Obesity in England 'rising at a slower rate'

“Britons have stopped getting fatter,” The Daily Telegraph reports, while the Mail Online headlines that it’s “the fat getting fatter”. Despite a drop in the nation’s overall obesity rate, they say that the heaviest people continue to put on weight.

In fact, it appears the English are (on the whole) still “getting fatter”, the rate of increase has slowed, although those classed as overweight or obese are increasing at a greater rate than the rest.

These reports are based on a study that used data obtained from the annual Health Survey for England to explore body mass index (BMI) trends among adults between 1992 and 2010.

Overall, the research shows gradual increases in the average BMI over time, from 25.6kg/m2 to 27.5kg/m2 in men; and from 24.5kg/m2 to 26.5kg/m2 for women. Most of this increase occurred before 2001, after this there has been a much slower rate of increase.

The average BMI in the overweight or obese category tended to increase over the 19 years, while the average BMI in the normal weight category showed little change over time.

It is important to note that the surveys did not necessarily include the same people each year, so it can’t tell us what was happening to individuals. Rather, the study gives an overview of how BMI has been changing in England as a whole.

Along with other research, it may help public health officials to plan how to target interventions to prevent and reduce overweight and obesity going forward.

Where did the story come from?

The study was carried out by researchers from Lancaster University and University of Manchester and was funded by a grant from ESRC Obesity. The study was published in the peer-reviewed International Journal of Obesity.

The headlines are generally representative of the mixed findings of this modelling study. The Mail is correct (and rather blunt) in saying that “the fat are getting fatter” but the seemingly conflicting Telegraph headline is slightly misleading. It says that “Britons have stopped getting fatter”, which is not supported by the evidence showing that the population is still increasing in BMI but at a slower rate than in previous years.

What kind of research was this?

This was a time-trend study using cross-sectional data from the Health Survey for England, conducted annually between 1992 and 2010, to investigate trends in reported BMI.

The researchers say that the proportion of people who are overweight or obese (BMI 25kg/m2 or greater) has been increasing over the past 30 years, but their theory is that the rate of increase has been slowing in recent times.

The proportion of a population who are overweight or obese in any given year is affected by the number of new cases becoming overweight and obese each year, and the duration that individuals stay in this category or leave the category due to weight loss or death.

The authors suggested that if there is no population-wide intervention that reduces the duration of people’s obesity or being overweight significantly, then the proportion may reach a position of ‘saturation’. This is where the rate of people becoming obese or overweight roughly equals the rate of obese or overweight people dying or losing weight.

Such cross-sectional data from a large, representative population sample of England can provide us with information on trends in BMI across England over the past roughly 20 years. However, it can’t tell us about what happened to individuals, or provide any explanations for the trend.

What did the research involve?

The annual Health Survey for England (HSE) takes a representative sample of households living in England and invites householders to participate in interviews on various aspects of health.

Assessments were conducted by trained interviewers at the homes of participants, and included measurement of height and weight using standard procedures. HSE response rates are reported to be around 70% agreeing to interview, around 90% of whom have BMI measurements taken.

The UK Data Archive was used to download key data from the HSEs on age, BMI, and other sociodemographic factors, including smoking status, educational level, social class, and household income.

The researchers’ analyses included data on 164,166 adults (aged 20 to 74 years) with valid BMI data available. Changes in the median (average) BMI over the years were explored using computer models to look at trends in BMI. The models were specific to gender and used age brackets of 20 to 34 years, 35 to 49 years, and 50 to 74 years to reflect, respectively, early, middle and late adulthood. The association between other sociodemographic variables and BMI trends was explored.

What were the basic results?

The dataset included 76,382 men and 87,773 women. Across the 19-year study period 1992 to 2010 there were increases for both men and women in mean age, mean height, proportion of never smokers and increases in proportion with higher education.

Across the 19-year period median BMI increased for men from 25.6kg/m2 in 1992 to 27.5kg/m2 in 2010. For women median BMI increased from 24.5 kg/m2 in 1992 to 26.5kg/m2 in 2010. However, the increase was not evenly distributed over time, and there was a slower increase of median BMI after 2001. The increase was 0.14kg/m2 per year in both sexes prior to 2001; after 2001, in men the increase was 0.038kg/m2 per year, and in women 0.055kg/m2 per year.

When looking separately at those men and women who were overweight or obese (roughly a quarter of men and a third of women), the median age-adjusted BMIs increased significantly from 1992 to 2010.

For men, the median age-adjusted BMIs increased from 26.9kg/m2 in 1992 to 31.2kg/m2 in 2010 (a change of 0.304kg/m2 per year before 2001 versus 0.173 kg/m2 after). For women the median age-adjusted BMIs increased from 27.4kg/m2 in 1992 to 30.8kg/m2 in 2010 (a change of 0.234kg/m2 per year before 2001 versus 0.103 kg/m2 after)

By contrast, those in the ‘normal’ BMI group showed smaller increases in BMI over this period, with consistent 0.049kg/m2 yearly increases in men and 0.031kg/m2 in women.

How did the researchers interpret the results?

The researchers conclude that gender-specific median BMI values increased steadily from 1992 to 2001, with the rate of increase slowing down between 2001 and 2010, but not levelling off completely.

They say that the trends were “consistent with a hypothesis of a high BMI sub-population getting ‘fatter’; the slowing down of trend increases being explained by a majority proportion ‘resistant’ normal BMI population”.

Conclusion

This study using cross-sectional survey data explores the change in trend in BMI among adults in England over the 19-year period 1992 to 2010. It suggests a possible tailing off in the increase in BMI at a population level.

The research benefits from using data collected from the Health Survey of England, which is described as a “high-quality data source”. The survey benefits from using annually collected height and weight measurements taken by trained interviewers to measure BMI, rather than by self-reporting, which can be inaccurate.

There are some limitations, including the potential for response bias. As the researchers say, BMI was available from only 63% of the people approached by the survey and there could be some differences in BMI trends among those who consented from those who declined. Also the researchers note that the survey data does not include a good representation of ethnic minority groups.

Overall, the research shows gradual increases in BMI over time, but that there are larger increases in average BMI in the overweight and obese category than in the normal weight category (which shows minimal variation over time). What the study can’t tell us is what happened to individuals, or the specific reasons for the trends seen (such as the influence of diet and activity).

Along with other research, this type of study may help public health officials to plan how to target interventions to prevent and reduce overweight and obesity going forward.