Skip to main content
PMC home page
SSM - Population Health logoLink to SSM - Population Health
. 2016 Dec 28;3:172–178. doi: 10.1016/j.ssmph.2016.12.012

Socioeconomic inequality in morbid obesity with body mass index more than 40 kg/m2 in the United States and England

Helen P Booth 1,⁎, Judith Charlton 1, Martin C Gulliford 1
PMCID: PMC5769007  PMID: 29349213

Abstract

Introduction

This study evaluated socioeconomic inequality in morbid obesity (body mass index, BMI ≥40 kg/m2) through an analysis of population health survey data in the United States (US) and England (UK).

Methods

We analysed data for the National Health and Nutrition Examination Survey and the Health Survey for England for 2011 to 2014. Age-adjusted odds ratios (AOR) were used to evaluate income- and education-inequality.

Results

There were 26,898 eligible UK and 10,628 US participants. Morbid obesity was more frequent in women than men, and higher in the US than the UK (men: US, 4.8%; UK, 1.7%; women US, 9.6%; UK, 3.7%). In the UK, morbid obesity showed graded income-inequality in both genders (AOR, for lowest income quintile: men, 1.83, 95% confidence interval 1.16 to 2.88; women, 2.18, 1.55 to 3.07), as well as education-inequality (AOR for no school qualifications, men 2.57, 1.64 to 4.02; women, 2.18, 1.55 to 3.07). In the US, morbid obesity showed a consistent gradient only for income in women (AOR for lowest income quintile 1.97, 1.19 to 3.25). When compared with all other US groups, having college education (AOR, men, 0.56, 0.29 to 1.08; women, 0.36, 0.22 to 0.60) or household income ≥$75 000 (AOR, men 0.52, 0.27 to 0.98; women, 0.51, 0.33 to 0.80) appeared to protect against morbid obesity.

Conclusions

Morbid obesity is associated with lower socioeconomic status in men and women in the UK. In the US, morbid obesity was twice as prevalent, but less strongly associated with socioeconomic status, suggesting that morbid obesity may now have spread to all but the highest socioeconomic groups.

Abbreviations: (NHANES), National Health and Nutrition Survey; (HSE), Health Survey for England; (CSE), certificate of secondary education; (AOR), age-adjusted odds ratio; NCD, Non-communicable disease

Keywords: Obesity, Morbid, Income, Education, Socioeconomic position, Health surveys

Highlights

  • •

    This study evaluated the socioeconomic patterning of morbid obesity in the UK and US.

  • •

    Inequalities in morbid obesity can be observed in men and women in the UK and US.

  • •

    A graded association between morbid obesity and SES is clearer in the UK than the US.

  • •

    This finding may be explained by different obesity prevalence in the two countries.

1. Introduction

1.1. Background

Obesity is a major global health problem (NCD Risk Factor Collaboration, 2016a) with important implications for population health (NCD Risk Factor Collaboration, 2016b). People with morbid obesity (body mass index, BMI ≥40 kg/m2) are disproportionately affected by the health consequences of obesity, often experiencing the premature onset of multiple morbidities (Booth et al., 2014b). Diabetes is particularly important, developing in up to 3% per year (Booth, PrevostGulliford, 2014a).

Morbid obesity affects 0.64% of men and 1.6% of women worldwide (NCD Risk Factor Collaboration, 2016a, NCD Risk Factor Collaboration, 2016a) but the prevalence of morbid obesity is considerably greater in high income countries, where the rate of increase has been very rapid. In England, 0.2% of men and 1.4% of women had morbid obesity in 1993, but by 2014 morbid obesity affected 1.8% of men and 3.6% of women (Joint Health Surveys Unit, 2014). In the United States, morbid obesity increased from 3.9% of the population in 2000 to 6.6% in 2010 (Sturm & Hattori, 2013).

1.2. Socioeconomic status and obesity

The rise in obesity appears to result from changes in the social environment that facilitate the development of obesity in susceptible individuals. Social environmental exposures may be differentially distributed across socioeconomic groups with men and women showing differing patterns of association. Previous studies demonstrate an important gender distinction in the association of socioeconomic status with simple obesity (BMI≥30 Kg/m2). In their seminal review, Sobal and Stunkard, (1989) showed that in high-income countries obesity was associated with lower socioeconomic position in women, but this pattern of association was not generally observed in men. This is in contrast to the situation in low- and middle-income countries where obesity may be associated with affluence. Recent empirical studies have confirmed the observation of Sobal and Stunkard, that a socioeconomic gradient in obesity is generally only observed in women in high-income countries. (Devaux and Sassi, 2013, García Villar and Quintana-Domeque, 2009, Sassi et al., 2009) McLaren (2007) analysed 1914 estimates from 333 published studies. In women from high income countries, obesity was negatively associated with occupational level for 100/146 (68%) of estimates. In men, obesity showed either no association or non-linear associations with education (35/50, 70% of estimates) or employment level (28/33, 85% of estimates) (McLaren, 2007).

The causal mechanisms underlying the association of obesity with lower socioeconomic status may be complex and possibly bidirectional (Department of Health Public Health Research Consortium et al., 2007; Finkelstein, Ruhm & Kosa, 2005). Lower availability and affordability of healthy foods (Drewnowski, 2009) and lower participation in physical activity (Beenackers et al., 2012) may be important factors in lower socioeconomic groups. Increased susceptibility to poverty, and to the effects of poverty, in women may play a role in gender differences. Reverse causation may also contribute to gender inequalities; an example of which is increased discrimination against overweight and obesity in the workplace, a phenomenon that appears to impact more heavily on women than men (García Villar & Quintana-Domeque, 2009).

1.3. Hypotheses and rationale for the study

Cross-country comparisons may offer important insights into the origins and determinants of population health and inequalities in health (Mackenbach et al., 2008). There may be substantial differences in health outcomes, and inequalities in health measures, even among countries with broadly similar aggregate levels of economic achievement (Wilkinson, 1997). Devaux and Sassi (2013) compared the prevalence of obesity in 11 OECD countries, showing that socioeconomic inequalities in obesity were greater than for overweight, and greater in women than men. In a recent study, which contrasted the U.S. and Canada, Siddiqi, Brown, Nguyen, Loopstra, and Kawachi (2015) suggested that the association of educational-level with all obesity may vary across countries. In Canada, having less than high school education was associated with obesity, while in the US all groups except for the college-educated were obese. Previous studies have not evaluated the social patterning of morbid obesity. Morbid obesity is a particular concern to public health because it is associated with disproportionately large health impacts and costs (Arterburn, Maciejewski & Tsevat, 2005; Rudisill, Charlton, Booth & Gulliford, 2016).

The United States (US) and England (UK) have among the highest rates of obesity, and morbid obesity, in the world. The two countries share cultural similarities and a ‘liberal’ economic system (Bambra, 2007, Hall and Soskice, 2001); the US is more affluent overall but access to services and social protection may often be more favourable in the UK. Analysing data from these two OECD countries offers an opportunity to compare the social patterning of overweight and obesity at different levels of overall prevalence. This analysis is timely in the context of the continued global rise of morbid obesity.

This study aimed to evaluate income- and education-related inequalities in morbid obesity through a comparison of national population health surveys from the UK and the US. We hypothesised that socioeconomic inequalities in morbid obesity may be more consistent than for all obesity (BMI ≥30 kg/m2), based on the observation that inequalities in obesity are greater than for overweight (Devaux & Sassi, 2013). If obese people are more likely to have a low socioeconomic position, it might be expected that those who attain extreme levels of obesity have an even greater likelihood of occupying a lower position in the socioeconomic gradient. We also hypothesised that inequalities in morbid obesity might be present in men as well as women.

2. Methods

2.1. Data source and collection

Data from the US National Health and Nutrition Examination Survey (NHANES) for 2011-12 and 2013-14 were analysed. The NHANES employs a multistage design aimed at selecting participants who are representative of the civilian United States (US) population (Centers for Disease Control, 2016). In NHANES, the response rate ranged from 45% in participants aged over 80 years to 71% in ages 30 to 39 (National Center for Health Statistics, 2015). Data from the Health Survey for England (HSE) were analysed for 2011 to 2014. The HSE also employs a multistage cluster sampling design to draw a representative sample of the non-institutional population in England (Mindell et al., 2012). Annual response rates for measurements ranged from 56% to 62% (Mindell et al., 2012). Participants who were under the age of 20 at the time of the survey were excluded from these analyses, as were those who did not have a valid BMI measurement or were pregnant during the time of the survey. Multiple years were selected to give a larger sample size. Response rates were similar for the two surveys, consistent with reducing participation rates observed in national surveys (Mindell et al., 2015).

2.2. Exposures, outcomes and co-variables

Height and weight records were obtained by the interviewer through standardised measurements (Mindell et al., 2012; Centers for Disease Control and Prevention, 2016) and used to calculate BMI. Morbid obesity was defined as a BMI of ≥40 kg/m2.

Questionnaire data for highest educational qualification and household income were used to evaluate socioeconomic position. In NHANES, participants were asked ‘what is the highest grade or level of school you have completed or the highest degree received’. Responses were grouped into the categories ‘less than 9th grade’, ‘9th to 11th grade’ at ages 14 to 17, ‘high school graduate’ typically at age 18, ‘some college or associate degree’, ‘college graduate or above’, ‘refused’ and ‘not known or missing’. In the HSE, the highest educational qualification was coded into the categories: university or college degree; higher education; A-level school examinations taken at 18 years; O-level or GCSE school examinations taken at 16 years; certificate of secondary education (CSE) taken at 14–16 years at a lower level than GCSE; no qualifications; and not disclosed. The resulting education categories were judged to be broadly comparable when mapped using the International Standard Classification of Education (ISCED) (UNESCO Institute of Statistics, 2012).

Total household income was consistently recorded between the two surveys and was used for the analysis. Household income data were collected using pre-defined categories. In NHANES, annual household income was grouped into the categories: ≥$75 000, $55 000 to $74 999, $35 000 to $54 999, $20 000 to $34 999, <$20 000 and not disclosed. In HSE, total household income was divided into the categories ≥£52 000, £33 800 to £52 000, £23 400 to £33 800, £13 000 to £23 400, <£13 000 and not disclosed.

Self-reported ethnicity from the HSE was analysed using the categories: ‘white’, ‘black’ (including black African, black Caribbean and black other), ‘Asian’, ‘mixed’, ‘other’ and ‘not disclosed’. Items concerning race and ethnicity from NHANES were mapped to the same categories with the additional categories of ‘Mexican American’ and ‘Other Hispanic’.

2.3. Analysis

Analyses were conducted separately in men and women so as to test our second hypothesis relating to gender differences. The ‘survey’ commands in the Stata and R programs were employed to account for the sampling design, based on the primary sampling units from the Health Survey for England and the sampling weights from NHANES (R Core Team, 2016; Stata Corp, 2015). Prevalence rates for morbid obesity were calculated by age group, ethnicity, income and education. Prevalence rates were age-standardised using the direct method based on 2000 US census data. The US has a younger population structure than the UK, but we chose to use one reference population to allow comparison. UK rates standardised to the European Standard Population are presented in a Supplementary file.

Logistic regression models were used to estimate the relative odds of morbid obesity by category of income or education, adjusting for age. Age-adjusted logistic regression models were also used to compare rates of morbid obesity in the highest socioeconomic category with the remaining participants. Sensitivity analyses were performed to assess the effect of adjusting for ethnicity because some ethnic groups are known to have higher obesity rates. The analyses were repeated using equivalised income, which adjusted for household size, to test the robustness of household income as a measure of socioeconomic status. These results are presented in the Supplementary file.

3. Results

From 2011 to 2014, there were 32,225 adults aged 20 years and older who participated in the HSE of whom 26,898 (83%) provided valid BMI measurements. There were 11,317 participants in NHANES 2011 from 2014, of whom 10,628 (94%) had valid BMI values.

The distribution of the sample by age and ethnicity and the prevalence of morbid obesity are presented in Table 1. The overall prevalence of morbid obesity in men was 1.7% in the UK and 4.8% in the US. For women the figures were 3.7% and 9.6% respectively. Morbid obesity was high in ‘black’ women but less so in men, with 16.0% of non-Hispanic black women in the US and 5.4% of ‘black’ women in the UK having morbid obesity.

Table 1.

Prevalence of morbid obesity in men and women from England and United States. Figures are frequencies except where indicated.

England
 United States
n/N Prevalence (%) n/N Prevalence (%)
MEN


 

 

 

 

 


Total 209/12,161 1.72 249/5219 4.84
Age group 20 to 34 32/2422 1.32 68/1421 4.79
35 to 44 44/2179 2.02 53/879 6.03
45 to 54 50/2293 2.18 40/861 4.65
55 to 64 52/2097 2.48 55/919 5.98
65 to 74 23/1943 1.18 25/646 3.87
75 and over 8/1227 0.65 8/493 1.62


 

 

 

 

 


Ethnic group White 195/10,932 1.80 101/2091 5.13
Mexican – – 31/627 5.10
Other Hispanic – – 16/460 3.58
Mixed race 4/128 4.79 – –
Asian 8/753 0.95 2/665 0.25
Black 0/257 – 86/1215 7.10
Other 2/81 2.04 13/161 7.38
Not disclosed 0/10 – – –
WOMEN


 

 

 

 

 


Total 555/14,737 3.77 514/5409 9.57
Age group 20 to 34 108/3155 3.42 124/1302 9.52
35 to 44 112/2727 4.11 95/979 9.70
45 to 54 127/2855 4.45 114/966 11.80
55 to 64 100/2412 4.15 112/940 11.91
65 to 74 75/2113 3.55 51/697 7.32
75 and over 33/1475 2.24 18/525 3.43


 

 

 

 

 


Ethnic group White 500/13,142 3.82 194/2150 9.63
Mexican – – 59/609 9.06
Other Hispanic – – 36/554 6.19
Mixed race 6/175 3.23 – –
Asian 22/899 2.59 6/691 0.90
Black 25/392 5.79 200/1258 16.0
Other 1/112 0.77 19/147 13.6
Not disclosed 1/17 11.8 – –
Open in a new tab

The age-standardised prevalence of morbid obesity according to income and education category is presented for English and US participants in Fig. 1 and Table 2, Table 3 for men and women respectively. Fig. 1 reveals consistent gradients in the distribution of morbid obesity according to income and education in both men and women in England. In English men, the prevalence of morbid obesity was 1.3% in the highest category of income and 2.3% in the lowest; in English women, the equivalent figures were 2.0% and 5.0%. English men in the highest category of educational qualification had a prevalence of morbid obesity of 0.9% compared with 2.4% in the lowest category; in women, the equivalent figures were 2.2% and 4.9%.

Fig. 1.

Fig. 1

Open in a new tab

Age-standardised prevalence of morbid obesity by household income (upper panel) and education (lower panel) in England and the USA. Black bars, men; gray bars, women.

Table 2.

Age-standardised prevalence and logistic regression model of morbid obesity in men from England and the USA by income and education category. Figures are frequencies except where indicated.

England
United States
n/N Prevalence (%) AORa(95% CI) n/N Prevalence (%) AORa(95% CI)
Household income Household income
Highest (≥£52,000) 37/2446 1.35 1.00 Highest (≥$75,000) 48/1284 3.69 1.00
£33,800 to £52,000 24/1784 1.28 0.93 (0.55 to 1.55) $55,000 to $74,999 28/504 5.18 2.65 (1.08, 6.53)
£23,400 to £33,800 29/1758 1.64 1.22 (0.75 to 1.98) $35,000 to $54,999 43/834 5.24 1.93 (0.83, 4.49)
£13,000 to £23,400 42/2167 2.11 1.62 (1.03 to 2.54) $20,000 to $34,999 53/988 5.39 2.17 (0.97, 4.88)
Lowest (<£13,000) 36/1618 2.35 1.83 (1.16 to 2.88) Lowest (<$20,000) 65/1232 6.22 1.50 (0.75, 3.01)
Not disclosed 41/2388 1.74 1.37 (0.88 to 2.13) Not disclosed 12/377 3.31 0.65 (0.22, 1.93)
Education Level Education Level
Degree 33/3342 0.94 1.00 College or above 40/1328 2.90 1.00
Higher education 41/1656 2.48 2.62 (1.65 to 4.16) Some college 91/1454 6.18 2.31 (1.13, 4.69)
A Level/NVQ3 27/1697 1.60 1.67 (1.00 to 2.79) High school 65/1199 5.63 1.60 (0.73, 3.53)
O Level/NVQ2 43/2184 1.97 2.02 (1.28 to 3.19) 9th to 11th grade 39/760 5.26 1.00 (0.43, 2.34)
CSE/NVQ1 12/645 1.85 2.16 (1.11 to 4.18) <9th grade 14/478 2.92 1.06 (0.25, 4.40)
No qualifications 52/2589 2.48 2.57 (1.64 to 4.02) – –
Not disclosed 1/48 2.26 2.89 (0.39 to 21.5) – –
Open in a new tab
a

AOR Age-adjusted Odds Ratio; CI, confidence interval; n, number with morbid obesity; N, total number in category.

Table 3.

Age-standardised prevalence and logistic regression model of morbid obesity in women from England and the USA by income and education category. Figures are frequencies except where indicated.

England
 United States
n/N Prevalence (%) AORa(95% CI) n/N Prevalence (%) AORa(95% CI)
Household income Household income
Highest (≥£52,000) 51/2538 2.01 1.00 Highest (≥$75,000) 75/1224 5.78 1.00
£33,800 to £52,000 57/1931 2.88 1.53 (1.04 to 2.24) $55,000 to $74,999 40/514 6.99 1.67 (0.85, 3.30)
£23,400 to £33,800 76/1978 3.81 2.08 (1.44 to 3.01) $35,000 to $54,999 78/881 9.47 1.85 (1.03, 3.32)
£13,000 to £23,400 140/2739 5.28 3.01 (2.14 to 4.23) $20,000 to $34,999 140/1024 13.87 2.75 (1.65, 4.56)
Lowest (<£13,000) 117/2500 5.11 2.92 (2.07 to 4.12) Lowest (<$20,000) 163/1408 12.04 1.97 (1.19, 3.25)
Not disclosed 114/3051 4.15 2.18 (1.55 to 3.07) Not disclosed 18/358 5.03 0.79 (0.30, 2.05)
Education Level Education Level
Degree 84/3723 2.23 1.00 College or above 75/1355 5.34 1.00
Higher education 46/1428 3.24 1.54 (1.06 to 2.23) Some college 206/1777 11.81 3.11 (1.83, 5.28)
A Level/NVQ3 84/2167 3.65 1.75 (1.27 to 2.41) High school 113/1111 10.67 2.68 (1.50, 4.79)
O Level/NVQ2 140/3163 4.60 2.07 (1.55 to 2.76) 9th to 11th grade 77/714 10.92 2.04 (1.09, 3.80)
CSE/NVQ1 35/474 7.38 3.47 (2.30 to 5.24) <9th grade 43/452 9.24 1.87 (0.93, 3.74)
No qualifications 152/3395 4.90 2.61 (1.95 to 3.48) – –
Not disclosed 14/387 2.76 2.26 (1.25 to 4.09) – –
Open in a new tab
a

AOR Age-adjusted Odds Ratio; CI, confidence interval; n, number with morbid obesity; N, total number in category.

In the US, there was a gradient in the distribution of morbid obesity by income in women: 5.8% of the highest income category had morbid obesity compared with 12.0% in the lowest income category. In US men, there was no consistent gradient. US men in the highest income category had a prevalence of morbid obesity of 3.7%. The remaining income categories all showed values between 5% and 6%, except the lowest category at 6.2%. In the US, the highest category of education showed the lowest prevalence of morbid obesity (2.9% in men and 5.3% in women), but the second highest education category showed the highest prevalence of morbid obesity (6.2% in men and 11.8% in women).

Age-adjusted odds ratios (AOR) of morbid obesity by income and by education are presented in Table 2, Table 3. These estimates confirm a graded association of morbid obesity with income and education category in both English men and women. In the lowest category of income, compared with the highest, the relative odds of morbid obesity were 1.83 (95% confidence interval 1.16 to 2.88) in men and 2.92 (2.07 to 4.12) in women. In the lowest category of education, compared with the highest, the relative odds of morbid obesity were 2.57 (1.64 to 4.02) in men and 2.61 (1.95 to 3.48) in women. In US women, there was evidence of a gradient in morbid obesity related to income, with relative odds for the lowest income category of 1.97 (1.19 to 3.25). In US men, the greatest odds of morbid obesity were for the second highest category of income (AOR 2.65, 1.08 to 6.53). In both US men and women, the greatest odds of morbid obesity were for the second highest category of education (some college education or associate degree, men 2.31, 1.13 to 4.69; women, 3.11, 1.83 to 5.28).

Inspection of estimates in Fig. 1 suggested that, in the US, people with highest level of education or income might have some protection against morbid obesity, when compared with all other groups. Table 4 presents a comparison of the prevalence of morbid obesity in those from the highest income (greater than £52 000 or $75 000) or education (degree or college) categories in both settings, compared with all others. The likelihood of morbid obesity for US men in the highest category of income was approximately half that of the remainder of the population (AOR 0.53, 0.27 to 0.98). A similar pattern was observed for US women, for both the highest category of income (AOR 0.51, 0.33 to .80 and the highest category of education (AOR 0.36, 0.22 to 0.60). This finding was not statistically significant for education as a predictor in US men (AOR 0.56, 0.29 to 1.08). UK men were less likely to be morbidly obese if they were in the highest education category (AOR 0.46, 0.31 to 0.66), but not if they were in the highest income category (AOR 0.73, 0.50 to 1.06). In the UK data, the association was stronger in women than men (AOR for income 0.42, 0.31 to 0.57; AOR for education 0.48, 0.37 to 0.61). Adjusting for ethnicity did not alter the results.

Table 4.

Logistic regression model comparing morbid obesity in highest income or education categories with all others in US and UK men and women.

Men
 Women
Freq (%) AOR P value Freq (%) AOR (95% CI) P value
(95% CI)
England


 

 

 

 

 

 

 


Household ≥£52,000 37/2,446 (1.5) 0.73 (0.50 to 1.06) 0.097 51/2538 (2.0) 0.42 (0.31 to 0.57) <0.001
income
All othersa 131/7,327 (1.8) Ref. 390/9148 (4.3) Ref.
Education Degree 33/3,342 (1.0) 0.46 (0.31 to 0.66) <0.001 84/3723 (2.3) 0.48 (0.37 to 0.61) <0.001
qualifications
All othersa 175/8,771 (2.0) Ref. 457/10,627 (4.3) Ref.
United States


 

 

 

 

 

 

 


Household ≥$75,000 48 / 1284 (3.7) 0.52 (0.27 to 0.98) 0.045 75/1224 (6.1) 0.51 (0.33 to 0.80) 0.004
income
All others 201 / 3,935 (5.1) Ref. 439/4185 (10.5) Ref.
Education College education 40 / 1,328 (3.0) 0.56 (0.29 to 1.08) 0.084 75/1355 (5.5) 0.36 (0.22 to 0.60) <0.001
level
All others 209 / 3,891 (5.4) Ref. 439/4054 (10.8) Ref.
Open in a new tab
a

Not disclosed category removed; AOR Age-adjusted Odds Ratio; CI, confidence interval.

4. Discussion

4.1. Summary of findings

The present results provide new evidence of socioeconomic inequality in morbid obesity in two high-income countries with differing obesity profiles. While the results affirm that socioeconomic disparities are generally greater among women, the findings support the hypothesis that inequalities in morbid obesity are evident in men as well as women.

The study provided evidence of consistent socioeconomic gradients in morbid obesity according to income and education in England, where morbid obesity is less frequent overall. The lowest rates of morbid obesity in any US socioeconomic group were greater than the highest rates in England, suggesting that social environmental exposures, characterised as the ‘obesogenic environment’, may be more pervasive across social strata in the U.S. (Banks et al., 2006, Siddiqi et al., 2015). In the US, socioeconomic gradients were less consistent, with the highest rates found at intermediate socioeconomic positions. The high prevalence of morbid obesity in the US may account for a more widespread distribution throughout the socioeconomic scale. The distribution of morbid obesity in the US is not consistent with the hypothesis that inequalities will be more apparent with the prevalence of morbid obesity is higher. However, in both the US and England, participants with the highest levels of education or income had substantially lower odds of morbid obesity compared to the rest of the population.

4.2. Comparison with previous studies

Most previous studies have evaluated socioeconomic gradients in the distribution of all obesity considered as a single condition (Devaux & Sassi, 2013; McLaren, 2007). Greater income inequality has been associated with higher obesity prevalence, with the US experiencing both high income inequality and high obesity rates (Pickett, Kelly, Brunner, Lobstein & Wilkinson, 2005). The pattern of inequality in obesity may be changing over time. A recent NHANES study showed that the prevalence of obesity, and inequalities in obesity, increased up to the year 2000, followed by a more gradual increase in obesity prevalence that was more evenly distributed among socioeconomic groups (Pak et al., 2016). In an analysis of Scottish data, Zhu et al. (2015), found that education and income inequality in obesity reduced as obesity prevalence increased over time.

In the present study, the highest levels of household income and educational attainment were consistently associated with lower morbid obesity. This is consistent with the analysis of Siddiqi et al., which found that in the U.S., college-educated individuals showed a lower prevalence of obesity than all other groups. Consistent with Siddiqi et al. we find that there are cross-national differences in the prevalence of morbid obesity, in the shape of the socioeconomic distribution and the absolute and relative magnitude of inequalities (Siddiqi et al., 2015). Inequalities in income and education may influence health via multiple downstream mediators including, but not restricted to, lifestyle choices, diet quality and access to resources for physical activity (Benach & Muntaner, 2007; Devaux, 2013; Devaux & Sassi, 2013; Gaglioti, Petterson, Bazemore & Phillips, 2016). Educational attainment is generally expected to be associated with higher income but this effect may be modified by other characteristics including age-group, ethnicity and gender (Braveman, Cubbin & Egerter, 2005). The health effects of socioeconomic variables may also depend on context. For example, education may sometimes play a greater role in influencing lifestyle than income, with the opposite true in other settings (El-Sayed, Scarborough & Galea, 2012). Consequently, associations may differ between countries, or between socio-demographic groups within countries.

At the individual level, the association between education and morbid obesity may be exercised through access to health-related information, the ability to interpret this in the context of awareness of risks, and capacity to regulate choices (Devaux, Sassi, Church, Cecchini & Borgonovi, 2011). Low income may be associated with reduced access to a broad range of health resources contributing to consistent associations of income inequality with a range of health outcomes in England and the US (Martinson, 2012). Morbid obesity rates were sometimes lower at the lowest income levels which might be associated with food insecurity (Franklin, Jones, Love, Puckett, Macklin & White-Means, 2012) or high rates of occupational physical activity in this group (Bonauto, Lu & Fan, 2014).

4.3. Limitations

This study drew on national survey data from England and the United States, employing carefully standardised measurement techniques. However, cross-national comparisons may encounter differences in approach and data definitions. In this study, we compared available measures of educational level and household income, but definitions were not standardised across countries. Mapping education categories to international standards suggested that measures from the two surveys were broadly comparable but there was limited differentiation among the more central categories by international standards.

We evaluated total household income without consideration of household size. This may have biased estimates for larger households or if the respondent was not the main earner. However, our approach is generally consistent with the one used in other studies (Kakinami, Gauvin, Barnett & Paradis; Martinson, 2012) Alternatives to household income may have included the ratio of family income to poverty in NHANES and equivalised income in HSE, but consistent definitions were not available in either survey. Sensitivity analyses demonstrated that the conclusions were not altered by varying the definition of household income.

We used income data from different survey years without adjusting for purchasing power. Neither survey incorporated a measure of ‘wealth’, a potentially more revealing measure that may be difficult to obtain (Galobardes, Lynch & Smith, 2007). The data were cross-sectional and we are not able to evaluate causal pathways; we have focused on the effects of socioeconomic status on obesity rather than the effect of obesity on social mobility. In cross-sectional analyses, it may be difficult to distinguish between covariates that contribute confounding and those that contribute to causal relationships. Longitudinal analyses are required to increase understanding of the determinants of more extreme forms of obesity.

5. Conclusions

In the UK, both men and women with lower income or education, are more likely to have morbid obesity than those in higher socioeconomic groups. Consistent socioeconomic gradients in morbid obesity are less apparent in the US, where only those with the highest levels of income and education consistently demonstrated lower morbid obesity. The higher overall prevalence of morbid obesity in the US suggests that social environmental influences impact on obesity across social strata. In the context of high overall prevalence, socioeconomic position may now have a more limited impact on the distribution of morbid obesity. This suggests that any continuing increase in morbid obesity in the UK might result in morbid obesity spreading into higher socioeconomic categories.

Occupying the highest socioeconomic positions appeared to offer protection against the development of morbid obesity in both England and the US. This is consistent with known graded association between socioeconomic status and health, and reinforces the importance of social factors in determining health (Commission on Social Determinants of Health, 2008). A more explicit understanding of how high socioeconomic position confers protection against morbid obesity may offer insights that might inform policies and interventions for prevention and treatment. Further work should focus on ensuring obesity interventions are accessible and effective across all social strata, and investigating whether the health consequences and costs in people with morbid obesity are socially patterned.

Competing interests

The authors declare no conflict of interest.

Funding

Professor Gulliford is funded by the BIHR Biomedical Research Centre at Guy's and St Thomas’ NHS Foundation Trust. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Contributorship

HB conducted the analyses, contributed to interpretation of the results and wrote the manuscript. JC conducted the analyses, contributed to interpretation of the results and writing of the manuscript. MG devised the study and contributed to the analyses, interpretation of the results and writing of the manuscript.

Footnotes

Appendix A

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.ssmph.2016.12.012.

Appendix A. Supplementary material

Supplementary material

mmc1.docx (24KB, docx)

.

References

  1. Arterburn D., Maciejewski M., Tsevat J. Impact of morbid obesity on medical expenditures in adults. International Journal of Obesity and Related Metabolic Disorders. 2005;29(3):334–339. doi: 10.1038/sj.ijo.0802896. [DOI] [PubMed] [Google Scholar]
  2. Bambra C. Going beyond the three worlds of welfare capitalism: regime theory and public health research. Journal of Epidemiology and Community Health. 2007;61:1098–1102. doi: 10.1136/jech.2007.064295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Banks J., Marmot M., Oldfield Z., Smith J.P. Disease and disadvantage in the United States and in England. JAMA. 2006;295(17):2037–2045. doi: 10.1001/jama.295.17.2037. [DOI] [PubMed] [Google Scholar]
  4. Beenackers M.A., Kamphuis C.B., Giskes K., Brug J., Kunst A.E., Burdorf A., van Lenthe F.J. Socioeconomic inequalities in occupational, leisure-time, and transport related physical activity among European adults: a systematic review. The International Journal of Behavioral Nutrition and Physical Activity. 2012;9(1):1–23. doi: 10.1186/1479-5868-9-116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Benach J., Muntaner C. Precarious employment and health: developing a research agenda. J Epidemiol Community Health. 2007;61(4):276–277. doi: 10.1136/jech.2005.045237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bonauto D.K., Lu D., Fan Z.J. Obesity prevalence by occupation in Washington state, behavioral risk factor surveillance system. Preventing Chronic Disease. 2014;11:E04. doi: 10.5888/pcd11.130219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Booth H.P., Prevost A.T., Gulliford M.C. Impact of body mass index on prevalence of multimorbidity in primary care: cohort study. Family Practice. 2014;31(1):38–43. doi: 10.1093/fampra/cmt061. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Booth H.P., Khan O., Prevost T., Reddy M., Dregan A., Charlton J., Gulliford M.C. Incidence of type 2 diabetes after bariatric surgery: population-based matched cohort study. Lancet Diabetes & Endocrinology. 2014 doi: 10.1016/S2213-8587(14)70214-1. [DOI] [PubMed] [Google Scholar]
  9. Braveman P.A., Cubbin C., Egerter S. Socioeconomic status in health research: one size does not fit all. JAMA. 2005;294(22):2879–2888. doi: 10.1001/jama.294.22.2879. [DOI] [PubMed] [Google Scholar]
  10. Centers for Disease Control and Prevention (2016). National Health and Nutrition Examination Survey. Retrieved from: 〈http://www.cdc.gov/nchs/nhanes/index.htm〉
  11. Commission on Social Determinants of Health (2008). Closing the gap in a generation: health equity through action on the social determinants of health. Final report of the commission on social determinants of health. Retrieved from: 〈http://www.who.int/social_determinants/thecommission/finalreport/en/〉
  12. Department of Health Public Health Research Consortium, Law C., Power C., Graham H., Merrick D. Obesity and health inequalities. Obesity Reviews. 2007;8:19–22. doi: 10.1111/j.1467-789X.2007.00312.x. [DOI] [PubMed] [Google Scholar]
  13. Devaux M. Income-related inequalities and inequities in health care services utilisation in 18 selected OECD countries. The European Journal of Health Economics. 2013;16(1):21–33. doi: 10.1007/s10198-013-0546-4. [DOI] [PubMed] [Google Scholar]
  14. Devaux M., Sassi F. Social inequalities in obesity and overweight in 11 OECD countries. European Journal of Public Health. 2013;23(3):464–469. doi: 10.1093/eurpub/ckr058. [DOI] [PubMed] [Google Scholar]
  15. Devaux M., Sassi F., Church J., Cecchini M., Borgonovi F. Exploring the relationship between education and obesity. OECD Journal: Economic Studies. 2011;1:121–159. [Google Scholar]
  16. Drewnowski A. Obesity, diets, and social inequalities. Nutrition Reviews. 2009;67(S1) doi: 10.1111/j.1753-4887.2009.00157.x. [DOI] [PubMed] [Google Scholar]
  17. El-Sayed A.M., Scarborough P., Galea S. Unevenly distributed: a systematic review of the health literature about socioeconomic inequalities in adult obesity in the United Kingdom. BMC Public Health. 2012;12 doi: 10.1186/1471-2458-12-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Finkelstein E.A., Ruhm C.J., Kosa K.M. Economic causes and consequences of obesity. Annual Review of Public Health. 2005;26(1):239–257. doi: 10.1146/annurev.publhealth.26.021304.144628. [DOI] [PubMed] [Google Scholar]
  19. Franklin B., Jones A., Love D., Puckett S., Macklin J., White-Means S. Exploring mediators of food insecurity and obesity: a review of recent literature. Journal of Community Health. 2012;37(1):253–264. doi: 10.1007/s10900-011-9420-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gaglioti A.H., Petterson S., Bazemore A., Phillips R. Access to primary care in US counties is associated with lower obesity rates. The Journal of the American Board of Family Medicine. 2016;29(2):182–190. doi: 10.3122/jabfm.2016.02.150356. [DOI] [PubMed] [Google Scholar]
  21. Galobardes B., Lynch J., Smith G.D. Measuring socioeconomic position in health research. British Medical Bulletin. 2007;81–82(1):21–37. doi: 10.1093/bmb/ldm001. [DOI] [PubMed] [Google Scholar]
  22. García Villar J., Quintana-Domeque C. Income and body mass index in Europe. Economics and Human Biology. 2009;7(1):73–83. doi: 10.1016/j.ehb.2009.01.006. 〈http://dx.doi.org/10.1016/j.ehb.2009.01.006〉 [DOI] [PubMed] [Google Scholar]
  23. Hall P., Soskice D. Oxford University Press; Oxford, UK: 2001. Varieties of capitalism: the institutional foundations of comparative advantage. [Google Scholar]
  24. Joint Health Surveys Unit (2014). Health Survey for England: Trend Tables 2013. Retrieved from London, UK.
  25. Kakinami, L., Gauvin, L., Barnett, T.A., & Paradis, G. Trying to Lose Weight. Am J Prev Med, 46 (6), pp. 585–592. doi:10.1016/j.amepre.2014.01.022 [DOI] [PubMed]
  26. Mackenbach J.P., Stirbu I., Roskam A.-J.R., Schaap M.M., Menvielle G., Leinsalu M., Kunst A.E. Socioeconomic inequalities in health in 22 European countries. NEJM. 2008;358(23):2468–2481. doi: 10.1056/NEJMsa0707519. [DOI] [PubMed] [Google Scholar]
  27. Martinson M.L. Income inequality in health at all ages: a comparison of the United States and England. American Journal of Public Health. 2012;102(11):2049–2056. doi: 10.2105/AJPH.2012.300929. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. McLaren L. Socioeconomic status and obesity. Epidemiologic Reviews. 2007;29(1):29–48. doi: 10.1093/epirev/mxm001. [DOI] [PubMed] [Google Scholar]
  29. Mindell J., Biddulph J.P., Hirani V., Stamatakis E., Craig R., Nunn S., Shelton N. Cohort profile: the health survey for England. The International Journal of Epidemiology. 2012 doi: 10.1093/ije/dyr199. [DOI] [PubMed] [Google Scholar]
  30. Mindell J.S., Giampaoli S., Goesswald A., Kamtsiuris P., Mann C., Männistö S., on behalf of the HES Response Rate Group Sample selection, recruitment and participation rates in health examination surveys in Europe – experience from seven national surveys. BMC Medical Research Methodology. 2015;15:78. doi: 10.1186/s12874-015-0072-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. National Center for Health Statistics (2015). NHANES Response Rates and Population Totals. Retrieved from: 〈http://www.cdc.gov/nchs/nhanes/response_rates_cps.htm〉
  32. NCD Risk Factor Collaboration Trends in adult body-mass index in 200 countries from 1975 to 2014: a pooled analysis of 1698 population-based measurement studies with 19·2 million participants. The Lancet. 2016;387(10026):1377–1396. doi: 10.1016/S0140-6736(16)30054-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. NCD Risk Factor Collaboration Worldwide trends in diabetes since 1980: a pooled analysis of 751 population-based studies with 4.4 million participants. The Lancet. 2016;387(10027):1513–1530. doi: 10.1016/S0140-6736(16)00618-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Pak T.-Y., Ferreira S., Colson G. Measuring and tracking obesity inequality in the United States: evidence from NHANES, 1971–2014. Population Health Metrics. 2016;14(1):1–13. doi: 10.1186/s12963-016-0081-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Pickett K.E., Kelly S., Brunner E., Lobstein T., Wilkinson R.G. Wider income gaps, wider waistbands? An ecological study of obesity and income inequality. The Journal of Epidemiology and Community Health. 2005;59(8):670–674. doi: 10.1136/jech.2004.028795. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. R Core Team . R Foundation for Statistical Computing; Vienna, Austria: 2016. R: A language and environment for statistical computing.〈https://www.R-project.org/〉 (Retrieved from) [Google Scholar]
  37. Rudisill C., Charlton J., Booth H.P., Gulliford M.C. Are healthcare costs from obesity associated with body mass index, comorbidity or depression? Cohort study using electronic health records. Clinical Obesity. 2016;6(3):225–231. doi: 10.1111/cob.12144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Sassi F., Devaux M., Cecchini M., Rusticella E. OECD Publishing; 2009. The obesity epidemic: analysis of past and projected future trends in selected OECD countries.〈http://www.oecd.org/officialdocuments/publicdisplaydocumentpdf/?doclanguage=en&cote=delsa/hea/wd/hwp(2009)3〉 Retrieved from: [Google Scholar]
  39. Siddiqi A., Brown R., Nguyen Q.C., Loopstra R., Kawachi I. Cross-national comparison of socioeconomic inequalities in obesity in the United States and Canada. International Journal for Equity in Health. 2015;14:116. doi: 10.1186/s12939-015-0251-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. StataCorp . StataCorp LP; College Station, TX: 2015. Stata statistical software: release 14. [Google Scholar]
  41. Sturm R., Hattori A. Morbid obesity rates continue to rise rapidly in the United States. International Journal of Obesity. 2013;37(6):889–891. doi: 10.1038/ijo.2012.159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. UNESCO Institute of Statistics (2012). International Standard Classification of Education (ISCED 2011). Retrieved from: Montreal, Quebec: 〈http://www.uis.unesco.org/Education/Documents/isced-2011-en.pdf〉
  43. Wilkinson R.G. Socioeconomic determinants of health: health inequalities: relative or absolute material standards? BMJ. 1997;314(7080):591. doi: 10.1136/bmj.314.7080.591. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Zhu J., Coombs N., Stamatakis E. Temporal trends in socioeconomic inequalities in obesity prevalence among economically-active working-age adults in Scotland between 1995 and 2011: a population-based repeated cross-sectional study. BMJ Open. 2015;5(6) doi: 10.1136/bmjopen-2014-006739. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary material

mmc1.docx (24KB, docx)

Articles from SSM - Population Health are provided here courtesy of Elsevier

RESOURCES