FTO Genotype and 2-Year Change in Body Composition and Fat Distribution in Response to Weight-Loss Diets

The POUNDS LOST Trial

Xiaomin Zhang; Qibin Qi; Cuilin Zhang; Frank B. Hu; Frank M. Sacks; Lu Qi

Disclosures

Diabetes. 2012;61(11):3005-3011. 

In This Article

Results

Characteristics of Study Population

Baseline characteristics of participants according to the FTO rs1558902 genotype are presented in Table 1. The minor allele frequency (MAF; A allele) was 0.402 in the total population. The genotype frequencies were significantly different among the sexes and ethnicities. After adjustment for age, sex, and ethnicity, all variables such as weight, BMI, WC, body composition, and fat distribution had no association with genotype at baseline. Baseline characteristics were similar among participants in the four diet groups (Supplementary Table 1). Likewise, no associations between the FTO genotypes and these measures were observed in the white participants (data not shown).

Effects of FTO rs1558902 Genotype on Weight and Waist: Overall and Two-factorial Analysis

After adjustment for age, sex, ethnicity, baseline BMI, and diet groups, no main effects of the FTO rs1558902 genotype on changes in weight or WC were found in any participants at 6 months and 2 years (data not shown). We next examined the genetic effects on changes in weight and WC following a two-factorial design: low versus high fat and low versus high protein. We found that the risk allele (A) was significantly associated with a 1.51-kg greater weight loss in the high-protein group (P = 0.010), but not in the low-protein group, by the end of intervention (2 years). The changes in weight and WC were less significant at 6 months (Table 2). In subgroups treated by different proportions of dietary fat, we did not find significant genetic effects on changes in weight and WC (all P > 0.05; Supplementary Table 2).Similarly, in the white participants, we found the risk allele was associated with a 1.38-kg greater weight loss in the high-protein group at 2 years (P = 0.028), but not in other subgroups (data not shown).

The FTO rs1558902 Genotype and Changes in Body Composition by Diet Intervention

Consistent with the observations of change in body weight, we found that the rs15589002 risk allele (A) was associated with greater loss of total fat, FFM, FM%, and percentage of trunk fat at 2 years in the high-protein group, but not in the low-protein group (Table 2). Tests for genotype-diet protein interaction were significant on changes in FFM and FM% (for interactions, P = 0.034 and 0.049, respectively) adjusted for age, sex, ethnicity, carbohydrate, baseline BMI, and the baseline value of body composition (Fig. 1). At 6 months, we only observed gene by protein diet interaction on changes in FFM (P = 0.008 for interaction; Fig. 1). The risk allele carriers in the high-protein group had greater loss of FFM than noncarriers, but those in the low-protein group had less loss of FFM compared with noncarriers (Table 2).

Figure 1.

Interaction between the FTO rs1558902 genotype and dietary protein intervention on changes in total fat (A), FFM (B), FM% (C), and percentage of trunk fat (D) at 6 months and 2 years. P values are adjusted for age, sex, ethnicity, carbohydrate, baseline values for respective outcomes, and baseline BMI. Data included 52 and 60 (TT), 66 and 73 (TA), and 31 and 28 (AA) participants in the low-protein group and the high-protein group at 6 months, respectively (total n = 310), and 34 and 44 (TT), 46 and 61 (TA), and 19 and 20 (AA) participants in the low-protein group and the high-protein group at 2 years, respectively (total n = 224).

We did not find significant genetic effect and interactions between the FTO variant and dietary fat intake on changes in body composition in total participants (Supplementary Table 2 and Supplementary Fig. 1). The results in the white participants were similar (data not shown).

The FTO rs1558902 Genotype and Changes in Fat Distribution by Diet Intervention

We further analyzed body fat distribution measured by CT. At 2 years, we found significant interactions between the FTO rs1558902 genotype and protein diet intervention on changes in TAT, VAT, and SAT (for interactions, P = 0.001, 0.012, and 0.002, respectively; Fig. 2). The risk allele (A) was associated with greater loss of TAT and VAT in the high-protein group but with less loss of TAT and SAT in the low-protein group (Table 2). At 6 months, gene-protein interactions were observed on changes in TAT and SAT (for interactions, P = 0.026 and 0.050, respectively; Fig. 2), and the risk allele carriers in the high-protein group had greater loss of TAT and VAT than noncarriers, but those in the low-protein group had less loss of SAT than noncarriers (Table 2).

Figure 2.

Interaction between the FTO rs1558902 genotype and dietary protein intervention on changes in TAT (A), VAT (B), DSAT (C), and SAT (D) at 6 months and 2 years. P values are adjusted for age, sex, ethnicity, carbohydrate, baseline values for respective outcomes, and baseline BMI. Data included 17, 18, 18, and 17 (TT); 30, 35, 35, and 30 (TA); and 10, 13, 13, and 10 (AA) participants in the low-protein group and 18, 22, 22, and 18 (TT); 26, 35, 35, and 26 (TA); and 12, 14, 14, and 12 (AA) in the high-protein group for TAT, VAT, DSAT, and SAT at 6 months (total n 137); and 12, 15, 15, and 12 (TT); 18, 25, 25, and = 18 (TA); and 8, 9, 9, and 8 (AA) participants in the low-protein group and 15, 17, 17, and 15 (TT); 23, 30, 30, and 23 (TA); and 8, 9, 9, and 8 (AA) in the high-protein group for TAT, VAT, DSAT, and SAT at 2 years (total n 105).

We did not find significant genetic effects and interactions on changes in fat distribution in subgroups treated by different dietary fat components (all P > 0.05; Supplementary Table 2 and Supplementary Fig. 2).Similar results were found in the white participants (data not shown).

The Trajectory of Changes in Body Composition and Fat Distribution by FTO rs1558902 in Response to Protein Diet

In a secondary analysis, we used linear mixed models to assess the genotype by time effect over the 2-year trial in those treated by the four dietary compositions. We observed significant genotype-time interactions on changes in total fat, FM%, and percentage of trunk fat in response to the high-protein diet. When assigned to the high-protein diet, participants who carried the AA genotype had greater loss in total fat, FM%, and percentage of trunk fat than those without this genotype. No genotype-time interaction on changes in body composition was found in the low-protein group (Fig. 3).

Figure 3.

Changes in total fat (A), FFM (B), FM% (C), and percentage of trunk fat (D) in the low-protein and the high-protein diet groups according to the FTO rs1558902 genotype from baseline to 6 months and 2 years. P values are adjusted for age, sex, ethnicity, carbohydrate, baseline values for respective outcomes, and baseline BMI. Data included 198, 149, and 99 in the low-protein group and 193, 161, and 125 in the high-protein group for body composition at baseline, 6 months and 2 years, respectively. (A high-quality color representation of this figure is available in the online issue.)

We also observed significant genotype-time interactions on changes in fat distribution in response to the low- and high-protein diets. Participants with the AA genotype had greater decrease in fat distribution in response to the high-protein diet compared with those without this genotype. In contrast, participants with the AA genotype were associated with less loss in TAT and SAT in response to the low-protein diet (Fig. 4).

Figure 4.

Changes in TAT (A), VAT (B), DSAT (C), and SAT (D) in the low-protein and high-protein diet groups according to the FTO rs1558902 genotype from baseline to 6 months and 2 years. P values are adjusted for age, sex, ethnicity, carbohydrate, baseline values for respective outcomes, and baseline BMI. Data included values at baseline and at 6 months and 2 years for 80, 57, and 38 participants, respectively, for TAT and SAT and 89, 66, and 49 participants, respectively, for VAT and DSAT in the low-protein group; and 71, 56, and 46 participants, respectively, for TAT and SAT and 86, 71, and 56 participants, respectively, for VAT and DSAT in the high-protein group. (A high-quality color representation of this figure is available in the online issue.)

We found genotype-time interactions on changes in total fat, FM%, and percentage of trunk fat in response to the low-fat diet (Supplementary Fig. 3), but no genotype-time interactions on changes in fat distribution were found in response to the low- or high-fat diets (Supplementary Fig. 4).A similar trend was observed in the white population (data not shown).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....