) compared to minor allele carriers. Rs10305420 CC had higher HOMA2-%B (β[95% CI]: 1.09%[1.01, 1.17]), and higher stimulated insulin secretion at 30-minutes (β[95% CI]: 27.62 uIU/mL [3.00, 25.24]) and 120-minutes (β[95% CI]: 18.94 uIU/mL [1.04, 36.84), when compared to carriers of the minor allele.
Conclusions:
GLP-1R polymorphisms were associated with faster BMI growth across development, and lower estimated insulin sensitivity and higher compensatory insulin secretion during adolescence. GLP-1R polymorphisms should be considered in future pediatric studies of genetic susceptibility for obesity and diabetes.
Keywords: adolescence, BMI, Glucagon-like-protein-1 receptor, insulin, pediatric, polymorphism
Introduction
Studies have shown that common genotypes of the glucagon-like-peptide-1 receptor gene (GLP-1R) are associated with adult metabolic traits. Polymorphisms of the GLP-1R gene have been associated with adult glycemia (1–4), glucose-dependent insulin response (1,5,6), early insulin secretion (1), insulin sensitivity (5), beta cell health (1,7,8), and anthropometrics (3,9–13). The major homozygous genotype CC for GLP-1R polymorphism rs10305420 has been associated with higher glycated hemoglobin (HbA1c) among adults with type 2 diabetes (14). A genome-wide association study (GWAS) showed that carrying the major C allele for rs10305420 was associated with increased risk of adult type 2 diabetes (15). Among a Spanish populations of adults with obesity, carriers of the major homozygous genotype GG for GLP-1R polymorphism rs6923761 had higher fasting insulin (5), HOMA-IR (5), BMI (5,9,11,12,16), compared to carriers of the minor allele. A large study from the UK Biobank showed that the major homozygous genotype CC for GLP-1R polymorphism rs1042044 was marginally associated with higher adult HbA1c and BMI (3).
Few studies have assessed GLP-1R polymorphisms in the children (17,18), and none have focused on obesity and glucose-insulin homeostasis. It is unknown if GLP-1R polymorphisms are associated with BMI growth in early life, and if so, when effects may begin to develop. No studies have examined whether GLP-1R polymorphisms are associated with glucose-insulin metabolism during adolescence, a developmental stage when youth-onset type 2 diabetes typically develops (19).
In this study, we focus on three previously studied, high frequency polymorphisms of the GLP-1R gene: rs10305420, rs6923761, and rs1042044. These polymorphisms were described by Koole et al. (2011) or were identified by GWAS (3,15). When compared across all populations in LDLink (https://ldlink.nih.gov), the three polymorphisms are in linkage equilibrium, and thus have independent effects. The three polymorphisms make amino acid changes in different domains of the GLP-1R protein (Table 1).
Table 1:
A summary of characteristics for the three common missense GLP-1R polymorphism included in this study.
| rs10305420 | rs6923761 | rs1042044 | |
|---|---|---|---|
| Major allele [residue (frequency in global population)] | C (0.67) | G (0.70) | C (0.55) |
| Minor allele [residue (frequency in global population)]] | T (0.33) | A (0.30) | A (0.45) |
| Functional domain | N-terminal signal peptide | Transmembrane 1 | Second intracellular loop |
| Protein residue | 7 | 168 | 260 |
| Amino acid change from major to minor allele | Proline -> Leucine | Glycine -> Serine | Leucine -> Phenylalanine |
| Change in hydrophobicity | No | Yes | No |
| Change in capability for phosphorylation | No | Yes | No |
| Proposed mechanism (s) | ↓translocation of GLP1 receptor to cell surface |
We leveraged data from the Exploring Perinatal Outcomes among CHildren (EPOCH) study (20–22). Participants were genotyped for all three polymorphisms, underwent one 2-hour 75g oral glucose tolerance test (OGTT) during adolescence, and provided longitudinal BMI measurements from birth through adolescence. We assessed two aims. First, assuming a dominant genetic model, we characterized separate associations between GLP-1R genotype groups and BMI growth. Second, we characterized associations between GLP-1R genotypes and markers of glucose-insulin homeostasis from OGTT.
Materials and Methods
Study Population
EPOCH is a historical prospective cohort that studied associations between in utero exposure to maternal gestational diabetes (GDM) and offspring outcomes (23), in Denver, Colorado. EPOCH recruited 604 mother/child dyads from Kaiser Permanente of Colorado (KPOC). Two study visits were conducted: in childhood at ~10 years of age (2006–2010), and in adolescence at ~16 years of age (2012–2016). Our analytic sample includes all participants with genotyping in childhood (N = 464). The study was approved by the Colorado Multiple Institutional Review Board (protocol no. 05–0623). Mothers provided written informed consent and children provided verbal assent.
GLP-1R genotyping
Venous blood was collected from children and stored at −80°C. DNA isolation, concentration, and purification has been described elsewhere (24,25). DNA was genotyped in two batches. The first batch of 336 samples was genotyped using the Illumina Infinium Omni2.5–8 v1.1 BeadChip (Illumina, San Diego, California). The second batch of 140 samples was genotyped using the Illumina Multi-Ethnic Global Array (MEGA) v1.0. Variants genotyped on the Omni2.5–8 v1.1 BeadChip were filtered to retain variants that were represented at the same chromosome and position on the most up-to-date version of the Omni 2.5 array (Omni2.5–8 v1.4). Genetic principal components were calculated using variants that were directly genotyped and passed quality control on both arrays to capture the array type batch effect as well as population substructure. Linkage disequilibrium pruning was performed on variants with a minor allele frequency of at least 5%. Independent variants were retained with a pairwise correlation of less than 0.2. GLP-1R polymorphisms rs6923761 and rs1042044 were directly genotyped. The Omni and MEGA data were combined for analysis.
Imputation of rs10305420
Separately for each array batch, the Michigan Imputation Server (v1.0.4) was used to phase and impute missing genotype, using the 1000 Genomes Phase 3 (v5) reference panel and Eagle v2.5 phasing. To estimate the number of copies of the T allele, imputed dosage was computed as a continuous value ranging from 0 to 2 and then categorized as 0 if < 0.5, 1 if ≥ 0.5 and < 1.5, and 2 if ≥ 2.
Description of GLP-1R variants
For consistency across polymorphisms, we required that the reference allele align with the major allele observed in the EPOCH cohort. For all three polymorphisms, the major allele observed in the EPOCH cohort matched the major allele observed within the global population on dbSNP (26). The reference alleles shown on dbSNP for rs10305420 and rs6923761 correspond to the global population major allele (C and G, respectively). The reference allele shown on dbSNP for rs1042044 corresponds to the global population minor A allele (Table 1). Consequently, we flipped the reference and alternate alleles for rs1042044, so that the reference aligned with the global population major allele (C).
Genetic models
While minor alleles are often associated with risk, the major homozygous genotypes for rs10305420, rs6923761, and rs1042044 are associated with worse anthropometric or glucose and insulin traits in adults (3,14,27). Each polymorphism was categorized using dominant effect coding with respect to the minor allele (i.e. comparing individuals with one or two copies of the minor allele to those with two copies of the major allele). Polymorphism rs10305420 was categorized as 0 for genotype CC and 1 for genotypes CT or TT. Polymorphism rs6923761 was categorized as 0 for genotype GG and 1 for genotypes GG or GA. Polymorphism rs1042044 was categorized as 0 for genotype CC and 1 for genotypes CA or AA. Although rs10305420 was imputed, the distribution of estimated minor allele dosage was highly clustered around the values of 0, 1, and 2 (Fig. 1).
Fig. 1:
Distribution of imputation results for rs10305420.
Assessment of BMI
The EPOCH study collected measurements of BMI during research visits and by medical record abstraction. During research visits, height and weight were measured in light clothing and without shoes. The average of duplicate measures was used for analysis. Height was measured to the nearest 0.1cm using a portable Seca stadiometer (Chino, CA). Weight was measured using a portable electronic Seca scale (Chino, CA). Medical record abstraction extracted heights and weights from birth through adolescence. Over 10,000 measurements of height and weight were collected across 584 participants, and 73% of participants had nine or more repeated measurements of height and weight. Participants trajectories of height, and then weight, were cleaned using a validated algorithm that detects and removes biologically implausible measurements of height or weight from longitudinal pediatric growth data (28).
BMI was calculated by dividing weight in kilograms by height in meters squared (kg/m2). BMI analysis used measurements collected at or after 27 months of age, when children were able to stand for height measurements. This approach matches that of Crume et al (29).
Oral glucose tolerance testing
EPOCH study participants underwent a 2-hour, 75g OGTT at the second research visit (N = 387). Glucose and insulin concentrations were measured from venous blood draws performed at baseline, 30-minutes, and 120-minutes post-glucose challenge. Glucose was measured using enzymatic kits and an AU400e Chemistry Analyzer. Insulin was measured via radioimmunoassay (Millipore, Darmstadt, Germany).
Homeostasis model assessment calculators (HOMA2, https://www.rdm.ox.ac.uk/about/our-clinical-facilities-and-mrc-units/DTU/software/homa) were used to estimate insulin sensitivity (HOMA2-IR) and insulin secretion ((HOMA2-%B) in the fasting state (30).
Insulinogenic index and beta cell function were used to assess glucose-insulin homeostasis in a stimulated state. Insulinogenic index (31) evaluates how insulin levels change from a fasting state to 30-minutes after glucose intake, while also accounting for changes in plasma glucose levels [(Fasting insulin – 30 minute insulin) / (fasting glucose – 30 minute glucose)]. We used oral disposition index (32) to estimate insulin secretion in the context of insulin sensitivity {[(Fasting insulin – 30 minute insulin) / (fasting glucose – 30 minute glucose)] × (1/fasting insulin)}.
Other measurements
Participant demographics were collected by self-report at the first EPOCH visit. Self-identified race and ethnicity, which we view as a social construct (33), was categorized as Hispanic, non-Hispanic White, non-Hispanic Black, and non-Hispanic Other. Participant sex assigned at birth was self-reported by participants and mothers. Age was calculated as the difference between the date of the respective visit and the date of birth. Overweight or obesity were defined as a BMI-for-age Z-scores ≥ 1.
Statistical Methods
Participant characteristics were stratified by genotype for each polymorphism. Continuous variables were reported as mean and standard deviation, and group comparisons were conducted using the Student’s t-test. Categorical variables were reported as number and percent, and group comparisons were conducted using either the chi-squared test or the Fisher’s exact test.
We assessed if there were differences in participant characteristics between the full cohort and the subsets of data used for each model. Models of BMI trajectories required that participants have genotyping and at least two measurements of BMI. Models of OGTT data required that participants have genotyping, adolescent OGTT, and all model covariates. For each pair of datasets, we compared participant sex, age at visit 2, race/ethnicity, and overweight or obesity status at both research visits.
BMI trajectories across childhood and adolescence
Separate general linear mixed models were fit to assess the association between each GLP-1R genotype and repeated measurements of BMI between the ages of 27 months and 19 years. The predictors included polynomials of age (age, age2, and age3), an indicator for genotype, and the interactions between genotype and each age polynomial. Models were adjusted with an indicator for female sex and the first three genetic principal components, following methods of Stanislawski et al (24). A random intercept and slope, with unstructured covariance between the random effects, were fit to account for within-participant correlation between repeated measurements and for increasing variance in BMI across time. Model assumptions were assessed using jackknife studentized residuals. We used a planned sequential testing approach to examine the utility of age polynomial-by-genotype interaction terms, and then higher order terms in age.
In the best-fitting model for each GLP-1R polymorphism, we considered three hypothesis tests. For the primary hypothesis, we tested if GLP-1R genotype was associated with the BMI over time. The hypothesis test included all main effect and interaction terms including the GLP-1R genotype that remained in the model and used a multiple degree of freedom chunk test to assess significance. If the primary hypothesis test was significant, we conducted two additional tests. First, we assessed associations between GLP-1R genotype and BMI velocity. Second, we assessed for differences in BMI at ages 10 and 16 years between GLP-1R genotypes. Significance was determined using the Wald test with Kenward-Roger (34) degrees of freedom and an alpha level of 0.05. Age-specific estimates, 95% confidence intervals, F-statistics, and p-values were reported for each association of interest. Graphs provided model-based estimates for each genotype group.
Past studies have identified sex-specific differences in BMI across development (22,35). Thus, as a sensitivity analysis, we fit separate models for each GLP-1R polymorphism to assess whether sex modified the association between GLP-1R genotype and BMI across childhood and adolescence via two- and three-way interactions between sex, GLP-1R polymorphism, and age polynomial terms.
Markers of insulin sensitivity, insulin secretion, and beta cell function
Separate general linear models were fit to assess the associations between each GLP-1R genotype and each marker of glucose-insulin homeostasis. Outcomes included glucose and insulin measured at baseline, 30, and 120 minutes across OGTT; the natural log of HOMA2-IR; the natural log of HOMA2-%B; insulinogenic index; and oral disposition index. Models were adjusted with an indicator variable for female sex, age at EPOCH visit 2, and the first three genetic principal components. Model assumptions were tested using jackknife studentized residuals. An alpha level of 0.05 was used to determine significance. Beta estimates and 95% confidence intervals were reported for each association. Results for homeostasis model of assessment variables were back-transformed by exponentiation.
As a sensitivity analyses, we fit separate models to assess whether sex modified the association between GLP-1R genotype and each marker of glucose-insulin homeostasis. Modification was tested via a two-way interaction between sex and the GLP-1R polymorphism of interest.
Results
Genotyping data were available for 464 EPOCH participants. Participant characteristics are shown in Table 2. All participants with genotyping data had at least two measurements of BMI. Approximately 50% of participants were female. The majority of youth self-identified as non-Hispanic White (54%), followed by Hispanic (35%), non-Hispanic Black (7%), and non-Hispanic Other (4%). Approximately 70% of participants with genotyping data completed the OGTT at visit 2 and had complete data for all covariates (N = 333). There were no significant differences in participant characteristics between the overall EPOCH cohort (N = 593); the subset of participants with genotype and BMI data (N = 464); or the subset of participants with genotype, adolescent OGTT, or covariate data (N = 333). Dataset comparisons not shown.
Table 2:
Characteristics of EPOCH participants who have genotyping data and have at least one outcome measurement. Continuous variables are presented as mean and standard deviation, with comparisons test using the t-test. Categorical variables are presented as number and percent. Unless otherwise noted, comparisons of categorical variables were conducted using the chi-square test.
| rs10305420 | rs6923761 | rs1042044 | |||||||
|---|---|---|---|---|---|---|---|---|---|
| CT or TT | CC | p | GA or AA | GG | p | CA or AA | CC | p | |
| 261 | 203 | 216 | 248 | 312 | 152 | ||||
| Female, n (%) | 132 (51%) | 100 (49%) | 0.78 | 95 (44%) | 137 (55%) | 0.016 | 164 (53%) | 68 (45%) | 0.11 |
| Race/Ethnicity,* n(%) | <0.0001 | < 0.0001 | 0.029 | ||||||
| Non-Hispanic White | 158 (61%) | 93 (46%) | 140 (65%) | 111 (45%) | 180 (58%) | 71 (47%) | |||
| Hispanic | 87 (33%) | 74 (37%) | 66 (31%) | 95 (38%) | 95 (31%) | 66 (43%) | |||
| Non-Hispanic Black | 6 (2%) | 25 (12%) | 5 (2%) | 26 (11%) | 20 (6%) | 11 (7%) | |||
| Non-Hispanic Other | 10 (4%) | 11 (5%) | 5 (2%) | 16 (6%) | 17 (5%) | 4 (3%) | |||
| Ancestral PCs | |||||||||
| PC1 | 0.0084 | −0.0041 | <0.0001 | 0.0096 | −0.0027 | < 0.0001 | 0.0030 | 0.0030 | > 0.99 |
| PC2 | −0.0032 | 0.0000 | 0.22 | −0.0052 | 0.0011 | 0.014 | −0.0035 | 0.0017 | 0.059 |
| PC3 | −0.0004 | −0.0002 | 0.95 | −0.0006 | −0.0001 | 0.84 | −0.0009 | 0.0008 | 0.51 |
| EPOCH visit 2 | |||||||||
| Age, years | 16.6 (1.2) | 16.7 (1.2) | 0.66 | 16.7 (1.2) | 16.7 (1.2) | 0.94 | 16.7 (1.2) | 16.6 (1.2) | 0.72 |
| Overweight/Obesity, n(%) | 54 (26%) | 51 (33%) | 0.18 | 43 (25%) | 62 (32%) | 0.14 | 77 (31%) | 28 (25%) | 0.27 |
Due to small cell sizes, group comparisons of race/ethnicity were conducted using the Fisher’s Exact test.
Abbreviations: PC: Genetic Ancestral Principal Component
Over a third of the EPOCH population was homozygous major for at least one of the three polymorphisms: 44% carried genotype CC at rs10305420, 53% carried genotype GG at rs6923761, and 33% carried genotype CC at rs1042044. We observed differences in the proportion of participants who were homozygous for the major allele across all self-identified racial and ethnic groups. Among participants who self-identified as Non-Hispanic White, there was a higher proportion of youth with at least one minor allele.
rs10305420
Parameter estimates for the model of BMI growth appear in Table 3. Rs10305420 was not associated with BMI growth across childhood and adolescence (p > 0.05).
Table 3:
Model parameters for the association between rs10305420 and BMI growth across childhood and adolescence.
| Parameter | Estimate | 95% CI | p-value |
|---|---|---|---|
| Intercept | 18.45 | 17.88, 19.02 | < 0.0001 |
| Age | −1.16 | −1.33, −1.00 | < 0.0001 |
| Age2 | 0.17 | 0.15, 0.19 | < 0.0001 |
| Age3 | −0.0046 | −0.005, −0.004 | < 0.0001 |
| SNP | 0.23 | −0.24, 0.70 | 0.33 |
| SNP × age | −0.092 | −0.17, −0.014 | 0.021 |
| Female | −0.41 | −0.80, −0.033 | 0.033 |
| PC1 | −4.64 | −12.54, 3.26 | 0.25 |
| PC2 | 5.36 | −1.63, 12.36 | 0.13 |
| PC3 | 7.28 | −0.57, 15.13 | 0.069 |
| p-value for SNP effecta | 0.067 |
A multiple degree of freedom chunk test was conducted to determine whether rs10305420 was associated with BMI over time: F = 2.72, ndf = 2, ddf = 404, p = 0.067.
GLP-1R polymorphism rs10305420 was associated with differential responses to OGTT for glucose and insulin (Table 4, Fig. 2). There were significant associations between rs10305420 and insulin levels at all three OGTT timepoints. When compared to minor allele carriers, participants who were homozygous for the major allele had higher insulin levels at baseline [β = 2.7 uIU/mL (95% CI: 0.4, 5.0)], 30-minutes [β = 27.6 uIU/mL (3.0, 52.2)], and 120-minutes [β = 18.94 uIU/mL (95% CI: 1.04, 36.84)]. While glucose levels were similar at baseline and 120 minutes, participants who were homozygous for the major allele had marginally higher, though statistically non-significant, glucose levels at 30-minutes as compared to minor allele carriers [β = 7.12 mg/dL (95% CI: 0.00, 14.24), p = 0.050]. Sex did not modify any of these associations (p > 0.05).
Table 4:
Association of GLP-1R polymorphisms rs10305420, rs1042044, and rs6923761 with surrogate measures of glucose and insulin traits derived from OGTT. Models were adjusted for female sex, age at OGTT, and the first three ancestral principal components.
| rs10305420 | rs6923761 | rs1042044 | ||||
|---|---|---|---|---|---|---|
| Beta (95% CI) | P | Beta (95% CI) | P | Beta (95% CI) | P | |
| Fasting | 2.38 (−1.85, 6.61) | 0.27 | 2.20 (−2.02, 6.42) | 0.31 | −2.72 (−7.10, 1.66) | 0.22 |
| 30-minutes | 7.12 (0.00, 14.24) | 0.050 | 0.068 (−7.07, 7.21) | 0.99 | −3.80 (−11.20, 3.61) | 0.31 |
| 120-minutes | 4.19 (−3.11, 11.49) | 0.26 | 0.12 (−7.16, 7.40) | 0.97 | −1.99 (−9.51, 5.52) | 0.60 |
| Insulin (uIU/mL) | ||||||
| Fasting | 2.68 (0.37, 4.99) | 0.023 | 1.74 (−0.58, 4.06) | 0.14 | 0.23 (−2.18, 2.64) | 0.85 |
| 30-minutes | 27.62 (3.00, 52.24) | 0.028 | 17.52 (−7.14, 42.17) | 0.16 | −1.98 (−27.68, 23.72) | 0.88 |
| 120-minutes | 18.94 (1.04, 36.84) | 0.038 | 1.45 (−16.50, 19.40) | 0.87 | 15.91 (−2.52, 34.35) | 0.090 |
| HOMA2-%IRǂ | 1.19 (1.06, 1.32) | 0.0032 | 1.13 (1.01, 1.26) | 0.033 | 0.97 (0.86, 1.08) | 0.56 |
| HOMA2-%Bǂ | 1.09 (0.01, 0.16) | 0.027 | 1.04 (0.96, 1.13) | 0.29 | 1.02 (0.94, 1.11) | 0.69 |
| Insulinogenic index | 0.08 (−0.86, 1.02) | 0.86 | 0.23 (−0.71, 1.16) | 0.64 | 0.16 (−0.81, 1.14) | 0.74 |
| Oral disposition index | 0.55 (−1.43, 2.54) | 0.58 | 1.59 (−0.38, 3.56) | 0.11 | 0.07 (−1.99, 2.13) | 0.95 |
Data were transformed for statistical modeling. Beta estimates and 95% confidence intervals were exponentiated for this table.
Fig. 2:
Glucose and insulin levels across oral glucose tolerance testing are different by rs1030420 genotype.
GLP-1R polymorphism rs10305420 was associated with markers of glucose-insulin homeostasis during the fasting state, but not during the early post-prandial state (Table 4). Compared to minor allele carriers, participants who were homozygous for the major allele had higher estimated insulin resistance (HOMA2-IR) and insulin secretion (HOMA2-%B), with, on average, 1.19% (1.06, 1.32) higher HOMA2-IR and 1.09% (95% CI: 1.01, 1.17) higher HOMA2-%B. GLP-1R polymorphism rs10305420 was not significantly associated with insulinogenic index or oral disposition index (p > 0.05). Sex did not modify any of these associations (p > 0.05).
rs6923761
Parameter estimates for the model of BMI growth appear in Table 5. GLP-1R polymorphism rs6923761 was associated with BMI growth across childhood and adolescence. There was a significant difference between participants who were homozygous for the major allele (genotype GG) and minor allele carriers (genotype GA or AA) in the BMI trajectory (p = 0.0037) and in the velocity of BMI growth (p = 0.0014). Participants who were homozygous for the major allele had significantly faster BMI velocity during childhood, but not adolescence, when compared to minor allele carriers [Average velocity (95% CI): age 10 / genotype GG: 0.86 kg/m2●year (0.80, 0.92), age 10 / genotype GA or AA: 0.71 kg/m2●year (0.64, 0.77), age 16 / genotype GG: 0.67 kg/m2●year (0.59, 0.75), age 16 / genotype GA or AA: 0.69 kg/m2●year (0.60, 0.77)]. Compared to carriers of the minor allele, participants who were homozygous for the major allele had higher BMI by age 10 [average difference (95% CI): 0.74 kg/m2 (0.07, 1.40); p = 0.030], with continued divergence through age 16 [average difference (95% CI): 1.29 kg/m2 (0.22, 2.37); p = 0.019]. Model predicted BMI trajectories are shown in Fig. 3. Sex did not modify the association between rs6923761 and BMI growth (p > 0.05 for all terms).
Table 5:
Model parameters for the association between rs6923761 and BMI growth across childhood and adolescence.
| Parameter | Estimate | 95% CI | p-value |
|---|---|---|---|
| Intercept | 18.68 | 17.99, 19.37 | < 0.0001 |
| Age | −1.30 | −1.53, −1.08 | < 0.0001 |
| Age2 | 0.19 | 0.16, 0.21 | < 0.0001 |
| Age3 | −0.0052 | −0.0060, −0.0044 | < 0.0001 |
| SNP | −0.23 | −1.19, 0.72 | 0.63 |
| SNP × age | 0.19 | −0.14, 0.51 | 0.26 |
| SNP × age2 | −0.037 | −0.071, −0.0031 | 0.033 |
| SNP × age3 | 0.0013 | 0.00022, 0.0024 | 0.019 |
| Female | −0.41 | −0.79, −0.025 | 0.037 |
| PC1 | −5.12 | −12.97, 2.73 | 0.070 |
| PC2 | 5.57 | −1.45, 12.60 | 0.12 |
| PC3 | 7.27 | −0.57, 15.12 | 0.069 |
| p-value for SNP effecta | 0.0037 |
A multiple degree of freedom chunk test was conducted to determine whether rs6923761 was associated with BMI over time: F = 3.92, ndf = 4, ddf = 1080, p = 0.0037.
Fig. 3:
BMI growth differs across childhood and adolescence by GLP-1R polymorphism rs6923761 genotype status.
GLP-1R polymorphism rs6923761 was associated with higher estimated insulin resistance (HOMA2-IR), but no other markers of glucose-insulin homeostasis (Table 4). When compared to minor allele carriers, participants who were homozygous for the major allele were more insulin resistant, with 1.13% (95%: 1.01, 1.26, p = 0.033) higher HOMA2-IR. Sex did not modify the association between rs6923761 and HOMA2-IR (p > 0.05).
rs1042044
Parameter estimates for the model of BMI growth appear in Table 6. GLP-1R polymorphism rs1042044 was not significantly associated with BMI growth across childhood and adolescence (p > 0.05) or with markers of glucose-insulin homeostasis (all p > 0.05, Table 4).
Table 6:
Model parameters for the association between rs1042044 and BMI growth across childhood and adolescence.
| Parameter | Estimate | 95% CI | p-value |
|---|---|---|---|
| Intercept | 18.56 | 17.96, 19.16 | <0.0001 |
| Age | −1.20 | −1.37, −1.03 | <0.0001 |
| Age2 | 0.17 | 0.15, 0.19 | <0.0001 |
| Age3 | −0.0046 | −0.0051, −0.0040 | < 0.0001 |
| SNP | 0.01S | −0.47, 0.50 | 0.94 |
| SNP × age | −0.016 | −0.099, 0.067 | 0.71 |
| Female | −0.41 | −0.80, −0.03 | 0.034 |
| PC1 | −4.89 | −12.53, 2.75 | 0.21 |
| PC2 | 5.39 | −1.64, 12.41 | 0.13 |
| PC3 | 7.19 | −0.66, 15.04 | 0.073 |
| p-value for SNP effecta | 0.92 |
A multiple degree of freedom chunk test was conducted to determine whether rs1042044 was associated with BMI over time: F = 0.08, ndf = 2, ddf = 398, p = 0.92.
Discussion
Summary
To our knowledge, this is the first study to assess whether there are early life phenotypic differences in BMI or glucose-insulin homeostasis associated with GLP-1R gene variant differences. We are also the first to study if variant genotypes of the GLP-1R gene are related to differential rates of anthropometric growth across child development. Among a diverse population of youth, we found associations of GLP-1R gene polymorphisms rs10305420 and rs6923761 with BMI growth and markers of glucose-insulin homeostasis in early life. Participants who were homozygous for the major allele at rs6923761 (i.e., genotype GG) displayed differences in BMI across childhood and adolescence. While average BMI was similar between genotype categories in early childhood, participants who were homozygous for the major allele had higher average BMI beginning in middle childhood with increasing divergence throughout adolescence. For both rs10305420 and rs6923761, participants who were homozygous for the major allele (i.e. genotypes CC and GG, respectively) had lower estimated insulin sensitivity during adolescence. Additionally, participants who were homozygous for the rs10305420 major allele had higher compensatory insulin secretion in the fasting state and across OGTT. These findings suggest that GLP-1R gene variation is associated with observable differences in anthropometric characteristics, insulin sensitivity and insulin secretion as early as middle childhood. These results should be interpreted with caution, however, since we do not know if these polymorphisms are causally associated with these outcomes, or if they are simply tagging functional variants.
GLP-1R polymorphisms and childhood BMI growth
Our results suggest that GLP-1R polymorphisms rs6923761 is associated with BMI and BMI growth as early as middle childhood. These findings are important, because higher BMI and BMI growth in middle childhood have been associated with an increased risk of lifelong excess adiposity (36–38). For rs6923761, participants who were homozygous for the major allele had faster rates of BMI growth across childhood and had different BMI trajectories across childhood and adolescence, when compared to minor allele carriers.
Our findings in children mirror those of de Luis et al. (5,9,11,12,16,39,40), who studied associations between rs6923761 and anthropometrics among adults from Spain. Similar to our findings, de Luis et al. showed that adults who were homozygous for the major allele had higher insulin resistance (5), and higher BMI, body weight, fat mass, and waist circumference (9,11,12) than adults who carried at least one minor allele. There are no similar studies in youth.
GLP-1R polymorphisms and glucose-insulin homeostasis traits
Participants who were homozygous for the rs10305420 major allele had differences in markers of glucose-insulin homeostasis, when compared to minor allele carriers. These participants were less insulin sensitive, and, subsequently, secreted more insulin during the fasting state. In a glucose-stimulated state, participants who were homozygous for the major allele had higher insulin concentrations at 30- and 120-minutes post-challenge. However, there were not significant differences in insulinogenic index or oral disposition index.
We had several null results, which may reflect true lack of associations, or lack of power. The results will have to be replicated in other studies. If our null results hold, they can be interpreted as follows. Null results for insulinogenic index suggest that although participants who are homozygous for the rs10305420 major allele are less insulin sensitive, their glucose-stimulated insulin secretion is comparable to carriers of the minor allele. Null results for oral disposition index suggest that although participants who are homozygous for the rs10305420 major allele require higher levels of insulin secretion to compensate for their reduced insulin sensitivity, their beta cells are still healthy enough to meet the increased demands. The fact that these participants had slightly higher, though statistically non-significant, glucose levels at 30-minutes post-glucose load suggests that, over time, the persistent stress of higher insulin secretory demand may lead to observable differences in oral disposition index and higher fasting and stimulated glucose levels. This finding aligns with previous research from Vujkovic et al. (15), who noted that the major allele for rs10305420 was associated with increased risk for type 2 diabetes among adults.
While it is possible that the findings may reflect false positive associations, we did not correct for multiple comparisons for the following two reasons. First, we only considered three sets hypotheses, rather than millions. Second, the three sets of hypotheses were based on SNPs identified a priori, based on evidence from the literature that all three of the polymorphisms were associated with either obesity, markers of glucose-insulin homeostasis, or both.
Participants who were homozygous for the rs6923761 major allele had lower estimated insulin sensitivity, but the polymorphism was not associated with any other markers of glucose-insulin homeostasis. These findings align with those of de Luis et al. (5) who found that adults who were homozygous for the rs6923761 major allele had higher insulin resistance, based on HOMA1-IR. On the other hand, a study of 53 adult participants from Europe (41) reported that participants who were homozygous for the rs6923761 major allele had higher post-prandial glucose levels after a mixed meal. It is possible that the differences in this result stem from differences in metabolism related to glucose ingestion versus a mixed meal. Additionally, it is possible that while youth who are homozygous for the rs6923761 major allele are more insulin resistant than their peers, they still have the capability to elicit a sufficient compensatory insulin secretion to maintain normoglycemia (42).
GLP-1R polymorphism rs1042044
We did not find significant associations between GLP-1R polymorphism rs1042044 and anthropometrics or glucose-insulin homeostasis. This is contrary to studies in adults. Though the results did not withstand multiple testing corrections, a study from the UK biobank (3) provided evidence that people who were homozygous for the rs1042044 major allele had higher BMI and glycated hemoglobin as compared to people who carried the minor allele. Among children from the EPOCH cohort, participants who were homozygous for the major allele had higher average BMI than participants who carried the minor allele, but the difference between genotype categories was not significant. It is worth noting, that the UK Biobank study utilized data from over 300,000 participants. Our study was much smaller (N = 464), so it is possible that we did not have adequate power to detect these more subtle differences. Consequently, rs1042044 needs to be studied further in a larger cohort of youth.
Strengths and Limitations
The study had several strengths. First, our study collected repeated measurements of height and weight from childhood through late adolescence. Since most participants had nine or more measurements of BMI across childhood and adolescence, we were able to characterize detailed trajectories of BMI across development. Second, we used results from OGTT during adolescence to derive indices reflective of glucose-insulin homeostasis, including insulin sensitivity and secretion.
There are several limitations of this study. First, we imputed genotypes for rs10305420 because we did not have direct genotyping available. However, rs10305420 had high imputation quality with R2 > 0.95 on both arrays, indicating that it is was well-imputed (43,44). Second, we have a relatively small sample size (N = 464) and results from only one OGTT. It is possible that we did not have enough power to detect all true associations between the three GLP-1R polymorphisms considered and markers of glucose-insulin homeostasis. Third, this project assessed associations of each GLP-1R polymorphism independently, but not in combination, thereby precluding assessments of potential synergistic effects. However, we felt that assessing each polymorphism singly was an important first step given the limited literature on these polymorphisms in relation to glycemia in youth, and because the three polymorphisms are in linkage equilibrium. Future studies are required to assess the combined effects of the three polymorphisms. Finally, it is possible that the GLP-1R polymorphisms are not causal. The results may indicate instead associations with true causal variants through linkage disequilibrium.
Conclusion
We have shown that common polymorphisms of the GLP-1R gene are associated with differences in BMI growth, estimated insulin sensitivity, and insulin secretion in youth. These findings suggest that GLP-1R variants should be considered when studying genetic associations with body size and glucose-insulin metabolism. Larger studies of pediatric populations with GLP-1R sequencing data and repeated oral glucose tolerance tests are needed to replicate and refine these findings. If further studies show that these GLP-1R polymorphisms are causal variants, it would suggest the utility of genotyping people for GLP-1R to promote early intervention. If further studies show that these GLP-1R polymorphisms are not causal variants, it would suggest that further studies must be considered to find the true causal variants.
Supplementary Material
Acknowledgements
We thank the participants and families who took part in the EPOCH study. The secondary data from the EPOCH study was collected during R01DK068001, for which DD was the principal investigator. A version of this paper was submitted to the University of Colorado Denver in partial fulfillment of the requirements for the PhD in Epidemiology for Dr. Kylie K. Harrall. The views expressed in this manuscript cannot be inferred to be those of the NHLBI or the NIH.
Footnotes
Disclosure summary: The authors have no conflicts of interest related to the research presented in this article.
Data availability
De-identified data from the EPOCH study may be made available, upon reasonable request, and after review of EPOCH publications committee.
Sources Cited
- 1.Wessel J, Chu AY, Willems SM, Wang S, Yaghootkar H, Brody JA, Dauriz M, Hivert M-F, Raghavan S, Lipovich L, Hidalgo B, Fox K, Huffman JE, An P, Lu Y, Rasmussen-Torvik LJ, Grarup N, Ehm MG, Li L, Baldridge AS, Stančáková A, Abrol R, Besse C, Boland A, Bork-Jensen J, Fornage M, Freitag DF, Garcia ME, Guo X, Hara K, Isaacs A, Jakobsdottir J, Lange LA, Layton JC, Li M, Hua Zhao J, Meidtner K, Morrison AC, Nalls MA, Peters MJ, Sabater-Lleal M, Schurmann C, Silveira A, Smith AV, Southam L, Stoiber MH, Strawbridge RJ, Taylor KD, Varga TV, Allin KH, Amin N, Aponte JL, Aung T, Barbieri C, Bihlmeyer NA, Boehnke M, Bombieri C, Bowden DW, Burns SM, Chen Y, Chen Y-D, Cheng C-Y, Correa A, Czajkowski J, Dehghan A, Ehret GB, Eiriksdottir G, Escher SA, Farmaki A-E, Frånberg M, Gambaro G, Giulianini F, Goddard WA, Goel A, Gottesman O, Grove ML, Gustafsson S, Hai Y, Hallmans G, Heo J, Hoffmann P, Ikram MK, Jensen RA, Jørgensen ME, Jørgensen T, Karaleftheri M, Khor CC, Kirkpatrick A, Kraja AT, Kuusisto J, Lange EM, Lee IT, Lee W-J, Leong A, Liao J, Liu C, Liu Y, Lindgren CM, Linneberg A, Malerba G, Mamakou V, Marouli E, Maruthur NM, Matchan A, McKean-Cowdin R, McLeod O, Metcalf GA, Mohlke KL, Muzny DM, Ntalla I, Palmer ND, Pasko D, Peter A, Rayner NW, Renström F, Rice K, Sala CF, Sennblad B, Serafetinidis I, Smith JA, Soranzo N, Speliotes EK, Stahl EA, Stirrups K, Tentolouris N, Thanopoulou A, Torres M, Traglia M, Tsafantakis E, Javad S, Yanek LR, Zengini E, Becker DM, Bis JC, Brown JB, Cupples LA, Hansen T, Ingelsson E, Karter AJ, Lorenzo C, Mathias RA, Norris JM, Peloso GM, Sheu WH-H, Toniolo D, Vaidya D, Varma R, Wagenknecht LE, Boeing H, Bottinger EP, Dedoussis G, Deloukas P, Ferrannini E, Franco OH, Franks PW, Gibbs RA, Gudnason V, Hamsten A, Harris TB, Hattersley AT, Hayward C, Hofman A, Jansson J-H, Langenberg C, Launer LJ, Levy D, Oostra BA, O’Donnell CJ, O’Rahilly S, Padmanabhan S, Pankow JS, Polasek O, Province MA, Rich SS, Ridker PM, Rudan I, Schulze MB, Smith BH, Uitterlinden AG, Walker M, Watkins H, Wong TY, Zeggini E, EPIC-InterAct Consortium, Laakso M, Borecki IB, Chasman DI, Pedersen O, Psaty BM, Tai ES, van Duijn CM, Wareham NJ, Waterworth DM, Boerwinkle E, Kao WHL, Florez JC, Loos RJF, Wilson JG, Frayling TM, Siscovick DS, Dupuis J, Rotter JI, Meigs JB, Scott RA, Goodarzi MO. Low-frequency and rare exome chip variants associate with fasting glucose and type 2 diabetes susceptibility. Nat Commun 2015;6:5897. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kwak SH, Chae J, Lee S, Choi S, Koo BK, Yoon JW, Park J-H, Cho B, Moon MK, Lim S, Cho YM, Moon S, Kim YJ, Han S, Hwang MY, Cho YS, Lee M-S, Jang HC, Kang HM, Park T, Cho NH, Kim K, Kim J-I, Park KS. Nonsynonymous Variants in PAX4 and GLP1R Are Associated With Type 2 Diabetes in an East Asian Population. Diabetes 2018;67(9):1892–1902. [DOI] [PubMed] [Google Scholar]
- 3.Ward J, Lyall LM, Strawbridge RJ, Stanciu I, Veldsman M, Garfield V, Celis-Morales C, Newby D, Stewart W, Pell JP, Sattar N, Lyall DM. Testing for association between exonic glucagon-like peptide 1 receptor mutation with physical and brain health traits in UK Biobank. Diabetes, Obesity and Metabolism 2023;25(2):623–627. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chen J, Spracklen CN, Marenne G, Varshney A, Corbin LJ, Luan J, Willems SM, Wu Y, Zhang X, Horikoshi M, Boutin TS, Mägi R, Waage J, Li-Gao R, Chan KHK, Yao J, Anasanti MD, Chu AY, Claringbould A, Heikkinen J, Hong J, Hottenga J-J, Huo S, Kaakinen MA, Louie T, März W, Moreno-Macias H, Ndungu A, Nelson SC, Nolte IM, North KE, Raulerson CK, Ray D, Rohde R, Rybin D, Schurmann C, Sim X, Southam L, Stewart ID, Wang CA, Wang Y, Wu P, Zhang W, Ahluwalia TS, Appel EVR, Bielak LF, Brody JA, Burtt NP, Cabrera CP, Cade BE, Chai JF, Chai X, Chang L-C, Chen C-H, Chen BH, Chitrala KN, Chiu Y-F, de Haan HG, Delgado GE, Demirkan A, Duan Q, Engmann J, Fatumo SA, Gayán J, Giulianini F, Gong JH, Gustafsson S, Hai Y, Hartwig FP, He J, Heianza Y, Huang T, Huerta-Chagoya A, Hwang MY, Jensen RA, Kawaguchi T, Kentistou KA, Kim YJ, Kleber ME, Kooner IK, Lai S, Lange LA, Langefeld CD, Lauzon M, Li M, Ligthart S, Liu J, Loh M, Long J, Lyssenko V, Mangino M, Marzi C, Montasser ME, Nag A, Nakatochi M, Noce D, Noordam R, Pistis G, Preuss M, Raffield L, Rasmussen-Torvik LJ, Rich SS, Robertson NR, Rueedi R, Ryan K, Sanna S, Saxena R, Schraut KE, Sennblad B, Setoh K, Smith AV, Sparsø T, Strawbridge RJ, Takeuchi F, Tan J, Trompet S, van den Akker E, van der Most PJ, Verweij N, Vogel M, Wang H, Wang C, Wang N, Warren HR, Wen W, Wilsgaard T, Wong A, Wood AR, Xie T, Zafarmand MH, Zhao J-H, Zhao W, Amin N, Arzumanyan Z, Astrup A, Bakker SJL, Baldassarre D, Beekman M, Bergman RN, Bertoni A, Blüher M, Bonnycastle LL, Bornstein SR, Bowden DW, Cai Q, Campbell A, Campbell H, Chang YC, de Geus EJC, Dehghan A, Du S, Eiriksdottir G, Farmaki AE, Frånberg M, Fuchsberger C, Gao Y, Gjesing AP, Goel A, Han S, Hartman CA, Herder C, Hicks AA, Hsieh C-H, Hsueh WA, Ichihara S, Igase M, Ikram MA, Johnson WC, Jørgensen ME, Joshi PK, Kalyani RR, Kandeel FR, Katsuya T, Khor CC, Kiess W, Kolcic I, Kuulasmaa T, Kuusisto J, Läll K, Lam K, Lawlor DA, Lee NR, Lemaitre RN, Li H, Lifelines Cohort Study, Lin S-Y, Lindström J, Linneberg A, Liu J, Lorenzo C, Matsubara T, Matsuda F, Mingrone G, Mooijaart S, Moon S, Nabika T, Nadkarni GN, Nadler JL, Nelis M, Neville MJ, Norris JM, Ohyagi Y, Peters A, Peyser PA, Polasek O, Qi Q, Raven D, Reilly DF, Reiner A, Rivideneira F, Roll K, Rudan I, Sabanayagam C, Sandow K, Sattar N, Schürmann A, Shi J, Stringham HM, Taylor KD, Teslovich TM, Thuesen B, Timmers PRHJ, Tremoli E, Tsai MY, Uitterlinden A, van Dam RM, van Heemst D, van Hylckama Vlieg A, van Vliet-Ostaptchouk JV, Vangipurapu J, Vestergaard H, Wang T, Willems van Dijk K, Zemunik T, Abecasis GR, Adair LS, Aguilar-Salinas CA, Alarcón-Riquelme ME, An P, Aviles-Santa L, Becker DM, Beilin LJ, Bergmann S, Bisgaard H, Black C, Boehnke M, Boerwinkle E, Böhm BO, Bønnelykke K, Boomsma DI, Bottinger EP, Buchanan TA, Canouil M, Caulfield MJ, Chambers JC, Chasman DI, Chen Y-DI, Cheng C-Y, Collins FS, Correa A, Cucca F, de Silva HJ, Dedoussis G, Elmståhl S, Evans MK, Ferrannini E, Ferrucci L, Florez JC, Franks PW, Frayling TM, Froguel P, Gigante B, Goodarzi MO, Gordon-Larsen P, Grallert H, Grarup N, Grimsgaard S, Groop L, Gudnason V, Guo X, Hamsten A, Hansen T, Hayward C, Heckbert SR, Horta BL, Huang W, Ingelsson E, James PS, Jarvelin M-R, Jonas JB, Jukema JW, Kaleebu P, Kaplan R, Kardia SLR, Kato N, Keinanen-Kiukaanniemi SM, Kim B-J, Kivimaki M, Koistinen HA, Kooner JS, Körner A, Kovacs P, Kuh D, Kumari M, Kutalik Z, Laakso M, Lakka TA, Launer LJ, Leander K, Li H, Lin X, Lind L, Lindgren C, Liu S, Loos RJF, Magnusson PKE, Mahajan A, Metspalu A, Mook-Kanamori DO, Mori TA, Munroe PB, Njølstad I, O’Connell JR, Oldehinkel AJ, Ong KK, Padmanabhan S, Palmer CNA, Palmer ND, Pedersen O, Pennell CE, Porteous DJ, Pramstaller PP, Province MA, Psaty BM, Qi L, Raffel LJ, Rauramaa R, Redline S, Ridker PM, Rosendaal FR, Saaristo TE, Sandhu M, Saramies J, Schneiderman N, Schwarz P, Scott LJ, Selvin E, Sever P, Shu X-O, Slagboom PE, Small KS, Smith BH, Snieder H, Sofer T, Sørensen TIA, Spector TD, Stanton A, Steves CJ, Stumvoll M, Sun L, Tabara Y, Tai ES, Timpson NJ, Tönjes A, Tuomilehto J, Tusie T, Uusitupa M, van der Harst P, van Duijn C, Vitart V, Vollenweider P, Vrijkotte TGM, Wagenknecht LE, Walker M, Wang YX, Wareham NJ, Watanabe RM, Watkins H, Wei WB, Wickremasinghe AR, Willemsen G, Wilson JF, Wong T-Y, Wu J-Y, Xiang AH, Yanek LR, Yengo L, Yokota M, Zeggini E, Zheng W, Zonderman AB, Rotter JI, Gloyn AL, McCarthy MI, Dupuis J, Meigs JB, Scott RA, Prokopenko I, Leong A, Liu C-T, Parker SCJ, Mohlke KL, Langenberg C, Wheeler E, Morris AP, Barroso I, Meta-Analysis of Glucose and Insulin-related Traits Consortium (MAGIC). The trans-ancestral genomic architecture of glycemic traits. Nat Genet 2021;53(6):840–860. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.de Luis DA, Ballesteros M, Lopez Guzman A, Ruiz E, Muñoz C, Penacho MA, Iglesias P, Maldonado A, Puigdevall V, Delgado M. rs6923761 gene variant in glucagon-like peptide 1 receptor: Allelic frequencies and influence on cardiovascular risk factors in a multicenter study of Castilla-Leon. Clin Nutr 2018;37(6 Pt A):2144–2148. [DOI] [PubMed] [Google Scholar]
- 6.Nishiya Y, Daimon M, Mizushiri S, Murakami H, Tanabe J, Matsuhashi Y, Yanagimachi M, Tokuda I, Sawada K, Ihara K. Nutrient consumption-dependent association of a glucagon-like peptide-1 receptor gene polymorphism with insulin secretion. Sci Rep 2020;10(1):16382. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sathananthan A, Man CD, Micheletto F, Zinsmeister AR, Camilleri M, Giesler PD, Laugen JM, Toffolo G, Rizza RA, Cobelli C, Vella A. Common genetic variation in GLP1R and insulin secretion in response to exogenous GLP-1 in nondiabetic subjects: a pilot study. Diabetes Care 2010;33(9):2074–2076. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Luo P, Fan Y, Xiong Y, Feng H, Yang Z, Zhang C, Mei B. Genetic variants of the GLP-1R gene affect the susceptibility and glucose metabolism of gestational diabetes mellitus: a two-center nested case‒control study. Diabetol Metab Syndr 2022;14(1):190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.de Luis DA, Aller R, Izaola O, de la Fuente B, Primo D, Conde R, Sagrado MG. Evaluation of weight loss and adipocytokine levels after two hypocaloric diets with different macronutrient distribution in obese subjects with the rs6923761 gene variant of glucagon-like peptide 1 receptor. Ann Nutr Metab 2013;63(4):277–282. [DOI] [PubMed] [Google Scholar]
- 10.Li P, Tiwari HK, Lin W-Y, Allison DB, Chung WK, Leibel RL, Yi N, Liu N. Genetic association analysis of 30 genes related to obesity in a European American population. Int J Obes (Lond) 2014;38(5):724–729. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.de Luis DA, Aller R, Izaola O, Lopez JJ, Gomez E, Torres B, Soto GD. Effect of rs6923761 gene variant of glucagon-like peptide 1 receptor on metabolic response and weight loss after a 3-month intervention with a hypocaloric diet. J Endocrinol Invest 2014;37(10):935–939. [DOI] [PubMed] [Google Scholar]
- 12.de Luis DA, Bachiller R, Izaola O, de la Fuente B, Aller R. Relation of the rs6923761 gene variant in glucagon-like peptide 1 receptor to metabolic syndrome in obese subjects. Ann Nutr Metab 2014;65(4):253–258. [DOI] [PubMed] [Google Scholar]
- 13.Xu T, Liu M, Liu Q, Wang B, Wang M, Qu M, Chen X, Wu J. Associations of TCF7L2 rs11196218 (A/G) and GLP-1R rs761386 (C/T) Gene Polymorphisms with Obesity in Chinese Population. Diabetes Metab Syndr Obes 2021;14:2465–2472. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Li Y, Yang Z, Ren S, Shen B, Zhang Y, Zong H, Li Y. Association between GLP-1R gene polymorphism and dyslipidemia in Chinese patients with type 2 diabetes mellitus: A case-control study. Gene 2023;878:147589. [DOI] [PubMed] [Google Scholar]
- 15.Vujkovic M, Keaton JM, Lynch JA, Miller DR, Zhou J, Tcheandjieu C, Huffman JE, Assimes TL, Lorenz K, Zhu X, Hilliard AT, Judy RL, Huang J, Lee KM, Klarin D, Pyarajan S, Danesh J, Melander O, Rasheed A, Mallick NH, Hameed S, Qureshi IH, Afzal MN, Malik U, Jalal A, Abbas S, Sheng X, Gao L, Kaestner KH, Susztak K, Sun YV, DuVall SL, Cho K, Lee JS, Gaziano JM, Phillips LS, Meigs JB, Reaven PD, Wilson PW, Edwards TL, Rader DJ, Damrauer SM, O’Donnell CJ, Tsao PS, HPAP Consortium, Regeneron Genetics Center, VA Million Veteran Program, Chang K-M, Voight BF, Saleheen D. Discovery of 318 new risk loci for type 2 diabetes and related vascular outcomes among 1.4 million participants in a multi-ancestry meta-analysis. Nat Genet 2020;52(7):680–691. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.de Luis DA, Aller R, Izaola O, Bachiller R, Pacheco D. Cardiovascular risk factors and adipocytokines levels after two hypocaloric diets with different fat distribution in obese subjects and rs6923761 gene variant of glucagon-like peptide 1 receptor. J Endocrinol Invest 2014;37(9):853–859. [DOI] [PubMed] [Google Scholar]
- 17.Sheikh HI, Dougherty LR, Hayden EP, Klein DN, Singh SM. Glucagon-like peptide-1 receptor gene polymorphism (Leu260Phe) is associated with morning cortisol in preschoolers. Prog Neuropsychopharmacol Biol Psychiatry 2010;34(6):980–983. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Díaz M, García-Beltran C, López-Bermejo A, de Zegher F, Ibáñez L. GLP-1 and IGF-I levels are elevated in late infancy in low birth weight infants, independently of GLP-1 receptor polymorphisms and neonatal nutrition. Int J Obes (Lond) 2018;42(4):915–918. [DOI] [PubMed] [Google Scholar]
- 19.Kelsey MM, Zeitler PS. Insulin Resistance of Puberty. Curr Diab Rep 2016;16(7):64. [DOI] [PubMed] [Google Scholar]
- 20.Sauder KA, Hockett CW, Ringham BM, Glueck DH, Dabelea D. Fetal overnutrition and offspring insulin resistance and β-cell function: the Exploring Perinatal Outcomes among Children (EPOCH) study. Diabet Med 2017;34(10):1392–1399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Hockett CW, Harrall KK, Moore BF, Starling AP, Bellatorre A, Sauder KA, Perng W, Scherzinger A, Garg K, Ringham BM, Glueck DH, Dabelea D. Persistent effects of in utero overnutrition on offspring adiposity: the Exploring Perinatal Outcomes among Children (EPOCH) study. Diabetologia 2019;62(11):2017–2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Hockett CW, Harrall KK, Glueck DH, Dabelea DM. Exposure to gestational diabetes and BMI trajectories through adolescence: Exploring Perinatal Outcomes in Children study. J Clin Endocrinol Metab 2023:dgad278. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Crume TL, Ogden L, West NA, Vehik KS, Scherzinger A, Daniels S, McDuffie R, Bischoff K, Hamman RF, Norris JM, Dabelea D. Association of exposure to diabetes in utero with adiposity and fat distribution in a multiethnic population of youth: the Exploring Perinatal Outcomes among Children (EPOCH) Study. Diabetologia 2011;54(1):87–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Stanislawski MA, Shaw J, Litkowski E, Lange EM, Perng W, Dabelea D, Lange LA. Genetic Risk for Hepatic Fat among an Ethnically Diverse Cohort of Youth: The Exploring Perinatal Outcomes among Children Study. The Journal of Pediatrics 2020;220:146–153.e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Stanislawski MA, Litkowski E, Raghavan S, Harrall KK, Shaw J, Glueck DH, Lange EM, Dabelea D, Lange LA. Genetic Risk Score for Type 2 Diabetes and Traits Related to Glucose-Insulin Homeostasis in Youth: The Exploring Perinatal Outcomes Among Children (EPOCH) Study. Diabetes Care 2021;44(9):2018–2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Phan L, Jin Y, Zhang H, Qiang W, Shekhtman E, Shao D, Revoe D, Villamarin R, Ivanchenko E, Kimura M, Wang Z, Hao L, Sharopova N, Bihan M, Sturcke A, Lee M, Popova N, Wu W, Bastiani C, Ward M, Holmes J, Lyoshin V, Kaur K, Moyer E, Feolo M, Kattman B. ALFA: Allele Frequency Aggregator. 2020. Available at:https://www.ncbi.nlm.nih.gov/snp/docs/gsr/alfa/.Accessed November 1, 2023. [Google Scholar]
- 27.de Luis DA, Diaz Soto G, Izaola O, Romero E. Evaluation of weight loss and metabolic changes in diabetic patients treated with liraglutide, effect of RS 6923761 gene variant of glucagon-like peptide 1 receptor. J Diabetes Complications 2015;29(4):595–598. [DOI] [PubMed] [Google Scholar]
- 28.Harrall KK, Bird SM, Muller KE, Vanderlinden LA, Payton ME, Bellatorre A, Dabelea D, Glueck DH. A better performing algorithm for identification of implausible growth data from longitudinal pediatric medical records. Sci Rep 2024;14:18276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Crume TL, Ogden L, Daniels S, Hamman RF, Norris JM, Dabelea D. The Impact of In Utero Exposure to Diabetes on Childhood Body Mass Index Growth Trajectories: The EPOCH Study. The Journal of Pediatrics 2011;158(6):941–946. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Wallace TM, Levy JC, Matthews DR. Use and abuse of HOMA modeling. Diabetes Care 2004;27(6):1487–1495. [DOI] [PubMed] [Google Scholar]
- 31.Tura A, Kautzky-Willer A, Pacini G. Insulinogenic indices from insulin and C-peptide: comparison of beta-cell function from OGTT and IVGTT. Diabetes Res Clin Pract 2006;72(3):298–301. [DOI] [PubMed] [Google Scholar]
- 32.Caprio S The Oral Disposition Index: A Valuable Estimate of β-Cell Function in Obese Youth. The Journal of Pediatrics 2012;161(1):3–4. [DOI] [PubMed] [Google Scholar]
- 33.Flanagin A, Frey T, Christiansen SL, AMA Manual of Style Committee. Updated Guidance on the Reporting of Race and Ethnicity in Medical and Science Journals. JAMA 2021;326(7):621–627. [DOI] [PubMed] [Google Scholar]
- 34.Kenward MG, Roger JH. An improved approximation to the precision of fixed effects from restricted maximum likelihood. Computational Statistics & Data Analysis 2009;53(7):2583–2595. [Google Scholar]
- 35.Bekelman TA, Ringham BM, Sauder KA, Johnson SL, Harrall KH, Glueck DH, Dabelea D. Adherence to index-based dietary patterns in childhood and BMI trajectory during the transition to adolescence: the EPOCH study. Int J Obes (Lond) 2021;45(11):2439–2446. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Guo SS, Wu W, Chumlea WC, Roche AF. Predicting overweight and obesity in adulthood from body mass index values in childhood and adolescence. Am J Clin Nutr 2002;76(3):653–658. [DOI] [PubMed] [Google Scholar]
- 37.Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH. Predicting obesity in young adulthood from childhood and parental obesity. N Engl J Med 1997;337(13):869–873. [DOI] [PubMed] [Google Scholar]
- 38.Freedman DS, Khan LK, Serdula MK, Dietz WH, Srinivasan SR, Berenson GS. The relation of childhood BMI to adult adiposity: the Bogalusa Heart Study. Pediatrics 2005;115(1):22–27. [DOI] [PubMed] [Google Scholar]
- 39.de Luis DA, Pacheco D, Aller R, Izaola O. Role of the rs6923761 gene variant in glucagon-like peptide 1 receptor gene on cardiovascular risk factors and weight loss after biliopancreatic diversion surgery. Ann Nutr Metab 2014;65(4):259–263. [DOI] [PubMed] [Google Scholar]
- 40.de Luis DA, Aller R, Izaola O, Romero E. Effects of a high-protein/low-carbohydrate versus a standard hypocaloric diet on adipocytokine levels and cardiovascular risk factors during 9 months, role of rs6923761 gene variant of glucagon-like peptide 1 receptor. J Endocrinol Invest 2015;38(11):1183–1189. [DOI] [PubMed] [Google Scholar]
- 41.Mashayekhi M, Wilson JR, Jafarian-Kerman S, Nian H, Yu C, Shuey MM, Luther JM, Brown NJ. Association of a glucagon-like peptide-1 receptor gene variant with glucose response to a mixed meal. Diabetes Obes Metab 2021;23(1):281–286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Lee JM. Why young adults hold the key to assessing the obesity epidemic in children. Arch Pediatr Adolesc Med 2008;162(7):682–687. [DOI] [PubMed] [Google Scholar]
- 43.Browning BL, Browning SR. A Unified Approach to Genotype Imputation and Haplotype-Phase Inference for Large Data Sets of Trios and Unrelated Individuals. Am J Hum Genet 2009;84(2):210–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Zheng H-F, Rong J-J, Liu M, Han F, Zhang X-W, Richards JB, Wang L. Performance of Genotype Imputation for Low Frequency and Rare Variants from the 1000 Genomes. PLoS One 2015;10(1):e0116487. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
De-identified data from the EPOCH study may be made available, upon reasonable request, and after review of EPOCH publications committee.