Cardiovascular Journal of Africa: Vol 27 No 6 (November/December 2016)

CARDIOVASCULAR JOURNAL OF AFRICA • Volume 27, No 6, November/December 2016

368

AFRICA

Methods

Participants included a sample of 64 women from the original

convenience sample of 240 apparently healthy premenopausal

black SA women who were tested in 2005/06,

and were followed

up approximately 5.5 years later, as previously described.

The

original cohort of women were recruited at baseline from church

groups, community centres, universities and through the local

press, and were included in the study if they were (1) 18–45 years

old; (2) had no known diseases and were not taking medication

for type 2 diabetes (T2D), hypertension, HIV/AIDS, or any

other metabolic diseases; (3) were not pregnant, lactating or

postmenopausal (self-reported); and (4) were of SA ancestry

(self-reported). At follow up, the original cohort of 240 women

were contacted and invited to participate in the longitudinal

follow-up study in 2010/11.

Testing procedures at baseline included body composition

measures, questionnaires on SES and reproductive health, and

an assessment of baseline physical activity and dietary intake.

The dietary and physical activity assessment was not included

at follow up. At follow-up testing, voluntary HIV screening was

included. Participants were excluded on the basis of a confirmed

positive HIV test (Sanitests Home Test Kits, SA). For ethical

reasons, those who declined HIV screening were not excluded

from the study.

The study was approved by the Human Research Ethics

Committee of the Faculty of Health Sciences of the University

of Cape Town. Before participating in the study, procedures

and risks were explained to the subjects, and written informed

consent was obtained.

Body composition was assessed using basic anthropometry

(weight, height and circumference), dual-energy X-ray

absorptiometry (DXA) and computerised tomography (CT)

scans. DXA was used to measure whole-body composition

(Discovery-W

, software version 12.7.3.7; Hologic, Bedford,

MA).

In vivo

precision (CV) was 0.7 and 1.67% for fat-free soft-

tissue mass and fat mass, respectively. Percentage fat mass for the

whole body was obtained and fat mass for the various regions of

interest, including the trunk, limbs, android and gynoid regions,

were derived using DXA cut-off lines positioned at anatomical

markers, as previously described.

CT was used to measure

abdominal visceral adipose tissue (VAT) and superficial adipose

tissue (SAT) areas (Toshiba X-press Helical Scanner

; Toshiba

Medical Systems, Tokyo, Japan) in 43 women at baseline and

follow up.

A Xhosa-speaking field worker administered the socio-

demographic questionnaire at baseline and follow up. The

questionnaire included measures of SES such as housing density,

asset index, educational level, current employment and household

sanitation. Housing density was defined as the number of persons

in the household divided by the number of rooms. Asset index

was based on 14 appliances/items, reflecting the individual

and household wealth and resources. These included electricity

in the home, ownership of a television, radio, motor vehicle,

fridge, stove/oven, washing machine, telephone, video machine,

microwave, computer, cellular telephone and paid television

channels (MNET or DSTV). Level of education was described as

those who had completed grade 12 (secondary school) or lower.

Participants were categorised as employed (including students) or

unemployed. Sanitation was described as access to running water

or a flush toilet inside or outside the house.

Behavioural factors included self-reported indicators

of current smoking status (smoker or non-smoker), alcohol

consumption (non-drinker or drinker of any alcohol), and

hormonal contraceptive use (none, oral or injectable). Parity was

defined as those who had children at baseline or follow up, and

those who had children during the follow-up period.

Physical activity was assessed at baseline using the global

physical activity questionnaire (GPAQ).

As walking was the

most frequent activity, walking for travel was used as a proxy

for physical activity. Dietary intake was determined using a

quantitative food frequency questionnaire,

which has been

validated in black SA women.

A higher diet quality index –

international (DQI-I) score represents a higher quality of dietary

intake.

Statistical analysis

Parametric data are presented as means and standard

deviations and non-parametric data are presented as medians

and interquartile ranges (IQR) and compared using paried

-tests and the Mann–Whitney

-test. Socio-demographic or

categorical data are presented as percentages and were compared

over the follow-up period using McNemar chi-squared tests.

For univariate analysis, Spearman’s rank correlations were

used to assess non-parametric associations between continuous

variables (housing density, asset index) and the changes in body

composition, while ANOVA was used to explore the effects

of categorical variables (parity, access to sanitation, smoking,

alcohol and walking for travel) on changes in body composition.

To analyse the effect of baseline age and BMI on changes in

body composition, median age and accepted BMI classifications

were used to create groups, and a two-way analysis of covariance

(ANCOVA), adjusting for age, was performed. Based on the

significant univariate associations with changes in body weight

and body composition (baseline age and BMI, access to

sanitation, parity, level of education and employment status,

and changes in these SES and lifestyle variables), multiple

stepwise linear regression was used to determine the independent

contribution of these variables to changes in weight gain

and body fat distribution over the 5.5-year follow-up period.

Statistical significance was set at

0.05. Data were analysed

using STATISTICA version 10 (Statsoft Inc. Tulsa, OK) and

STATA 12.1 (StataCorp, College Station TX).

Results

Subject characteristics, including body composition, SES and

lifestyle variables at baseline and follow up are presented in

Table 1. Mean percentage weight gain over the follow-up period

was 8.8%, with an average increase of 1.2 kg/year. There was a

significant increase in fat mass (16.4

26.9%,

0.001), but

no significant increase in fat-free soft-tissue mass (

0.234).

The increase in fat mass was largely attributed to an increase

in central fat mass, characterised by increases in trunk (as a

percentage of total fat mass) and android fat mass, as well as

both VAT and SAT areas. Conversely, there was a decrease in

peripheral fat mass (gynoid and leg fat mass as a percentage of

total fat mass).

The measures of SES of the participants increased over the

follow-up period, as characterised by increases in asset index