Do trends in the prevalence of overweight by socio-economic position differ between India’s most and least economically developed states?

We used the National Family Health Survey (NFHS) Surveys 2 (1998–99), 3 (2005–06) and 4 (2015–16). All three surveys collected health and demographic data on women aged 15–49, whereas surveys 3 and 4 collected data on men aged 15–54. The sampling method was designed to include a nationally-representative sample of individuals within a nationally-representative sample of households. Additionally, in India, the NFHS surveys are also representative at the level of the state.

The NFHS surveys select rural and urban samples separately. Specifically, in rural areas in all three waves analysed, rural samples were selected using two-stage sampling, whereby the first stage involved selecting primary sampling units (PSUs), or villages, with a probability proportional to size (PPS), and the second stage involved selecting random households from each village. In urban areas, NFHS 2 and 3 used a slightly different sampling procedure to the one in NFHS 4. In NFHS 2 and 3, three-stage sampling was adopted whereby in the first stage wards were selected with a PPS, in the second random census enumeration blocks (CEB) were chosen in each ward and, in the third, random households were chosen from each CEB [2, 3] On the other hand, NFHS 4 adopted a two-stage approach in urban areas, whereby CEBs served as the PSU, selected using a PPS, and households from each PSU randomly selected. Were a PSU to contain fewer than 40 households, the PSU was joined to the nearest PSU. The 2011 census helped determine the sampling frame in NFHS 4 [1].

In all three surveys Interviews used a uniform questionnaire and were conducted by survey teams. A woman’s eligibility for the survey was determined by whether they were between ages 15–49 and, for the NFHS 3 and 4, whether they spent the previous night in the selected households. Men aged 15–54 in the households were eligible for the Men’s survey in NFHS 3. Of the selected households in NFHS 4, a random sample of households were selected to determine eligibility for the men’s survey [1].

In India there are currently 36 States/Union Territories. We restricted our analysis to states that have been in existence since the collection of the NFHS 2 survey. States created between the surveys were not considered in the analysis. We selected five states to indicate the most and least developed states as the study aimed to demonstrate a divergence in the trends in their socioeconomic patterning. Our primary objective was to highlight variation in trends in the socioeconomic patterning of overweight within India. We therefore chose not to include all the states in India as this would lead to the inclusion of states that are closer to the average level of per capita net state domestic product for India. As a result, we would risk placing states at similar levels of economic development in the Most and Least developed states categories, consequently underestimating the extent of the variation in trends.

Our classification of states was based on the per capita net state domestic product (PCNSDP) in 2014–15 using the base year 2011–12. The most economically developed states were Goa, Maharashtra, Sikkim, Haryana and Kerala with a PCNSDP ranging from ₹112,444 to ₹241,081, compared to an all India average of ₹72,805. The least economically developed states included Bihar, Assam, Uttar Pradesh, Manipur, and Madhya Pradesh with NSDPPC ranging from ₹23,223 to ₹44,809 [20]. We limited our sample to non-pregnant women, whose inclusion could bias the associations we sought to identify. This left a total of 96,365 women and 18,729 men in the most developed states category, and 289,200 women and 54,669 men, respectively, in the least developed states category.

As NFHS-2 only sampled ever-married women, we restricted our samples in 2005–06 and 2015–16 to this population to allow the comparability of the study population across surveys. Additionally, respondents with missing height and weight data were also omitted from the sample, leaving 76,050 women (12,168 in 1998–99; 14,000 in 2005–06; 49,882 in 2015–16) and 18,729 men (8518 in 2005–06 and 10,211 in 2015–16) as the study population in the most economically developed states, and 213,195 women (22,266 in 1998–99; 20,459 in 2005–06; and 170,470 in 2015–16) and 54,669 men (19,377 in 1998–99; and 35,292 in 2015–16) in the least economically developed states. As multi-stage sampling approaches were adopted in the collection of the NFHS, we included the sampling weights included in the data set to account for unequal selection probabilities.

We used the Body Mass Index (BMI) variable included in the surveys (measured as the respondent’s weight divided by the square of their height) to separate individuals into two groups: overweight (BMI over 24.99 kg/m²), and not overweight (BMI 24.99 kg/m² or under). This categorisation is based on the WHO’s recommended cut-offs for BMI classification [5]. Rather than split the continuous BMI measure into multiple subcategories of overweight, we used this classification as the main aim of the paper was to analyse trends in excess adiposity, and research has found an elevated risk of NCDs and mortality beyond a BMI of 24.99 kg/m² [21, 22]. We did not use a continuous measure of nutritional status, as observed population-level increases in BMI we would expect to observe over the study period could be driven by a both individuals moving into overweight categories, and individuals moving from underweight to normal weight; the latter of which does not capture increases in excess adiposity.

Height and weight information on women aged 15–49 in NFHS-2, 3 and 4, and men aged 15–54 in NFHS-3 and 4, were collected by specially trained investigators. A solar-powered SECA digital scale was used to measure the weight of respondents, with the NFHS-2 report claiming an accuracy of ±100 g. The height of respondents in NFHS-2 and 3 was measured using a measuring board designed for use in survey data collection. In NFHS-4, the Seca 213 stadiometer was used to collect respondent’s height information [1‐3].

Our final models were adjusted for the respondent’s age (15–29; 30–39; and 40–49 (40–54 for men)) and marital status. Marital status was categorised as either ‘currently married’ or ‘not currently married’ and was included as married individuals have been found to be at higher risk of being overweight [26]. We would have also preferred to control for the respondent’s occupation. Higher prevalence of overweight may be expected to be observed among individuals in more sedentary jobs [6‐9], and sedentary labour may be expected to be more prevalent among higher SEP individuals. However, it was not possible to control for occupation in our research due to the fact that it was collected on a very limited subsample of the respondents in NFHS 4 (approximately 5% of women in the NFHS-4 national sample).

Results of our sensitivity analysis are presented in Table 3, and in Figs. 4 and 5. Using the SoL index as the main exposure, shows a similar trend of a notable convergence of overweight prevalence across SEP among women, particularly in urban areas of India’s most developed states. Additionally, it supports the considerably more mixed trend of overweight patterning among men we identified when using education as the exposure of interest.

This study has found that the trends in the socioeconomic patterning of overweight in India varied between India’s most and least developed states between 1998 and 2016. When examining the difference between overweight prevalence in the highest and lowest SEP groups, we found a converging trend of overweight prevalence across SEP among women between 1998 and 2016. As expected, this trend amongst women was more pronounced in India’s most developed states, particularly in urban areas, however, similar trends were observed in the least developed states and in rural areas. The converging trend amongst women appears to be driven by relatively smaller increases, and in some cases a decline, in the prevalence of overweight in higher SEP groups compared to lower SEP groups. This convergence appears to be limited to women, as amongst men, we did not identify any notable convergence in the socioeconomic patterning of overweight between 2005 and 2016. Using SoL as the main exposure of interest, we found similar, albeit a more attenuated convergence in the socioeconomic patterning of overweight.

Few studies have examined sub-national variation in the association of SEP and overweight in countries that have undergone rapid economic development. The only similar study we identified in India used a standard of living index as the primary exposure and reported a similar converging trend in the prevalence of overweight across SEP in states with a high overall prevalence of. On the other hand, they identified a diverging trend in Standard of Living in states with a high prevalence of underweight amongst women between 1998 and 2006 [27]. Our more up-to-date examination of sub-national trends in additional subpopulations suggests a more nuanced picture of the socioeconomic patterning. Notably we find no evidence of a diverging trend of overweight across SEP in any of the subpopulations, and that there are notable differences in the trends between the sexes and urban and rural areas.

Although there still remains a positive association between SEP and overweight across all the subpopulations we analysed in India, studies in other countries have identified a negative association between SEP and excess weight in more economically areas of countries that have undergone rapid economic development. One study in Brazil found a positive association between obesity and per capita household income in both more and less economically developed regions of Brazil in 1974/75. By 1996/97 the association was negative in the more economically developed regions, whereas the positive association in the less developed regions persisted [28]. This suggests that Brazil’s more developed regions in 1996/97 may have been at a more advanced stage of the epidemiological transition than India’s most developed states currently. Other studies using measures of household income and educational attainment as the primary exposures and focusing on women in China’s most economically prosperous regions have also found a negative association between SEP and prevalence of overweight [29, 30].

We also identified a particularly notable convergence in the prevalence of overweight by SEP among women when compared to men. Other studies have identified similar differences by sex. One study in China found high-income men and women with low education to be at highest risk of obesity in an economically prosperous province [29]. Another study in China found that higher education was associated with lower odds of overweight among women and higher odds of overweight among men [30]. In South Korea, a country that experienced a remarkable pace of economic growth in previous decades [31], one study still found a positive association of income with obesity among men, whereas they found a negative association among women [32].

The increased capacity of higher SEP individuals to afford to consume excess food [6‐9] is a commonly suggested reason as to why the association between overweight and SEP is positive in low- or low-middle-income countries like India. However, the smaller difference in overweight prevalence between lower and higher SEP women in India’s most developed states, particularly in the most recent period, may be due to an increased ability to afford expensive healthy foods and an increased level of health consciousness among higher SEP individuals [6, 33, 34]. On the other hand, particularly in India’s most developed states, lower SEP individuals may be increasingly able to afford cheap high calorie fatty foods [6, 35].

Our study has some limitations. Firstly, we would have ideally liked to have used additional indicators of excess adiposity to complement our findings. BMI may be limited in that it cannot distinguish between lean mass and body fat, nor does it have information on the distribution of body fat, potentially making it an inaccurate measure of central adiposity [36]. However, other studies have shown a strong correlation of BMI with measures of central adiposity among Indians, such as waist circumference [37]. Consequently, we would not expect the trends we report to vary considerably between adiposity measures. Another possible limitation associated with our use of the BMI variable to inform our main outcome of interest is the potential difference in the body fat percentage at any given BMI between higher and lower SEP groups. Research amongst children from higher income countries have shown that lower SEP groups may have a higher percentage of body fat at any given BMI compared to higher SEP groups [14, 38]. Although this may be limited to high income societies, we are unable to verify the association of body fat and BMI in our data as the NFHS does not collect body fat information. Were a similar phenomenon observed in India, this would imply a more rapid convergence in the socioeconomic patterning of overweight in India than we have reported.

We limited our study population of women in 2005–06 and 2015–16 to ever-married women, as this was the sampled population in 1998–99. However, a slightly higher proportion of the sample was never-married in 2015–16 (22.5%) than in 2005–06 (19.5%). Additionally, the prevalence of overweight is lower among never-married than in currently married women (prevalence of overweight was 6.6% and 25% among never married and currently married women, respectively, in 2015–16). This may have led us to potentially overestimate the prevalence we reported for 2015–16 and therefore underestimate the extent of convergence overweight prevalence across SEP (see Additional file 1: Table S3).

There are some slight differences in the rankings of states by PCNSDP in 2005–06 and 1998–99 compared to in 2014–15. For example, in 2005–06, the state of Odisha had a slightly lower PCNSDP than Manipur. Additionally, Gujarat’s economy is 19% larger than Sikkim’s in 2005–06, however, Sikkim’s economy almost tripled within a decade [20]. These discrepancies are however, unlikely to change the overall message of the study, and instead inclusion of these states is expected make results more relatively conservative.

Another limitation of our study involves our use of the standard of living index based in part on the ownership of assets. Common criticisms of an asset-based index like the one we used, includes the fact that it makes little accommodation for the quality of assets [24, 39], potentially leading to misclassification of households. For instance, televisions in poorer households may only receive terrestrial transmission, whereas in higher SEP households may receive digital transmission. Despite this, the simple collection of asset ownership information is not expected to affect the variable substantially [39]. Additionally, as we used three broad SoL categories across a large data set, any misclassification is not expected to be substantial. Another criticism of asset indices is that certain assets are likely to have different importance between broad geographical areas. Although we attempted to remedy this to an extent by calculating separate SoL indices in urban and rural areas, the importance of some assets may still differ between other geographical levels of aggregation [18]. Despite these issues, asset-based indices offer an affordable and stable long-term measure of household wealth for large surveys in low-income settings [24]. Furthermore, our use of two different measures of SEP in this study ensures that we have captured a large portion of the avenues through which SEP and overweight are associated.

Finally, the cross-sectional nature of the data we used did not enable us to draw conclusions about the causal relationship between overweight and SEP. Although this was not an explicit study aim, such information may have enriched our understanding of the reasons as to why overweight is more prevalent among particular socioeconomic groups in India.

Despite these limitations, we use the most recent state representative data, making our findings both generalisable and the most current estimates of these trends.

Some have suggested that overweight is a ‘disease of affluence’ in low and low-middle income countries [40, 41]. We find evidence of a much more nuanced picture of the socioeconomic patterning of overweight, when we examine sub-national trends. Whereas it may be an appropriate description of the positive associations between SEP and overweight we identified, were the identified trends to continue especially among women in India’s more economically developed states, there may be a negative association in the coming years. We find no evidence that were past trends to continue, there would be any change to the socioeconomic pattering among men.

The markedly higher increase in overweight among lower SEP Indians will be an important consideration in the near future as state governments are already tasked with tackling the burden of infectious diseases within this demographic. A state-specific approach will be needed to face the challenge of raising general access to staple foods whilst simultaneously trying to lower demand for unhealthy foods [27, 42‐44]. Additionally, attempts to close the difference in the association of overweight with SEP between men and women may wish to focus on improving health-related behaviors among men.