Content uploaded by Candy Day
Author content
All content in this area was uploaded by Candy Day on Dec 14, 2018
Content may be subject to copyright.
Authors:
115
Noluthando Ndlovui Candy Dayi
Benn Sartoriusii Jens Aagaard-Hanseniii
Karen Hofmaniv
Assessment of food environments
in obesity reduction: a tool for
public health action
Assessment of food environments
in obesity reduction: a tool for
public health action
By measuring the food
environment geographically,
healthy food access gaps
can be identified and
nutrition-sensitive preventive
interventions can be
developed
1313
i Health Systems Trust
ii School of Nursing and Public Health, Universit y of KwaZulu-Natal
iii Steno Diabetes Center Copenhagen, Gentofte, Denmark; and MRC Developmental Pathways
for Health Research Unit, Faculty of Health Science, University of the Witwatersrand, Johannesburg
iv PRICELESS SA, School of Public Health, University of the Witwatersrand, Johannesburg
The nutrition transition in sub-Saharan African countries has contributed to
increased incidence of overweight and obesity, which constitutes a major public
health risk. This is especially the case where dietary patterns are influenced by
the ready availability of fast foods, resulting in a high intake of fat, sugar and salt.
This low-quality diet increases the risk of non-communicable diseases (NCDs). By
measuring the food environment geographically, healthy food access gaps can be
identified and nutrition-sensitive preventive interventions can be developed.
Addresses of food retailers were geocoded to quantify the total number of grocery
stores (healthy options) and fast-food outlets (less-healthy options) within wards across
Gauteng, the most densely populated province in South Africa. The Modified Retail
Food Environment Index (mRFEI) was then computed, representing the percentage of
‘healthy’ food retailers in the area.
The mRFEI was widely heterogeneous across Gauteng, ranging from a minimum of
5% to a maximum of 100%, with an average of 33%. The index was highest in the
most affluent wards and lowest in the poorest wards, with the latter including a high
number of informal settlements. This diverse result was consistent with the high levels
of socio-economic inequality that have been observed in Gauteng.
For countries such as South Africa currently undergoing rapid nutritional transition,
it is imperative to be creative in finding cost-effective ways to identify the structural
drivers of NCDs. Through supporting healthy food environments, the public health
goals of reducing and preventing obesity and improving nutrition can be reached in
settings with a high and increasing burden of obesity.
Review
116 2018 SAHR
Introduction
In 2015, overweight and obesity contributed to four million deaths
globally, with cardiovascular disease accounting for 70% of those
deaths, followed by diabetes (15%).1 While obesity is prevalent
in both high-income countries (HICs) and low- and middle-income
countries (LMICs), it affects the poor disproportionately and
contributes to growing health inequities at all levels.2
Non-communicable diseases (NCDs) are driven by a complex
interplay of multiple risk factors. However, a low-quality diet,
which can lead to obesity, combined with reduced physical
activity, increases the risk of NCDs such as hypertension, diabetes,
cardiovascular disease and cancer.3,4 In HICs, NCDs have also
been inversely associated with socio-economic status, with some
studies finding increased consumption of fast food among low-
income and black populations.4–6 Morland and Filomena also
found disparities in the availability of healthy food between racially
segregated urban neighbourhoods in the USA, where there were
hardly any supermarkets in predominantly black areas.7 In another
study conducted in Australia, Burns and Inglis found that those living
in advantaged areas had better access to supermarkets, while those
living in disadvantaged areas lived in closer proximity to fast-food
outlets.8
In 2016, South Africa had the highest prevalence of obesity among
sub-Saharan African countries,9 with 68% of women and 31%
of men considered overweight or obese.10 Sub-Saharan African
countries have undergone a nutrition transition towards a diet high
in sugar and saturated fats but low in fibre,11 which has contributed
to the emergence of overweight and obesity as a critical public
health problem.6,12
There is currently a global discourse on the introduction of planning
laws to regulate the spread of fast-food stores13,14 and food
environments that are not supportive of healthy eating.15 ‘Food
environment’ can be defined as the physical, economic and social
factors that impact the availability, accessibility and adequacy of
food within a region, or as the everyday stimuli that encourage
a consumer’s food choices in a particular way.16 Various factors
influence the choices people make in acquiring and consuming
food; these include household income, proximity to food store
location, food price, pervasive and persuasive food marketing, and
convenience.3,17,18
Numerous studies have also found associations between the number
of neighbourhood fast-food outlets and obesity rates, as fast-food
consumption is linked to increased body mass index (BMI) and
weight gain.4,9,19–22 Promotion and low price of fast food, and
easy access to it, are probably major drivers of obesity and related
NCDs.1,23 However, there are no structured prevention interventions
to improve food environments in South Africa, and prevention is still
aimed largely at an individual level.3
Several studies conducted in other countries have found significant
associations between the number and proximity of fast-food outlets
and the high frequency of purchasing such foods.14,24–26 In South
Africa, the fast-food industry is experiencing exponential growth,
with a predicted annual growth rate of 9% for the 2014–2019
period.27 In measuring the food environment, food access gaps
can be identified, allowing for the development of nutrition-sensitive
preventive interventions that prioritise high-risk areas.9
Overview of study
The purpose of this study was to calculate the Modified Retail
Food Environment Index (mRFEI)28 at the ward level in Gauteng
(GP) and to assess whether food environments varied according to
socio-economic status, thereby generating evidence to inform policy
on the drivers of the obesity epidemic. Obesity is a risk factor for
most NCDs,29 yet measures to reverse the increasing prevalence
of overweight and obesity are still largely absent.30 Utilisation of a
tool such as the mRFEI is an example of an easy method that looks
beyond the health system in the prevention of obesity and NCDs.
Setting
Gauteng was selected as a relevant location to assess the food
environment as it has well-developed infrastructure, making it easier
to find geo-located food outlets as there are proper street addresses,
which would be more difficult in areas that are predominantly
rural. Furthermore, there is a high level of socio-economic inequity
in GP, making it an appropriate location to assess whether food
environments differ by socio-economic status. The study was
conducted at ward level. Based on 2011 demarcations, there were
508 wards, with population density ranging from 4 to 66 664
persons per km2 across the various wards.31
The mRFEI
The mRFEI is an environmental indicator of food access or the
proportion of ‘healthy stores’ within a defined neighbourhood
relative to all accessible stores. The definition of ‘healthy’ and ‘less-
healthy’ food retailers is based on the Centers for Disease Control
and Prevention (CDC) definition, which states that healthy food
retailers include grocery stores and supermarkets, while less healthy
food retailers are fast-food restaurants.32
The mRFEI was chosen to quantify the retail food environment
because it includes both unhealthy and healthy food outlets in a single
measure to give a comprehensive picture of the food environment.33
In the South African context, supermarkets and grocery stores were
used as a proxy for healthy food based on typical food available
in this type of retail format, while fruit and vegetable markets were
excluded due to lack of data. The assumption is that grocery stores
stock healthy foods such as fruit and vegetables, meat and whole
grains. The four major grocery store chains accounting for 97%
of sales in the South African formal food sector were selected for
calculation of the mRFEI; these were Shoprite Checkers, Pick ‘n
Pay, Spar, and Woolworths.11 Different size stores were included,
namely convenience stores, supermarkets and hypermarkets.34
Only fast-food outlets were chosen as a proxy for unhealthy foods in
the assessment. Full-service restaurants (e.g. Spur) were not included
as the quality of food differs between fast-food outlets and full-service
restaurants, with full-service restaurants often providing healthier
food options for health-conscious clients.35 Food outlet locations
were collected from the retailers’ websites and Google Maps and
geocoded using ArcMap version 10.5.36 Once the geographical
co-ordinates of the outlets were recorded, further analysis was
done in ArcMap. The mRFEI was then computed using a formula
developed by the CDC. The index measures the number of ‘healthy’
(grocery store) and ‘unhealthy’ (fast-food outlet) food retailers within
wards across GP, as defined by typical food offerings in the specific
store types. The mRFEI shows the percentage of retailers considered
‘healthy’ out of the total number of food retailers.
Assessment of food environments
SAHR 2018117
•
•
•
•
•
•
•
••
•
•
•
••
•
•
•
•
•
• •
•
••
•
•
••
•
••
•
••
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•••••
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
••
••
•
•
••
•
•
•
•
•••
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•••
•
•
•
••
•
••
•
•
•
•
•
•
•
•
•
•
•••
••
•
•
•
•
•
•
••
•
•••
••••
•
•••• •
•
•
•
•
••••
•
••
•••
••
• •
••
•
•••• •
••• •
•• ••• •
•
•
•
•••
• •
••
•• •• •
• •
• •
••
•••
••
•
••
•••
•
••
••
•
•••
•
• •
•••
•
•
••
•••
•
•
•
•
•
••
•
•
•
•
•
•••
•
•
••
•
•
••
•
•
•
•
•
••
•
•
••
••
•
•
••
•
•
•
•
•
••
•
•
•
•
•
••••
•
•
•
•
•
••
•
•
•
•
•
••
•
•
••
•
•
••
•
•
••
•••
•
•
••
• •
••
••
•
•••
•••
•
•
•
•
••
••
•
•
•
••
••
• •
•••
•
•
•
•
•
••
••
••
•
•
•
•
••
•
•
•
•
•••
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•••
•••
•
•
•
••
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
••
•
•
•
•
•
•
•
•
•
•
••
•
••
•
••
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•••
•
•
•
•
••
••
•
•
•
•
•
•
••
•
•
•
•
•
•
•
••
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
••
•
•
•
•
•
•
•
•
••• •
•
•••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•••
•
•
••
•
•
•
••
•
••
•
•
•
•
••••
•
••
•
•
•
•
•
•
•
•••
•
•
•
••
•
•
••
•
•
•
•
••
•
•
•
•
•
••
•• •
••
•
•
•
•
•
•
•
••
Decile
Grocery Stores
Municipality
endgeL
10 (= most deprived)
1 (= least deprived)
2
3
4
5
6
7
8
9
Wards
SAIMD
Tshwane
West Rand Johannesburg
Ekurhuleni
Sedibeng
0 510 20 30 40
Kilometers
Area-level deprivation and socio-economic
indicators
In addition to assessing the food environment, socio-economic
factors in the wards were also assessed to investigate if there were
any correlations between the food environment and socio-economic
factors. The South African Index of Multiple Deprivation (SAIMD)
was used to assess the socio-economic factors, together with census
and community survey data from Statistics South Africa (Stats
SA).37,38 The SAIMD is a relative measure of multiple deprivation
expressed at small-area (ward) level and takes into account the
four dimensions of deprivation, namely employment deprivation,
education deprivation, material deprivation, and living environment
deprivation. The four dimensions are combined and weighted
equally to construct the overall deprivation score.
All wards in the country are divided into 10 deciles according to
their poverty rates, with decile 1 being least deprived and decile 10
being the most deprived. Gauteng has very few wards in deciles 8,
9 and 10 (for a detailed breakdown of the indices and indicators
see Noble et al.39). The SAIMD deciles were calculated for the
entire country; consequently, the inequality among wards in GP was
masked as GP has low levels of deprivation compared with other
provinces in South Africa. This prompted an exploration of selected
socio-economic factors such as household income and employment
rates in individual wards, using data directly from Stats SA.
Key findings
In November 2016, there were 1 559 unhealthy food outlets and
709 healthy food outlets in GP (Table 1).
Table 1: Total number of food outlets in Gauteng, South Africa,
2016
Unhealthy outlets Total (N) Healthy outlets Total (N)
KFC 202 Checkers 245
Steers 194 Pick ′n Pay 201
Debonairs Pizza 182 Spar 151
Wimpy 164 Woolworths Food 112
ChesaNyama 159
McDonald’s 140
Nan do’s 115
Roman’s Pizz a 88
Chicken Licken 79
Fishaways 64
The Fish a nd Chip Co. 42
Burger K ing 29
Panarottis 23
Pizza Hut 21
Bar cel o’s 20
Anat 18
Andiccio24 16
Chickin Tyme 3
Total 1 559 Total 709
Distribution of healthy food outlets is highly inequitable in GP. Wards
with the highest number of stores with healthier food options were
located predominately in suburban areas (Figure 1).
Figure 1: Distribution of healthy food outlets and ward-level SAIMD deciles across Gauteng, 2016
118 2018 SAHR
•
•
•
•
•
•
•
•
••
••
•
•
••
•
•
••
•
•
•
•
••
•••
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
••
•
•
•
••
••
•
•
•
•
••
•
•
•
•
•
••
•
•
•
•
•
•
••
•
••
•
•
•
•
•
•
•
•
•
••
••
•
•
•
• •
•
••
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
••
•
•••
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•••
•
•
•
•
•• •
•
•
•
•
•
••••
•
••
•
••
•
•
••
•• •
••
••
•
•
•
•••
•
•••
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•••
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•••
•
•
••
•
•
••• •
•
•
•
•
•
•
•
••
•
••
•
••
••
•
•
•
••
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•••
•
•
•
•
•
•
•
•
•
•
•
•
• •
•
•
•
•
••
••
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
••
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•••
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
••
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
••
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
••
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
••
••
•
•
•
•
•
•
•
•
•
•
•
•
•
• •
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
••
•
•
•
•
••••••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
••
•
•
•
•
•
•
•
••
•
•
•
•
••
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
••
•
•
•• •
•
•
•
•
•
•
•
••
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
••
•
•
•
•
•
••
••
•••
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
••
•
•
•
••
•
•
•
•
•
•
•
•
•
•
••
•
•
•
•
••
••
••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Tshwane
West Rand Johannesburg
Ekurhuleni
Sedibeng
0 510 20 30 40
Kilometers
Decile
Fast food outlets
Municipality
endgeL
10 (= most deprived)
1 (= least deprived)
2
3
4
5
6
7
8
9
Wards
SAIMD
•
Although the distribution of unhealthy food outlets showed a similar
pattern, there was also a high concentration of fast-food outlets
in wards located in the inner city of Johannesburg and in black
communities (Figure 2). The highest number of unhealthy food
outlets in one ward was 29, while the highest number of healthy
food outlets in a ward was only 17 (Tables 2 and 3). Wards with
the highest number of unhealthy food outlets were located mainly in
Johannesburg (Table 2).
Table 2: Wards with the highest number of unhealthy food
outlets, Gauteng, 2016
Ward
no.
Main suburbs Municipality No. of fast-food
outlets
146 Lynwood Tshwan e 29
2106 Bryanston Johannesburg 28
374 Melros e North Johannesburg 23
4103 Sandton Johannesburg 21
593 Paulshof Johannesburg 20
654 Ridgeway Johannesburg 20
760 Braamfontein Johannesburg 20
8112 Noordwyk Johannesburg 20
9115 Craigavon Johannesburg 20
10 97 Wilgeheuwel Johannesburg 20
Figure 2: Distribution of unhealthy food outlets across Gauteng, 2016
Table 3: Wards with highest number of healthy food outlets,
Gauteng, 2016
Ward
no.
Main suburbs Municipality No. of
grocery
stores
191 Mooikloof Hills Tsh wa ne 17
278 Zwartkop, B ronber rik Tsh wa ne 14
3103 Sandton Johannesburg 13
422 Boksburg Noord Ekurhuleni 13
585 Waparand Tshw an e 12
646 Lynwood Tshwan e 10
7106 Dou glasda le, Brya nston,
Rivonia
Johannesburg 10
875 Welgedacht Ekurhuleni 10
992 Activia Park, Bar vallen Ekurhuleni 9
10 20 Bedfordview, Morninghill Ekurhuleni 9
The highest incidence of wards with no grocery stores was observed
in low-population-density wards, which was to be expected. The
maps (Figure 3) show how the mRFEI varied across the wards in
GP. The majority of wards had low mRFEI percentages for healthy
food outlets, either zero or in the range from 20% to 39.9%. Very
few wards had percentages above 59.9% (Table 4). The low mRFEI
percentages could be indicative of highly obesogenic environments.
Assessment of food environments
SAHR 2018119
City of Ekurhuleni Metropolitan Municipality
mRFEI
No retail food outlets
0 (no grocery store)
0.1 - 19.9
20 - 39.9
40 - 59.9
60 - 85
100 (no fast food outlet)
0 5,5 11 16,5 222,75
Kilometers
0 6.5 13 19.5 263.25
City of Johannesburg Metropolitan Municipality
FEI
No retail food outlets
0 (no grocery store)
0.1 – 19.9
20 – 39.9
40 – 59.9
60 – 85
100 (no fast food outlet)
mR
0 10 20 30 405
Kilometers
Sedibeng District Municipality
mRFEI
No retail food outlets
0 (no grocery store)
0.1 – 19.9
20 – 39.9
40 – 59.9
60 – 85
100 (no fast food outlet)
City of Tshwane Metropolitan Municipality
100 (no fast food outlet)
60 – 85
40 – 59.9
20 – 39.9
0.1 –19.9
0 (no grocery store)
No retail food outlets
IEFRm
Kilometers
20 30 400 105
West Rand District Municipality
mRFEI
No retail food outlets
0 (no grocery store)
0.1 – 19.9
20 – 39.9
40 – 59.9
60 – 85
100 (no fast food outlet)
Kilometers
3425.5178.54.250
Figure 3: mRFEI at ward level in the different municipalities in Gauteng, 2016
Table 4: Number of wards per mRFEI category, Gauteng, 2016
mRFEI categor y
(percenta ge of healthy food retai lers)
Number of wards
0% (No retail food outl ets) 208
0% (No groc ery) 83
0.1 –19 .9% 28
20– 39.9% 73
40–59.9% 70
60 –85% 17
100% (No fast food outlets) 29
On average, the municipality with the highest percentage of healthy
food outlets was Ekurhuleni, with an mRFEI of 45%. It was also the
municipality with the highest number of wards with percentages of
100, indicating that several wards only had healthier food outlets
available. The worst-performing municipalities were Johannesburg
and the West Rand, which on average had percentages of 28 and
27 respectively, indicating that only 27/28 out of 100 stores in those
municipalities were likely to provide healthier food options (Table 5).
Overall, in GP there are healthy food options in 33 out of 100
stores. The West Rand and Sedibeng, the most rural municipalities
in GP, had the highest number of wards with no retail food outlets.
120 2018 SAHR
Table 5: Descriptive statistics of mRFEI by municipality, Gauteng, 2016
Municipality Mean Standard
error
Median Min. Max. Number
of ward s
No retail
food outlets
Wards with
no groc ery
sto re (%)
Wards with
no fast-food
out le t (%)
Johannesburg 28 2.76 23 0100 13 0 32 21 5
Tsh wa ne 33 3.35 33 010 0 105 37 16 4
Ekurhuleni 45 3.85 46 0100 101 32 12 12
Sedibeng 29 7. 2 2 17 010 0 72 47 15 6
West Rand 27 4.88 13 0100 10 0 60 16 3
GP 33 1.76 30 0100 508 208 16 6
The third-lowest category of areas with a low percentage of healthy
food outlets (mRFEI 20–39.9) were high-population-density wards
in Johannesburg and Tshwane with low-income black residents.
Table 6 shows that most of the top 10 wards with only fast-food
outlets were in black areas, although only two of those wards fell
in the fourth SAIMD decile, indicating higher levels of deprivation
compared with the other wards. However, a few wards also fell in
the least-deprived decile; they had high average annual household
income compared with other wards, and were mostly occupied by
whites. Two wards with a high percentage of informal dwellings were
among the wards with only unhealthy food outlets. In one of those
wards, most households (55%) were living in informal dwellings.
Another ward in a black community had a very high population
density (over 10 000/km2) yet there wasn’t a single grocery store
or supermarket in the immediate area (Table 6).
Wards with the highest deprivation index in GP (between decile 7
and 10) had no formal food retail outlets at all. The majority of these
wards were in the West Rand, and the population was composed
mainly of low-income blacks living predominantly (over 90%) in
informal dwellings. A similar trend was observed in wards in the
other GP municipalities. The majority of wards with fast-food outlets
only fell in the third SAIMD decile, i.e. among the least-deprived
wards, and the number of wards decreased as the SAIMD decile
increased from low deprivation to high deprivation.
Table 6: Top 10 wards with fast-food outlets only (i.e. mREFI = 0), Gauteng, 2016
Ward
no.
Municipality Main suburb SAIMD Decile
(10=most
deprived)
Majority
ethnicity
Average
annual
household
income
Population
density
Informal
settlements
(%)
Employment
rate (%)
74 Johannesburg Melrose North 1White (47%) 115 1 0 0 1 929 0.3 74. 2
12 West Rand Welverdiend 2White (54%) 57 300 35 1.2 45.2
55 Johannesburg Lindbergh Park 1Black (50%) 57 300 4 132 0.8 60.3
96 Johannesburg Lion Par k informa l
settlement
4Black (74%) 29 400 454 54.6 60. 2
8West Rand Bhongweni 2Black (54%) 57 300 663 6.6 37. 4
122 Johannesburg Zakariyya Park 4Black (91%) 14 6 00 810 33.6 4 3.1
53 Johannesburg Slovoville 2Black (100%) 57 300 700 2.3 45.4
48 Johannesburg Dobsonville 2Black (99%) 29 400 10 159 11.4 45
94 Ekurhuleni Generaal
Albertspark
1White (48%) 230 700 955 0.9 69.8
15 Tsh wa ne Mamelodi 1Black (99%) 29 400 2 805 3.3 46 .1
Correlation between the mRFEI and socio-economic
factors
According to the SAIMD, Gauteng is the second-least-deprived
province in the country, and Tshwane and Johannesburg are among
the 10 least-deprived municipalities.39 The majority of the 10 worst-
performing wards in GP were in predominantly black areas, with the
exception of three wards where whites were slightly in the majority
(Table 6). These wards only had fast-food outlets, without a single
healthy food outlet.
Several township areas had high population densities yet there were
no food retail outlets in those wards. One such example was Zola in
Johannesburg, with a population density of approximately 14 000
people per square kilometre (km2). Zola is a low-income area where
100% of the population are black and 6.5% of households live in
informal settlements. Only 35% of the people living in this area are
employed.40
Wards with the lowest percentage of healthy food outlets also had
relatively low population densities (20–2 500 people/km2), with
the exception of one ward in Tshwane that had a high population
density of over 16 000 people/km2. This ward also predominantly
included black residents who were low middle-income earners.
However, the area had very few informal dwellings (0.2%), and
approximately half of the population was employed.
Assessment of food environments
SAHR 2018121
Discussion
Although individuals make decisions about food choices in a complex
set of physical and social environments,35 the social patterning
of NCDs is influenced by differential exposure to obesogenic
environments leading to the consumption of excess calories.41
The mRFEI revealed that GP is a highly obesogenic environment,
especially the wards in Johannesburg. The geographical distribution
of grocery stores in GP is similar to the pattern in Cape Town (the
Western Cape being the least-deprived province) in that grocery
stores are concentrated in higher socio-economic areas.11 This trend
is also similar to what has been observed in HICs such as the USA
and Australia, where the type of food outlet changes according to
neighbourhood economic status.7,8
According to Rudolph et al.,42 fast-food outlets, small shops and
restaurants play an important role in day-to-day provisioning among
the urban poor in GP, with 55% of households sourcing food from
these outlets at least once a week or more often, especially in the
inner city. In the lower- to middle-income and predominantly black
communities, fast-food outlets are typically more available than in
high-income and white communities in urban areas.35 In addition
to this, communities living in those areas had low average annual
household income. This pattern has also been observed in the UK,
where fast-food outlets cluster in areas of deprivation.2
Preference for unhealthy food is further encouraged and intensified
by the low price, as purchasing power is known to be a key
determinant in whether an individual is willing and able to pay
more for healthy food,16 and healthy food typically costs around
60% more than less-healthy food.35 Furthermore, due to the high
number of informal settlements in GP it is possible that families are
purchasing fast food as they do not have adequate utilities in the
home to cook food; this further emphasises the need for outlets that
provide healthier food options for purchase in such areas.
In many LMICs there are few regulatory frameworks preventing the
promotion of processed fast foods and sugar-sweetened beverages
(SSBs).43 However, it has become evident that policy interventions
against obesity should be directed at both the individual and
the food environment to support healthy choices, as effective
government policies and actions are necessary to increase healthy
food options.44
In targeting the food environment, healthy choices are significantly
easier to make at the individual level (rather than trying to compel
the individual alone to make healthy choices via health-promotion
and educational programmes). Such policies also tend to be more
sustainable as they affect the entire population, thus they can
concretely reverse the environmental drivers of obesity.3 Several
countries such as Ecuador, Australia, India, Brazil, Mexico45 and
Chile have implemented policies to prevent obesity (warning labels
on high-fat, sugar and salt foods). Other countries have gone even
further by increasing import and excise tariffs on SSBs and other
high-sugar products.44
Policy interventions that limit the number of fast-food outlets in
communities, and that lower the cost of healthy foods and increase
the cost of unhealthy foods, can assist in reversing the environmental
drivers of obesity.46 However, without formal structures and policies
similar to the restrictions placed on tobacco, food companies
will continue to shape and influence the polices that should be
controlling them, and the negative trajectory of fast-food expansion
will continue to result in collateral health damage.13
The mRFEI is a powerful tool for public health professionals and
provincial administrators to identify areas where access to healthy
food is limited. However, this study and the mRFEI tool also
have limitations. The study only considered the residential food
environment, and assumed that people live in the same areas in
which they work. Furthermore, assumptions were made about the
types of food sold in grocery stores. The study was also limited to
retail food outlets that could be geo-located via Google Maps. There
may have been a number of retail food outlets not included in the
analysis, in addition to food sold in the informal food sector, which
is quite significant in South Africa.
Further research should explore the links between the mRFEI and
epidemiological data such as NCD morbidity and mortality and
assess how the mRFEI differs in provinces that are not as highly
urbanised as GP. In addition, it would be worthwhile to investigate
the informal food environment in GP, especially in the areas that
were most deprived and that had no formal food retail outlets.
Conclusion
The NCD pandemic is widespread globally and is emerging as a
major public health issue in South Africa. Obesity has been identified
as a key driver, yet prevention strategies have targeted individual
behaviour-change. Policy makers need to address the structural
drivers of obesogenic environments. In addition, the available data
are often aggregated at high levels (low granularity, e.g. provincial)
thus hiding health disparities at local level. The mRFEI provides a
tool for policymakers to visualise the food environment at ward level,
allowing them to implement interventions to reduce obesogenic
environments.
The NCD burden can be prevented by addressing diet and creating
health-promoting living environments. Government should commit
to addressing unhealthy food environments by adopting a wide-
ranging, health-in-all policies approach. Municipalities can play
a fundamental role in this by introducing by-laws that limit the
number of fast-food outlets in communities. They can also zone
land using urban-planning tools and use intentional urban design
to promote citizen health. This process will necessitate multi-sectoral
collaboration with different departments and industries to ensure that
health is not negatively impacted by the activities of other sectors
such as trade and industry. Urgent action is needed to mitigate
the adverse effects of the rapidly changing food environment in
South Africa. Policy makers need to understand the structural and
environmental factors contributing to the health and wellbeing of
communities.
Acknowledgments
One of the authors (KH) was funded by the International Development
Research Centre, Grant number 108424–001.
122 2018 SAHR
References
1 GBD 2015 Obesity Collaborators, Afshin A, Forouzanfar MH,
Reitsma MB, Sur P, Estep K, et al. Health Effects of Overweight
and Obesity in 195 Countries over 25 Years. N Engl J Med.
2017;377(1):13–27.
2 Townshend T, Lake A. Obesogenic environments: current
evidence of the built and food environments. Perspect Public
Health. 2017;137(1):38–44.
3 winburn BA, Sacks G, Hall KD, et al. The global obesity
pandemic: shaped by global drivers and local environments.
Lancet. 2011;378(9793):804–14.
4 Svastisalee CM, Nordahl H, Glumer C, Holstein BE, Powell
LM, Due P. Supermarket and fast-food outlet exposure
in Copenhagen: associations with socio-economic
and demographic characteristics. Public Health Nutr.
2011;14(9):1618–26.
5 Winkleby MA, Kraemer HC, Ahn DK, Varady AN. Ethnic and
Socioeconomic Differences in Cardiovascular Disease Risk
Factors. JAMA. 1998;280(4):356–62.
6 Block JP, Scribner RA, DeSalvo KB. Fast food, race/ethnicity,
and income: a geographic analysis. American journal of
preventive medicine. 2004;27(3):211–7.
7 Morland K, Filomena S. Disparities in the availability of
fruits and vegetables between racially segregated urban
neighbourhoods. Public Health Nutr. 2007;10(12):1481–9.
8 Burns CM, Inglis AD. Measuring food access in Melbourne:
access to healthy and fast foods by car, bus and foot
in an urban municipality in Melbourne. Health Place.
2007;13(4):877–85.
9 Herforth A, Ahmed S. The food environment, its effects
on dietary consumption, and potential for measurement
within agriculture-nutrition interventions. Food Security.
2015;7(3):505–20.
10 South African National Department of Health, Statistics South
Africa, South African Medical Research Council, ICF. South
Africa Demographic and Health Survey 2016: Key Indicators.
Pretoria, South Africa and Rockville, Maryland, USA: NDoH,
Stats SA, SAMRC and ICF; 2017.
11 Battersby J, Peyton S. The Geography of Supermarkets in
Cape Town: Supermarket Expansion and Food Access. Urban
Forum. 2014;25(2):153–64.
12 Naik R, Kaneda T. Noncommunicable diseases in Africa:
Youth are key to curbing the epidemic and achieving
sustainable development. Washington, D.C.: Population
Reference Bureau; 2015. [Internet]. [cited 12 February
2016].
URL: http://www.prb.org/pdf15/ncds-africa-policybrief.pdf.
13 Gostin LO. Non-communicable diseases: Healthy living needs
global governance. Nature. 2014;511(7508):147–9.
14 Ni Mhurchu C, Vandevijvere S, Waterlander W, et al.
Monitoring the availability of healthy and unhealthy foods
and non-alcoholic beverages in community and consumer
retail food environments globally. Obes Rev. 2013;14 Suppl
1:108–19.
15 Penney TL, Brown HE, Maguire ER, Kuhn I, Monsivais P. Local
food environment interventions to improve healthy food choice
in adults: a systematic review and realist synthesis protocol.
BMJ Open. 2015;5(4):e007161.
16 Lartey A, Hemrich G, Amoroso L. Influencing food
environments for healthy diets. Rome: Food and Agriculture
Organization of the United Nations; 2016.
17 Krebs-Smith SM, Scott Kantor L. Choose a Variety of Fruits
and Vegetables Daily: Understanding the Complexities. The
Journal of Nutrition. 2001.
18 Life Sciences Research Office, Federation of American
Societies for Experimental Biology. Nutrition Monitoring in the
United States – An Update Report on Nutrition Monitoring.
Prepared for the US Department of Agriculture and the US
Department of Health and Human Services. Washington: U.S
Government Printing Office; 1989.
19 Li F, Harmer P, Cardinal BJ, Bosworth M, Johnson-Shelton
D. Obesity and the Built Environment: Does the Density of
Neighborhood Fast-Food Outlets Matter? American Journal of
Health Promotion. 2009;23(3):203–9.
20 Inagami S, Cohen DA, Brown AF, Asch SM. Body mass index,
neighborhood fast food and restaurant concentration, and car
ownership. J Urban Health. 2009;86(5):683–95.
21 Richardson AS, Boone-Heinonen J, Popkin BM, Gordon-
Larsen P. Neighborhood fast food restaurants and fast
food consumption: A national study. BMC Public Health.
2011;11(543).
22 Austin SB, Melly SJ, Sanchez BN, Patel A, Buka S, Gortmaker
SL. Clustering of Fast-Food Restaurants Around Schools: A
Novel Application of Spatial Statistics to the Study of Food
Environments. Am J Public Health. 2005;95(9):1575–81.
23 Claasen N, Van der Hoeven M, Covic N. Food environments,
health and nutrition in South Africa. Working Paper 34.
Cape Town: PLAAS, UWC and Centre of Excellence on Food
Security; 2016.
24 Boone-Heinonen J, Gordon-Larsen P, Kiefe CI, Shikany
JM, Lewis CE, Popkin BM. Fast food restaurants and food
stores: longitudinal associations with diet in young to
middle-aged adults: the CARDIA study. Arch Intern Med.
2011;171(13):1162–70.
25 Jaime P, Duran A, Sarti F, Lock K. Investigating environmental
determinants of diet, physical activity, and overweight among
adults in Sao Paulo, Brazil. J Urban Health. 2011;88:567–
81.
26 Wang R, Shi L. Access to food outlets and children’s nutritional
intake in urban China: a difference-in-difference analysis. Ital J
Pediatr. 2012;38(30).
27 Holmes T. SA’s ferocious fast food appetite. Mail & Guardian.
8 April 2016.
URL: https://mg.co.za/article/2016-04-11-sa-has-an-appetite-
for-fast-food
28 Centers for Disease Control and Prevention (CDC).
Census Tract Level State Maps of the Modified Retail Food
Environment Index (mRFEI). Atlanta, GA: Division of Physical
Activity, Nutrition and Obesity; 2011.
29 Steyn NP, McHiza ZJ. Obesity and the nutrition transition in
Sub-Saharan Africa. Ann N Y Acad Sci. 2014;1311:88–
101.
30 Ng M, Fleming T, Robinson M, et al. Global, regional, and
national prevalence of overweight and obesity in children
and adults during 1980–2013: a systematic analysis
for the Global Burden of Disease Study 2013. Lancet.
2014;384(9945):766–81.
31 Municipal Demarcation Board. 2011 Boundaries – Wards
Municipalities Centurion: Municipal Demarcation Board;
2011.
URL: https://www.demarcation.org.za.
32 Centers for Disease Control and Prevention (CDC). Children’s
Food Environment State Indicator Report, 2011. Atlanta, GA:
CDC; 2011.
33 Luan H, Law J, Quick M. Identifying food deserts and swamps
based on relative healthy food access: a spatio-temporal
Bayesian approach. Int J Health Geogr. 2015;14(1):37.
Assessment of food environments
SAHR 2018123
34 Ntloedibe M. Republic of South Africa Retail Foods. Pretoria:
Global Agricultural Information Network; 2015.
35 Michimi A, Wimberly MC. The Food Environment and
Adult Obesity in US Metropolitan Areas. Geospat Health.
2015;10(2).
36 Environmental Systems Research Institute. ArcGIS Desktop:
Release 10.5. Redlands, CA: ESRI; 2016.
37 Statistics South Africa. Community Survey 2016 Statistical
Release. P0301. Pretoria: Statistics SA; 2016.
38 Statistics South Africa. Census 2011. Statistical release
(Revised) P03014. Pretoria: Statistics SA; 2012.
39 Noble M, Zembe W, Wright G, Avenell D, Noble S. Income
Poverty at Small Area Level in South Africa in 2011. Cape
Town: Southern African Social Policy Research Institute; 2014.
40 Statistics South Africa. Census 2011. Census in Brief. Report
No 03–01–41. Pretoria: Statistics SA; 2012.
41 Whiting D, Unwin N, Roglic G. Diabetes: equity and social
determinants. Geneva: World Health Organization; 2010.
42 Rudolph M, Kroll F, Ruysenaar S, Dlamini T. The State of Food
Insecurity in Johannesburg. Urban Food Security Series No
12. Kingston and Cape Town: Queen’s University and African
Food Security Urban Network; 2012.
43 Boerma JT, Mathers C, AbouZahr C, et al. Health in 2015:
from MDGs, Millennium Development Goals to SDGs,
Sustainable Development Goals. Geneva: World Health
Organization; 2015.
44 Vandevijvere S, Swinburn B, for International Network for
Food and Obesity/non-communicable diseases (NCDs)
Research, Monitoring and Action Support (INFORMAS),. Pilot
test of the Healthy Food Environment Policy Index (Food-EPI)
to increase government actions for creating healthy food
environments. BMJ Open. 2015;5.
45 Zhang Q, Liu S, Liu R, Xue H, Wang Y. Food Policy
Approaches to Obesity Prevention: An International
Perspective. Curr Obes Rep. 2014;3(2):171–82.
46 Igumbor EU, Sanders D, Puoane TR, Tsolekile L, Schwarz C,
Purdy C, et al. ‘Big food,’ the consumer food environment,
health, and the policy response in South Africa. PLoS
medicine. 2012;9(7):e1001253.
124 2018 SAHR
