Income and obesity in an urban poor community: a cross-sectional study

Introduction

Obesity prevalence has tripled since the 1970s. Currently, it has reached epidemic proportions throughout the globe and become a significant cause of morbidity and mortality (WHO, 2020). It is a major risk factor in the development of the leading causes of global deaths, such as diabetes, cardiovascular diseases and cancer (Al-Goblan et al., 2014; De Pergola & Silvestris, 2013; Hubert et al., 1983).

Malaysia is the most obese country among the Southeast Asian nations (Ng et al., 2014). Two out of three Malaysians are overweight or obese (Aris et al., 2015). Large-scale obesity studies in the country showed a two-fold increase in the prevalence of people categorised as overweight, from 16.6% to 30%, and a four-fold increase in obesity prevalence, from 4.5% to 17.7%, in the last two decades (Aris et al., 2015; Lin et al., 2006). The main driver of obesity is an imbalance between energy consumption and expenditure (Romieu et al., 2017). However, our choices of eating and physical exercise are influenced by multiple factors and the concept of an obesogenic environment has been a topic of interest lately, that is, an environment that increases our risk of gaining weight and becoming obese (Lipek et al., 2015). Recent studies have shown complexity in the association of income and obesity: a higher income is associated with a higher risk of subsequent obesity in low- and middle income countries, but in high-income countries the reverse is observed, where those with a higher income are less likely to be obese (McLaren, 2007). The concept that richer people can afford to overeat may no longer be true among well-developed nations. The reversal hypothesis was systematically tested using individual- and aggregate-level data for 67 nations representing all regions of the world and findings were consistent with individual studies in different countries: the influence of socioeconomic status on obesity shifts from positive to negative with national income (Pampel et al., 2012).

According to income classification in Malaysia, the income groups B40, M40 and T20 represent the lowest 40%, middle 40%, and highest 20% of incomes, respectively, among the country’s population. A national large-scale obesity study in 2015 revealed a positive trend among the low-income (B40) group but a mixed trend among the middle- and high-income groups (Aris et al., 2015). Also worth noting is the high prevalence rate of obesity in states with the highest (Putrajaya) and lowest urbanization (Perlis) levels (Aris et al., 2015; Department of Statistics, Malaysia, 2010). Nevertheless, the observed trend may not be universally applicable across heterogenous rural and urban community settings, where pockets of poverty are often found.

The inconsistencies with regards to the positive and negative associations of income with obesity might be due to a lack of studies among poor communities. These communities might differ in obesity prevalence along with its associated factors, and there are no current studies that explore this phenomenon. Therefore, this study was performed to investigate the association of income and other sociodemographic factors associated with obesity among residents within the B40 income group in an urban community.

Methods

Participants

The sample size (n) was calculated using the Krejcie and Morgan formula for prevalence studies of a known population (Krejcie & Morgan, 1970). Prevalence used in the formula was the current prevalence of people who are overweight and obese for both genders, which is close to 50% (Aris et al., 2015). Out of 4229 residents, a total of 380 residents were expected to participate in the survey. Residents were approached at their home by data collectors using a systematic sampling procedure to ensure even representation of units from all the 21 floors. Data collectors were divided into 20 groups, each containing 7–8 collectors, 10 of which were assigned to building block A and 10 of which were assigned to building block B. Each group was assigned 2–3 floors of the 21-floor building block and were instructed to carry out systematic random sampling on their floors. If residents of one house did not consent, the data collectors moved to the next house. Within the selected units, all the residents were invited for an interview. The inclusion criteria included: (1) residents of the PPR; (2) Malaysian citizens; (3) aged =/>5 years old. Physically disabled/bedridden individuals were excluded (Figure 1).

Figure 1. Participant flow chart

Results

There was a total of 380 participants, of which 341 completed anthropometric measurements. Of these, there was a total of 328 respondents, including 261 adults and 80 children (aged below 18 years old) from 209 household units, who provided full sociodemographic information (Galvan, 2020). Table 1 shows the sociodemographic characteristics of the 328 participants. Among the 261 adult participants that completed anthropometric measurements, 29.1% and 45.6% were overweight and obese, respectively (Table 2). Among the respondents aged below 18 years, the prevalence of an overweight and obese BMI was 13.8 % and 20%, respectively (Table 3).

Discussion

The obesity prevalence in this study was significantly higher compared to the national prevalence (Aris et al., 2015). Contrary to the results from the national survey (Aris et al., 2015), income was negatively associated with BMI after controlling for other variables in this low-income (B40) population group; in other words, the individuals with a higher income had a lower BMI. The negative association was similar to that of high-income countries, where a socioeconomically developed environment may possibly contribute to the reversal (Pampel et al., 2012). In contrast to the built environment in rural areas, families with a higher income among this community may be able to afford accessible healthy food choices along with various available weight loss programs, walkable entertainment centres, shopping malls and healthcare facilities (Ellis et al., 2016); thus, they are less obese as compared to the rest of the community. This supportive environment decreases the risk of obesity for urban dwellers, while the concept of being able to afford to overeat may still be true among wealthy rural dwellers (Creatore et al., 2016). Our results confirm the arguments discussed in the review of 2,009 population-based studies done by the NCD Risk Factor Collaboration (Bixby et al., 2019). The researchers found that more than 55% of the rise in global obesity was from rural areas. Their findings negate the assumptions that urbanization was the main driver of the obesity epidemic. Assumptions of sedentary lifestyles with access to processed food in urban places, and the general preconceived notion of manual labor and healthy eating from one’s own garden among rural dwellers may not be true. Automation has influenced our agricultural activities and distribution or shipment of processed food to rural areas. (Bixby et al., 2019) In the national survey done by Aris et al. (2015), both the states with highest and lowest urbanization levels had high prevalence rates of obesity. Putrajaya, with 100% urbanization, reported a 43% rate of obesity and Perlis, with 51.4% urbanization, reported a rate of 36% rate of obesity, ranking 2^nd together with the state of Melaka. This report contradicts the general idea that obesity applies to urban communities and malnutrition to rural communities. In our report, this urban poor community has a greater prevalence of obesity compared to the general population but the distribution of obesity among them was skewed towards those with a low income. The driver could possibly be that the advantages of the available supportive environment, such as gyms and healthy food choices, in the urban-built environment are missed by those with a lower income.

Conclusion

The obesity rate in this urban poor community is high. The monthly household income is negatively associated with obesity; people with a higher income have a lower BMI.

To understand further the complexity of the relationship between income and obesity, we recommend further investigations involving more than one PPR community. Determining monthly food and health expenses of the families will increase the precision of information on the household income level. Additionally, a wider study could compare trends between urban and rural poor communities.

Limitations

Our findings are limited to one community, and the cross-sectional nature of this study does not permit any temporal relationships to be deduced. Monthly household income disclosed to the interviewers may not reflect the actual figures; the residents in this community are recipients of the government housing project and therefore represent the poorest 40% of the population and should receive a salary of ~3,000 ringgit or below. However, salaries higher than 3,000 ringgit were reported in this study.

Data availability