WHO IS GETTING FAT IN SOUTH AFRICA? Annibale Cois 1, 2 , Candy Day - - PowerPoint PPT Presentation

WHO IS GETTING FAT IN SOUTH AFRICA?

Obesity trends and risk factors in the South African adult population

1 -School of Public Health and Family Medicine, University of Cape Town 2 - Health Systems Trust

Health Systems Trust Conference 2016 Birchwood Conference Centre, Boksburg, Gauteng 6 May 2016 Annibale Cois 1, 2 , Candy Day 2

HST Conference 2016

LITTLE DOUBT EXISTS THAT THE PREVALENCE OF OBESITY AND OVERWEIGHT IS ON THE RISE

Background

GLOBALLY IN SOUTHERN AFRICA IN SOUTH AFRICA

!

HST Conference 2016

Background

Modified from: Finucane et al. Lancet. 2011;377(9765):557-67.

World Southern Africa South Africa

0.5

kg/m2 per decade

0.7

kg/m2 per decade

0.4

kg/m2 per decade

1.0

kg/m2 per decade

0.9

kg/m2 per decade

1.4

kg/m2 per decade

1.6

kg/m2 per decade

2.9

kg/m2 per decade

HST Conference 2016

Females Males

HST Conference 2016

Background

Estimated distribution of BMI in the South African adult population: Females underweight

• verweight
• bese

normal weight

Data from: Ardington & Case. Cape Town. SALDRU. 2009

HST Conference 2016

Background

Data from: Ardington & Case. Cape Town. SALDRU. 2009

Estimated distribution of BMI in the South African adult population: Males underweight

• verweight
• bese

normal weight

HST Conference 2016

Background

WHAT IS DRIVING THIS TREND?

? WHO

?

HST Conference 2016

?

Background

WHO IS DRIVING THIS TREND?

?

HST Conference 2016

Aims

IDENTIFY SUBGROUPS IN THE SOUTH AFRICAN ADULT POPULATION CHARACTERISED BY DIFFERENT TRAJECTORIES OF CHANGE IN BODY MASS INDEX IDENTIFY SOCIOECONOMIC, BIOLOGICAL AND BEHAVIOURAL CHARACTERISTICS DIFFERENTIANG THESE GROUPS

HST Conference 2016

A PUBLIC HEALTH PERSPECTIVE

TARGET INTERVENTIONS

UNMODIFIABLE AND MODIFIABLE CHARACTERISTICS

BURDEN OF DISEASE

Aims

HST Conference 2016

• Panel survey of a nationally representative sample of the SA population
• ~28 000 subjects interviewed in 2008 and recontacted every 2 years
• In this study: data on 10 100 subjects 18+ at baseline and successfully recontacted

in 2010 and 2012

First Wave: 2008 Second Wave: 2010 Third Wave: 2012 Fourth Wave: 2014

www.nids.uct.ac.za

Methods

HST Conference 2016

Growth Mixture Modelling (GMM) was used to identify latent classes of subjects characterised by distinct patterns of change in BMI during the study period. DATA ANALYSIS A multinomial logistic regression model was fit to identify significant predictors of class membership

Education Household income Urban vs. rural residence

SOCIOE BIO

Gender Age Waist circumference

BEHAV

Alcohol use Smoking Physical exercise

Methods

HST Conference 2016

Results

AVERAGE BMI (AND THUS OBESITY PREVALENCE) ARE STILL RISING IN THE SOUTH AFRICAN ADULT POPULATION

1

1.8

(1.1 – 2.6)

kg/m2 per decade

1.0

(0.1 – 1.9)

kg/m2 per decade

Females Males

1.6

kg/m2 per decade

2.9

kg/m2 per decade

HST Conference 2016

Results

AN ASIDE: WAIST CIRCUMFERENCE IS INCREASING, TOO…

Kernel density estimate, weighted

HST Conference 2016

Females

HST Conference 2016

Results

… AMONG MALES AS WELL

Kernel density estimate, weighted

HST Conference 2016

Males

HST Conference 2016

Results

POPULATION IS NOT HOMOGENEOUS: SOME PEOPLE ARE PUTTING ON WEIGHT FASTER THAN OTHERS (AND SOME ARE LOSING IT)

Normal (relatively) stable (82.1% of the population) Obese stable (12.3%)

2

HST Conference 2016

HST Conference 2016

PEOPLE THAT ARE INCREASING FASTER THEIR WEIGHT ARE:

3

WOMEN RATHER THAN MEN

ORadj = 2.7

AT THE EXTREMES OF THE INCOME DSITRIBUTION

ORadj = 2.3 2.0

WITH GREATER WAIST CIRCUMFERENCE

ORadj = 1.9 per 10 cm increase

RURAL DWELLERS

ORadj = 1.7

HST Conference 2016

ARE RURAL PEOPLE ‘CATCHING UP’ WITH URBANITES?

Results

2008 2010 2012 Linear regression estimates from NiDS survey, weighted BMI

HST Conference 2016

PEOPLE THAT ARE DECREASING THEIR WEIGHT FASTER ARE:

OBESE WITH SMALLER WAIST SMOKERS, NON DRINKERS, EXERCISE MODERATELY MEN RATHER THAN WOMEN

4

ORadj =14 ORadj = 2.0 per 10 cm decrease ORadj = 7.6 ORadj = 0.1 ORadj = 11.0

HST Conference 2016

Results

AGE HAS A COMPLEX EFFECT, COMPATIBLE WITH THE HYPOTHESIS THAT YOUNG PEOPLE ARE LESS STABLE. COMPARED TO OLDER PEOPLE, THEY ARE MORE LIKELY EITHER TO INCREASE OR DECREASE THEIR WEIGHT

5

HST Conference 2016

Conclusions

HST Conference 2016

HST Conference 2016

Conclusions

A strong positive trend in BMI remains in South Africa and obesity prevalence is likely to increase in the next years Trends are not homogeneous, and high risk groups (subjects at the extreme of the income distribution, rural dwellers, women) and modifiable risk factors (physical inactivity, alcohol use) can be targeted Subjects quitting smoking should receive additional weight-loss support in order that the numerous health benefits of cessation are not reduced by increasing BMI Further research is needed on the role of central obesity

HST Conference 2016

Strenghts & Weknesses

• Low reliability of self-report data for

risk factors

• Absence of reliable data on nutrition
• Low response rates and high

attrition in some social strata

• Choice of number of classes in GMM
• Only three waves available
• Large sample
• Longitudinal study
• Direct measurement
• f weight and height
• Efficient treatment of

missing data and random measurement error

STRENGTHS WEAKNESSES

THANK YOU!

annibale.cois@uct.ac.za

This study has received financial support from the Programme to Support Pro-Poor Policy Development II (PSPPD II), a partnership between the Presidency, Republic of South Africa and the European Union. Health Systems Trust Conference 2016, Birchwood Conference Centre, Boksburg, Gauteng