Phenotyping Ethnic Differences in Body Fat Depots
Abstract
There are remarkable ethnic differences in the incidence of metabolic
syndrome associated features; including insulin resistance, type 2 diabetes,
hypertension and cardiovascular diseases. Studies have suggested that
South Asians (SA) present an unfavourable body fat phenotype, which
includes a pattern of elevated visceral adipose tissue (VAT), and liver fat
content; depots strongly associated with the progression of metabolic
dysregulation. However, there are a limited number of studies examining
body fat composition by ethnicity.
The purpose of this thesis was to comprehensively phenotype VAT,
abdominal subcutaneous adipose tissue (ASAT) and liver fat content in
Caucasian (Cau), SA and Black African (BA) individuals from a large number
of distinct populations. Here, I include data from three adult cohorts: the UK
Biobank (n=9533) of mixed ethnicities, the DIRECT cohort (n=1553) of Cau
pre-diabetic individuals and The West London Observation (TWLO) cohort
(n=747) of mixed ethnicities. In addition, I present data from Pune Maternal
Nutrition study (PMNS) cohort; comprising 423 young adults of SA descent
in India.
Analyses of body fat phenotype in Cau pre-diabetic populations showed
higher VAT (mean differences= 0.5 litre, p<0.0001) and liver fat content
(mean differences= 0.6%, p<0.0001), but lower ASAT (mean differences= -
0.2 litre, p<0.0001) compared to Cau from the general population (free-living).
I also observed negative associations between VAT, ASAT, liver fat content
and day to day physical activity in both pre-diabetic and general populations
(pre-diabetic; VAT; r= -0.296, ASAT; r= -0.163, liver fat: r= -0.186 and general
population; VAT; r= -0.185, ASAT: r= -0.374, liver fat: r= -0.139, p<0.001 for
all).
Analysis of both the TWLO and UK Biobank revealed no differences in VAT
or liver fat in SA in UK compared to other ethnic groups (TWLO; VAT: SA:
3.0 ± 1.6 litres, Cau: 3.3 ± 2.1 litres; liver fat: SA= 6.4 ± 11.1%, Cau= 6.5 ±
13.6%, p=ns - UK Biobank; VAT: SA: 3.6 ± 1.6 litres, Cau: 3.8 ± 1.5 litres; liver fat: SA: 4.6 ± 4.6%, Cau: 4.2 ± 4.6%, p=ns). Analysis of both these
cohorts also revealed a more favourable body fat phenotype with BA males
presenting significantly less VAT than SA and Cau males (p<0.05 for both).
Data from the PMNS cohort revealed high levels of VAT in 18 year old Indiabased SA population. A high proportion (58.7%) of these lean individuals also
presented with the thin-outside fat inside (TOFI) phenotype (a ratio of VAT to
ASAT).
A key finding is the lack of an unfavourable body fat phenotype in UK based
SA. Therefore, the increased incidence of metabolic syndrome associated
features in the SA population may arise via a mechanism unrelated to
elevated levels of VAT or liver fat.