Liver Fat Accumulation in Response to Overfeeding With a High-fat Diet

A Comparison Between South Asian and Caucasian Men

Siti N. Wulan; Vera B. Schrauwen-Hinderling; Klaas R. Westerterp; Guy Plasqui

Disclosures

Nutr Metab. 2015;12(18) 

In This Article

Results

Subject Characteristics

South Asian subjects were Indian (n = 8) and Pakistani (n = 2). Caucasian subjects were Dutch (n = 3), Germans (n = 2), French (n = 1), British (n = 1), Danish (n = 1), Polish (n = 1) and Icelander (n = 1). South Asians were measured within 3 y of their stay in The Netherlands (n = 5) and within 1 y (n = 5). Subjects' characteristics are presented in Table 1. South Asian subjects were slightly older than Caucasians and had a significantly lower BMI (P = 0.04), but there was no difference in body fat percentage (P = 0.58). As consistently reported by many studies, South Asian men had a lower fat-free mass (P = 0.001) compared to Caucasian men, and a lower fat-free mass per square meter of height (17 kg/m2 versus 20 kg/m2).

Waist and hip circumference were higher in Caucasians than Asians, resulting in no difference in waist to hip ratio between ethnicities (P = 0.32). There was no difference in cardio-respiratory fitness corrected for fat-free mass between ethnicities (P = 0.15) and nor was physical activity (P = 0.68).

Body Fat Distribution

Body fat distribution of South Asian and Caucasian men are presented in Table 2. Caucasian men had a slightly higher absolute fat mass and fat mass per square meter of height but this was not significantly different (P = 0.85 and P = 0.65 respectively). Body composition of our two populations was further characterized, by calculating the fat mass to fat-free mass ratio.[39] In our study, the matching procedure in body fat percentage resulted in no difference in the fat mass to fat-free mass ratio (P = 0.63).

Abdominal obesity as shown by total abdominal fat area (TAT) and subcutaneous abdominal fat area (SAT) was slightly higher in Caucasians, whereas visceral fat area (VAT) was slightly higher in South Asians but did not reach statistical significance (P = 0.37, P =0.18 and P = 0.32 respectively). This may be due to the larger variation in the Caucasian group. However, as a percentage of the total abdominal fat area, VAT was higher in South Asians (P = 0.003) and so was the VAT to SAT ratio (P = 0.003). The abdominal fat scan was performed at the umbilical, except in one subject of the South Asian group, in this subject abdominal fat scan was performed at the sagital lumbar spine L3/L4 instead of the umbilical. Excluding this subject did not change the resulting P values for TAT, SAT, VAT, percentage VAT and VAT/SAT ratio (0.36; 0.18; 0.36; 0.004 and 0.004 respectively).

Fat in the extremities measured as biceps and triceps skinfold did not differ between ethnicities (P = 0.91 and P = 0.19 respectively). Truncal fat measured as subscapular skinfold was not different between groups (P = 0.36) nor was the suprailliac skinfold (P = 0.69).

Diet Composition and Compliance

There was no difference in compliance to the diet. Energy intake, as percentage overfeeding achieved, was 43 % and 52 % exceeding the requirement (P = 0.16) for South Asians and Caucasians respectively. Macronutrients composition of the actual energy intake during overfeeding was similar; with carbohydrate (26.2 % versus 26.4 %, P = 0.85), protein (14.6 % versus 14.9 %, P = 0.20) and fat (59.3 % versus 58.8 %, P = 0.53) for South Asians and Caucasians respectively. There was no difference in the proportion of saturated and unsaturated fatty acids in the diet (P = 0.49).

Hepatic Fat Content

Liver fat content before and after short-term overfeeding with a high fat diet is shown in Fig. 1. Data were available for 8 South Asians and 8 Caucasians matched for body fat percentage (P = 0.53). In the South Asian group, one had to be excluded because of poor signal to noise ratio in the spectra. In the Caucasian group, one subject could not undergo the measurement due to a technical problem. Therefore, these subjects were excluded from the analysis along with their body fat matched counterparts in the other group respectively.

Figure 1.

Individual (grey line) and the mean (black line) response of liver fat content to overfeeding with a high fat diet in South Asians (a) and Caucasians (b). Liver fat content before and after overfeeding with a high fat diet were assessed using ANOVA repeated measure. Data were available from 8 South Asians and 8 Caucasians matched for body fat percentage (25.0 ± 5.4 % and 23.2 ± 6.3 % for South Asian and Caucasian respectively, P = 0.53). Overfeeding with a high fat diet increased liver fat content (P = 0.01) but the increase did not differ between ethnicities (P = 0.47). SA: South Asian, C: Caucasian

Liver fat content at baseline did not differ between ethnicities (P = 0.48). Overfeeding with a high fat diet significantly increased liver fat content (P = 0.01) but the increase did not differ between ethnicities (P = 0.47). The mean increase was 33 % and 34 % for South Asians and Caucasians respectively. We assessed the association between liver fat at baseline and body fat percentage, it turned out that liver fat at baseline was positively associated with body fat percentage (R 2 = 0.44, P = 0.03, Fig. 2a). Furthermore, we assessed the effect of ethnicity on the baseline liver fat content with body fat percentage as a covariate (ANCOVA analysis) and found that baseline liver fat content was associated with body fat percentage (P = 0.02) but not with ethnicity (P = 0.21). We assessed the association between liver fat at baseline with visceral fat area, and found that liver fat at baseline had a stronger association with visceral fat area (R 2 = 0.62, P = 0.003, Fig. 2b). In a multiple regression analysis by including ethnicity in the model, visceral fat area was found to be the significant predictor of liver fat content at baseline (P = 0.002, R 2 = 0.56) and not ethnicity (P = 0.13). There was an outlier in Fig. 2a and b. Excluding the outlier, did not change the significant association between body fat percentage and baseline liver fat content in Fig. 2a (R 2 = 0.36, P = 0.018) and the significant association between visceral fat area and baseline liver fat content in Fig. 2b (R 2 = 0.76, P = 0.001).

Figure 2.

The association between liver fat at baseline with body fat percentage (a) and the association between liver fat at baseline with visceral fat area (b). Liver fat at baseline was associated with body fat percentage (R2 = 0.44, P = 0.03). Liver fat at baseline had a stronger association with visceral fat area (R2 = 0. 62, P = 0.003). In a multiple regression analysis, visceral fat area was the significant predictor of baseline liver fat (R2 = 0.56, P = 0.002) and no effect of ethncity was found (P = 0.13). Excluding the outlier, did not change the significant association between body fat percentage and baseline liver fat content in Fig. 2a (R2 = 0.36, P = 0.018) and the significant association between visceral fat area and baseline liver fat content in Fig. 2b (R2 = 0.76, P = 0.001)

processing....