Cover Page
Article Title: Use of data from the Health Survey for England in obesity policy making and monitoring
Author Names: Dr Oyinlola Oyebode, Dr Jennifer Mindell
Department and Institution to which authors belong: Research Department of Epidemiology and Public Health, UCL (University College London)
Three or four key words: “Health Examination Survey” “Evidence-based Policy” “Obesity Strategy”
A Running Title: Use of Health Survey for England in policy making and monitoring
Acknowledgements: We thank all those who were interviewed and/or provided information or suggestions of documents to review. Additional thanks go to Elizabeth Breeze for helpful comments and suggestions on the manuscript.
Corresponding Author: Dr Oyinlola Oyebode, Research Department of Epidemiology and Public Health, UCL, 1-19 Torrington Place, London WC1E 6BT
Funding: JM is funded by the Health and Social Care Information Centre (HSCIC) to work on the Health Survey for England (HSE) series; OO is funded by the London Deanery Public Health Training Programme. Both the HSCIC and the Department of Health (DH) fund the HSE. This study was unfunded; the HSE funders were not involved in the decision to undertake this work nor to publish it.
Summary
Health data and statistics are the foundation of health policy. Over the last 20 years, numerous government documents have been commissioned and published to inform obesity strategies in the UK. The Health Survey for England, an annual cross-sectional survey of a nationally representative, random general population sample in England. It collects information on health, lifestyle and socio-economic factors, physical measurements and biological samples. Heights and weights measured by the Health Survey for England are believed to have played a major part in promoting, shaping and evaluating obesity strategies. A formal review of how these data have been used has not been conducted previously.
This paper reviews government documents, demonstrating the contribution of Health Survey for England examination data to every stage of the policy-making process:
· quantifying the obesity problem in England (e.g. Chief Medical Officer’s reports);
· identifying inequalities in the burden of obesity (Acheson report);
· modelling potential future scenarios (Foresight);
· setting and monitoring specific, measurable, attainable targets (calorie reduction challenge in manufacturers’ Responsibility Deal);
· developing and informing strategies and clinical guidance; and
· evaluating the success of obesity strategies (Healthy Weights, Healthy Lives progress report).
·
Measurement data are needed and used by governments to produce evidence-based strategies to combat obesity.
1. Introduction
Obesity has become an important problem in populations worldwide. Many governments have produced strategies to tackle this problem. In the 20 years since the Health of the Nation Strategy was published in 1992, successive UK governments have commissioned and published numerous documents, targets and strategies to deal with growing levels of obesity.
Reliable health data and statistics are the foundation of health policies, strategies, evaluation and monitoring and evidence is also the foundation for sound health information for the general public ([1]). The main source of health data and statistics to inform obesity policies in the UK is the Health Survey for England ([2]).
The Health Survey for England (HSE) is an annual survey which uses a multistage stratified design to sample a nationally representative random cross-section of the free-living general population of England. Following an advance letter to the selected households, an interviewer visits to recruit up to ten adults and up to two children per household. The interviewer collects socio-economic data and information on health and health-related behaviours, and measures height and weight. Participants who agree are then visited by a nurse who measures waist and hip circumferences and blood pressure, and collects biological samples and information on medication use. The nurses have also collected infant length, demi span and mid-upper arm circumference in some years. Field-staff undergo training, with refresher training annually. Data on adults has been collected yearly since 1991 and on children from 1995 onwards. In addition to an annual published report, data are freely available through the UK Data Service to those in the public or academic sectors, and are provided to the Department of Health directly.
Other countries conduct similar health examination surveys on nationally representative samples of their population, a number are planning their first one ([3]); others are seeking evidence of cost-effectiveness to justify funding ([4]). Most health examination surveys have been conducted for a single time point, however others are long running, such as the National Health And Nutrition Examination Survey (NHANES) in the USA, which began in the 1960s and became continuous in 1999; and FINRISK in Finland, updated every 5 years, since 1972. The Health Survey for England has been annual since it began, with the logistics of a smaller annual survey more efficient in terms of staff recruitment, retention and training and survey output more responsive to changes in national priorities. The only other country to have annual measured height and weight for a general population sample is Japan ([5]).
The Health Survey for England results from a 1988 ‘command’ paper Public Health in England ([6]). This paper also led to the publication of the Health of the Nation strategy in 1992 ([7]). The Health of the Nation provided a strategic approach to improving the health of the population. It was the first time a UK government had set out to do this. The strategy included a series of targets; the Health Survey for England was to be used to monitor targets where a data deficit had been identified.
The Health of the Nation included an obesity target. This target was set after examination of the Office for National Statistics survey data The Heights and Weights of Adults in Great Britain, 1980 ([8]) where obesity prevalence was 6% of men and 8% of women and data from The Dietary and Nutritional Survey of British Adults, 1986/7 which showed that obesity prevalence was 8% of men and 12% of women ([9]). In both cases, these data described the population aged 16-64. The Health of the Nation target was set as follows: “To reduce the proportion of men and women aged 16-64 who are obese by at least 25% and 33% respectively by 2005 to no more than 6% of men and 8% of women” (7). In 2005, HSE data showed that 22% of men and 23% of women aged 16-64 were obese ([10]). Rather than demonstrating a government failure, the failure to reach the target illustrates the lack of understanding of the scale of the problem or the rate at which it was growing before Health Survey for England data were available. In retrospect, the Health of the Nation obesity target may be perceived as awry, but it was set by intelligent civil servants working with the best data available at the time. Had year-on-year data been available, a different target might have been set.
Policy-making includes several stages: from identifying, quantifying and promoting recognition of a health issue to seek or to justify policy intervention; through strategy development, including impact assessment and selection of targets; to target monitoring and policy evaluation. This paper examines where and how the measured height and weight data from the Health Survey for England have been used by obesity policy makers in the UK. Specific examples are presented. These are internationally relevant, where the potential of annual measured heights and weights to affect obesity policy is being considered.
2. Uses of the Health Survey for England
2.1 Quantifying the Obesity Problem in England
The Health Survey for England began in 1991, when 53% of men and 44% of women were found to be overweight (body mass index, BMI 25-29.9kg/m2) or obese (BMI ≥30kg/m2) ([11]). Children were first surveyed in 1995; 24% of boys and 26% of girls aged 2-15 years were overweight or obese (defined as on or above the 85th or 95th centile respectively for their age and sex, using the UK1990 centiles ([12]))(13). The latest figures show that obesity prevalence has risen to 65% in men and 58% in women [Figure 1], and 31% in boys and 28% in girls [Figure 2] and for men, boys and girls there have been higher peak prevalences ([13]). During the lifetime of the Health Survey for England, simply presenting the obesity prevalence figures, i.e. quantifying the problem, has been enough to shock, shame and urge action. Many of the papers that have influenced obesity policy in England have used Health Survey for England data in this way [Table 1].
A specific example of the use of health examination data to quantify the obesity problem in England, which highlighted this issue to government and was key to developing obesity policy, was the Chief Medical Officer’s Report of 2002 ([14]).
The Chief Medical Officer (CMO) is the government’s most senior medical advisor, responsible for providing expert advice on any issue relating to the population’s health, and independently representing the nation’s health in government. Through his or her annual report, the CMO identifies priorities in health and recommends action to improve public health.
Obesity was a main feature in the Chief Medical Officer’s 2002 annual report by Sir Liam Donaldson (Chief Medical Officer 1998-2010). His report warned of an obesity ‘time bomb’ that would have dire consequences for the future health of the population. Health Survey for England data were used to quantify both the scale of the problem, with two-thirds of all men and half of all women overweight or obese (24 million adults) at the time, and the trend was used to show that the issue was of increasing concern. Rates of obesity were demonstrated to have trebled in the previous 20 years (14) [Figure 1].
Childhood obesity data from the Health Survey for England were also presented. In 2001, 9% of six-year-olds and 15% of 15-year-olds were obese. The trend showed that between 1996 and 2001, the proportion of overweight children aged 6-15y increased by 7% and obesity by 4%. Health inequalities in obesity and ethnic differences were also reported, with higher rates in more disadvantaged groups (14)
The Chief Medical Officer returned to this issue, using Health Survey for England data to highlight obesity as a particular problem for the West Midlands (in 2003 ([15])) and the East Midlands (in 2004 ([16])). The 2005 Chief Medical Officer’s report used the Health Survey for England’s prevalence of obesity in children under 11 to highlight London’s poor performance in this area ([17]).
The 2002 Chief Medical Officer’s report was widely quoted in the media, and obesity rapidly became a matter of public concern. Sir Liam Donaldson said in 2008: “The publicity for my "timebomb" scenario was huge, and a 2004 evaluation of media coverage of my reports showed obesity became a bigger issue than either smoking or the MMR vaccine - both "hot topics".” ([18]). Giving a public health problem, like obesity, such a high profile, in the context of reputable and irrefutable data, ensured that there would be sustained government commitment to tackle the problem.
2.2 Identifying inequalities in the burden of obesity
Obesity both contributes to and results from health and social inequalities ([19]). As the Health Survey for England collects information on socio-economic circumstances of participants, analysing prevalence of obesity by a range of individual or household measures of socio-economic position, such as occupational social class, education level, equivalised household income (adjusted for the number of adults and children in the household), area deprivation (Index of Multiple Deprivation (IMD) quintile), overcrowding, or car ownership, is possible. Such analyses have been presented in the inequalities literature produced by and for government over the last 20 years [Table 1].
The Health Survey for England was used in the Independent Inquiry into Inequalities in Health Report, led by Sir Donald Acheson in 1998 (the Acheson Report ([20])), one of the key documents in the debate on health inequalities in the UK. The data were used to illustrate the marked social class gradient in obesity which was greater among women than among men. The prevalence of obesity among women in disadvantaged groups was almost twice that in the most affluent groups (25% in class V compared with 14% in class I).
Obesity in women is particularly important, as maternal obesity is a risk factor for adverse perinatal outcomes and infant mortality ([21],[22]). The specific target set in the wake of the Acheson Report was to reduce the inequality in health outcomes between different socio-economic groups by 10%, as measured by infant mortality and life expectancy at birth, by 2010 ([23]). Action on women’s obesity in routine and manual groups was therefore important for meeting the infant mortality goal.
Tackling Health Inequalities: 10 Years On ([24]) examined key interventions and their potential contribution to meeting the infant mortality target. Reducing obesity within routine and manual groups to 23% was shown to have the potential to contribute 2.8 percentage points reduction in the gap in infant mortality rate between routine and manual groups and higher socio-economic groups.
Tackling Health Inequalities: 10 Years On used a framework of indicators proposed by the WHO Commission on Health Inequalities ([25]) to help analyse the contribution of different social determinants to the overall burden of health inequalities within England. One of the 39 indicators is prevalence of obesity in adults by gender and income quintile. To assess performance against this indicator, the report examined the Health Survey for England data from 1998, 2003 and 2006. This showed that there was a consistent pattern of increasing obesity as household income decreased for women. The pattern for men was more complex, with data from 2003 and 2006 showing no relationship between equivalised household income and obesity. The continued disparity between obesity prevalence in women from routine and manual groups and women from other socio-economic groups showed there was more work to be done, which could help to achieve the inequalities target.