Chapter 3: Theory and evidence-based development of a weight loss intervention for
3.2 Methods and outcomes…
3.2.5 Stage 1: Identify target behaviours for weight loss
A. Review of existing research for identification of target behaviours for weight loss in primary school educators employed at public schools in low-income settings in the Western Cape
Identification of studies
A literature search was conducted in May 2019 on PUBMED using the search terms school teachers [MeSH] OR teachers OR educators AND Weight reduction OR weight loss OR dietary OR diet OR nutrition OR smoking OR exercise OR physical activity OR health interventions OR health behaviour change OR health behaviour change OR health education OR health promotion OR health promoting behaviours OR health awareness OR health programs OR health programmes.
The only comprehensive research in South Africa that provides insights in not only the health and weight status of educators, but also their lifestyle behaviours and body shape perceptions, is the HealthKick (HK) survey conducted by the MRC. [de Villiers et al., 2012]. Only one other survey, conducted in Brazil, reported on the weight status of educators with limited insights on lifestyle behaviours [Serra et al., 2015] and therefore did not provide insights for consideration in the intervention development process.
The MRC HK study
Background: In 2007, the MRC initiated a primary school diabetes prevention programme called HK which was aimed at learners. As part of the HK study, a situational analysis was conducted at 100 urban and rural disadvantaged schools in low socio-economic areas in the Western Cape (50 schools in the Cape Metropole and 50 schools in the Overberg/Breede River districts). Results showed that educators constituted an important environmental influence on children. The HK study was therefore expanded to include a health risk survey of educators employed at the participating schools [de Villiers et al., 2012].
NCD risks and weight perceptions of educators: Participating school principals identified lack of physical activity (33%) and NCDs (24%) as the main health priorities of primary school educators [de Villiers et al., 2012]. Findings showed that they were a high-risk group for NCD development. More than a third (37%) of males and 27% of females were overweight, while 35% of males and 55% of females were obese. In addition, 38% of males and 67% of females had a high waist circumference
101 (WC ≥ 102 centimetres in men, WC ≥ 88 centimetres in women) and 46% were hypertensive [Blood Pressure ≥ 140/90 millilitres mercury (mmHg)]. High non-fasting cholesterol levels were found in 30%
and high non-fasting blood glucose levels in 29% of the educators [Senekal et al., 2015a]. These figures exceed those reported for the general population in the South African National Health and Nutrition Survey (SANHANES) [Shisana et al., 2013]. It was further also evident that many educators were likely to underestimate their body weight, as well as being unaware of their personal health risks such as being overweight or having hypertension, high cholesterol or high blood glucose [Senekal et al., 2015a]. Senekal et al. (2015a) argue that this may reflect a lack of interest in health by the educators, which would need to be considered in the development of the weight loss intervention.
Dietary patterns of educators: The HK educator survey also investigated food choices of educators.
These were identified using a non-quantified indicator food frequency questionnaire that reflected the frequency of consumption of poor and healthy food choices in the past “week”. The food list was compiled to reflect foods/drinks/snacks associated with the development/prevention/management of obesity, diabetes and other NCDs, as well as cultural food choices (Seme et al., 2017). For data analysis purposes, the indicator foods were assigned to six food categories, namely fruit and vegetables (oranges/naartjies; apples; bananas; pears; green vegetables including spinach, peas, beans and broccoli; orange/yellow vegetables including sweet potato, pumpkin, butternut and carrots;
mixed vegetables; cabbage; cauliflower and lettuce.); high fat foods (processed meats; tinned meat;
chicken with skin; yellow cheese; margarine; fried foods, including chips, fat cakes, fish, chicken; take outs e.g. McDonalds; pies, sausage rolls, and samosas); energy-dense snacks/items (sugar, chocolate, sweets, cake, biscuits, doughnuts, crisps, jam, syrup, honey); white bread; cereal and legume fibre (breakfast cereals including all bran flakes, muesli, Weetbix, oats; brown bread and legumes) and processed (salty) foods (tinned meat/fish, margarine/butter, bread, cheese, peanut butter/peanuts, cakes, biscuits, doughnuts, crisp, take outs e.g. McDonald’s, pies, sausages rolls, samosas). Of note is that the frequency of intake of some foods allocated to the salty food group may have also been counted in one of the other 5 groups [Seme et al., 2017].
Educators reported eating fruit and vegetables choices twice (combined; classified as healthy choices) a day; legumes, cereals once a day (classified as healthy choices), white bread 3-4 times a week (classified as unhealthy choice), high fat foods twice a day (classified as unhealthy choices), energy dense snack choices once a day (mostly sugar, and sugar containing items, thus classified as unhealthy choices) and salty food choices three times a day (classified as unhealthy choices). These food choices reflect a dietary pattern where unhealthy food choices seem to be made more frequently than healthy food choices, with specific concerns being low intake of fruit and vegetables and high intake of fat and sugar rich and salty foods [Seme et al., 2017]. Educators’ understanding of healthy eating also seemed to be very limited. When asked what you need to eat to be healthy vegetables
102 was mentioned by 54%, fruits by 41.9%, proteins by 29%, fats by 14.6%, dairy by 8% and starches by 6% [Seme et al., 2017].
Physical activity levels of educators: Physical activity was assessed in the HK educator survey using the Global Physical Activity Questionnaire (GPAQ) that was developed by the WHO for the quantification of energy expenditure in subjects in developing countries [Armstrong & Bull, 2006]. The GPAQ focuses on physical activity which takes place in three settings (activity at work; travelling to and from places and recreational activities, as well as on sedentary behaviour. Metabolic equivalents (METs) per week were calculated and used to classify physical activity levels as high, moderate and low. Results show that the majority (86.7%) of educators were classified as being moderately active and 14.3% as having low levels of physical activity. Females and those older than 50 years were significantly more likely to be classified as having low levels of physical activity [Seme et al., 2017].
When interpreting these results, it should be noted that the major contributor to METs in the HK educator survey was activity in work time (1313 METs at work versus 224 METs for transport and 137 METs for recreational activity) [Seme et al., 2017]. Seme et al. (2017) argued that the high METs spent during work time could be explained by the possibility that educators spend a large part of their work time standing and walking (designated as moderate intensity activity in the GPAQ). The reported time spent by teachers in sedentary activity (3.4 hours per day) further supports the notion that physical activity levels, and thus classification of the majority as moderately active, could have been over estimated.
B. Salient beliefs of primary school educators employed at public schools in low-income settings in the Western Cape relating to healthy dietary and physical activity behaviours As mentioned, it is essential that behaviour change theories should be considered in the development of interventions aimed at improving behaviours for weight loss. Many studies, including weight loss studies, have used the TPB to understand the individual decision-making processes that underpin behaviour change [McEachan et al., 2011]. Azjen (1991) asserted that if one understood beliefs, interventions or intervention components could be tailored to target salient beliefs to attain sustained behavioural change.
For these reasons Steyn et al. (2014; Unpublished data) conducted an elicitation study to gain insights into the beliefs primary school educators employed at public schools in low-income settings in the Western Cape hold regarding fruit and vegetable, sugar, fat and salt intakes, as well as physical activity. Salient beliefs that were identified were further investigated in a cross-sectional survey using a Theory of Planned Behaviour questionnaire (TPB-questionnaire). The elicitation study comprised five focus groups (total n= 53 educators from schools in the target areas). Questions in the focus group guide covered beliefs educators and their families may hold regarding the mentioned
103 behaviours, as well as their beliefs regarding facilitators of and barriers to the execution of these behaviours. An expert panel (Senekal, Steyn, de Villiers, Evans, Booley and the PhD candidate), considered the salient beliefs that emerged from the focus group data. A list of 50 beliefs for inclusion in the TPB-questionnaire was generated based on insights of the panel in cultural, dietary and physical activity practices and other considerations in the target communities. Frequency of reporting in the focus groups was also considered bearing in mind the recommendation by Kreuger and Casey (2009:121) that “Although we pay attention to how frequently something is said, it is a mistake to assume that what is said most frequently is most important. Sometimes a key insight might have been said only once in a series of groups. You have to know enough about what you are studying to know a gem when it comes along. One person may be a visionary thinker and identify something that no one else has spotted or thought about yet.”
The expert panel familiar with the TPB and the development of TPB-questionnaires (Senekal, Steyn, de Villiers, Evans, Booley and PhD candidate), developed a belief-statement for each belief on the final list of 50, using the guidelines published by Azjen (2011). Beliefs were classified as behavioural, normative or control as outlined in the TPB [Azjen, 2011] (Addendum I, page 282).
The beliefs relating to the dietary and physical activity behaviours were mostly control beliefs relating to facilitators of and barriers to performing the behaviours. Behavioural beliefs that emerged related mostly to potential health benefits of the behaviours. Very few normative beliefs emerged and those that were identified mostly related to physical activity. The belief statements were pilot tested on five educators to ensure each statement was clearly written and understood as intended and it was then adapted accordingly if required. In the TPB-questionnaire respondents had to indicate their level of agreement with the statements on a Likert scale of 1-5 with 1 denoting the strongest disagreement, 3 being neutral and 5 denoting strongest agreement.
The final TPB-questionnaire was administered on 164 educators from schools in lower socio- economic areas in the Cape Metropole (median BMI was 30.5 kg/m2) [Senekal et al., 2015b;
Unpublished data]. The median interquartile range (IQR) score for each belief was calculated and considered in a final interactive session with the expert panel (Senekal, Steyn, de Villiers, Evans, Booley and PhD candidate), to identify messageable beliefs for targeting in a lifestyle weight loss intervention for primary school educators. Beliefs with a median score that allowed for a shift, either up or down, depending on whether the belief statement was formulated in a negative or positive format (scores of ≥ 4 or ≤ 1) and were deemed to have good potential to change, were retained. An example of this included the belief, ‘I would eat vegetables even if at times they look unappealing,’ which was deemed to have potential to change, while the belief, ‘Most people who are important to me eat fruits and vegetables every day’ was deemed to have little or no potential to change.
104 The final list of 24 beliefs that were considered messageable and could be targeted in the intervention was compiled (Table 3.2). Eighteen were control beliefs and 6 were behavioural beliefs. No normative beliefs remained after this final selection.
Table 3.2: Final list of messageable beliefs (24) regarding fruit and vegetable intake, sugar intake, fat intake and physical activity to be targeted by the intervention
Beliefs regarding fruit and vegetable intake (6 beliefs) Preparation of vegetables does not take
a long time (Control).
Fruits and vegetables are affordable (Control).
I am confident that I can eat the recommended amounts of fruit and vegetables every day (Control).
Eating fruits and vegetables every day will help me lose weight/ control my weight (Behaviour).
I would eat vegetables even if at times, they look unappealing (Control).
Fruit and vegetables are easy to find in stores nearby (Control).
Beliefs regarding sugar intake (4 beliefs) I turn to sugary foods/snacks/drinks
when I am stressed (Control).
Reducing the amount of sugary foods/snacks/drinks I eat and drink will make me feel unwell (moody or have a headache or tired) (Behaviour).
I have poor awareness of the sugar content in the food/drinks I eat/drink (Control).
I can reduce the amount of sugary foods/snacks/drinks I eat and drink (Control)
Beliefs regarding fat intake (7 beliefs) Eating less fat will help reduce the risk
of diseases e.g. heart disease (Behaviour).
Low-fat/healthy fat options are expensive (Control).
It is easy to exclude high-fat foods from my daily diet (Control)
Decreasing the amount of fat I eat will help me lose/ control my weight (Behaviour).
Healthy takeaways and/or street foods are easy to find in my surroundings (Control).
I do not have enough time to prepare healthy meals regularly (Control).
Low fat/ fat-free foods taste good / are tasty (Behaviour).
Beliefs regarding physical activity (7 beliefs) Being physically more active will make
me feel better about my appearance (Behaviour).
I could increase my physical activity levels (be physically more active) even if I were tired (Control).
There are no accessible, safe, affordable opportunities for me to be physically active (Control).
Knowing more about different types of physical activity I can do will help me to be more active (Control).
Finding time to be physically more active is possible (Control).
I can increase my levels of physical activity (be physically more active) (Control)
Having an exercise “buddy” will help me to be physically more active (Control).
C. Implications for intervention developments
There is consensus amongst leading health authorities that a comprehensive approach, which integrates healthy dietary intake, increased physical activity components and includes measures to support behaviour is the cornerstone of treatment for overweight or obese individuals [Wadden et al., 2020; Wharton et al., 2020; Kushner, 2018; Alamuddin et al., 2016; Bray et al., 2016; Raynor &
Champagne 2016; Yumuk et al., 2015; Jensen et al., 2014;].
105 Results from the only relevant study, the HK educator survey clearly showed that the following dietary behaviours would need to be addressed in a weight loss programme for educators: low intake of fruit and vegetables and frequent intake of foods/snacks/drinks high in sugar and fat. It is evident that these poor food choices are not unique to educators. As part of the 2012 SANHANES, scores were generated for fruit and vegetable, sugar and fat intake to investigate quality of food choices. With regards to fruit and vegetables, a quarter of both males and females had a poor intake (25%) while only 29% had a high fruit and vegetable intake. Approximately 20% of participants had a high sugar score (19.3% males and 20% females), with similar findings for the fat intake (18.7% males and 18%
females) [Shisana et al., 2013].
A further behaviour that evidently also requires attention in a weight loss programme for educators is the low to moderate levels of physical activity. Results from the 2012 SANHANES show that low levels of physical activity also characterise the general population, where one in four males (27.9%) and 45.2% of females were reported to be unfit [Shisana et al, 2013].
There is a paucity of information on beliefs educators hold relating to healthy lifestyle behaviours. The list of messageable beliefs was developed for integration as relevant in the BCW steps.
3.2.6 Stage 2: Understanding the behaviours