Abstract

Background

Estimates of the economic cost of risk factors for chronic disease to the NHS provide evidence for prioritization of resources for prevention and public health. Previous comparable estimates of the economic costs of poor diet, physical inactivity, smoking, alcohol and overweight/obesity were based on economic data from 1992–93.

Methods

Diseases associated with poor diet, physical inactivity, smoking, alcohol and overweight/obesity were identified. Risk factor-specific population attributable fractions for these diseases were applied to disease-specific estimates of the economic cost to the NHS in the UK in 2006–07.

Results

In 2006–07, poor diet-related ill health cost the NHS in the UK £5.8 billion. The cost of physical inactivity was £0.9 billion. Smoking cost was £3.3 billion, alcohol cost £3.3 billion, overweight and obesity cost £5.1 billion.

Conclusion

The estimates of the economic cost of risk factors for chronic disease presented here are based on recent financial data and are directly comparable. They suggest that poor diet is a behavioural risk factor that has the highest impact on the budget of the NHS, followed by alcohol consumption, smoking and physical inactivity.

Introduction

Chronic diseases are the leading cause of death globally.1 In the UK, each year >35% of total deaths (c. 200 000) are due to cardiovascular disease (CVD), with cancer causing a further 157 000 deaths.2 The total cost of CVD to the UK economy is estimated as £30 billion, 60% due to health-care costs.3 The National Institute for Health and Clinical Excellence has emphasized the importance of taking population level measures to prevent CVD in the UK.4

Many chronic diseases share common behavioural risk factors such as tobacco smoking, unhealthy diet, lack of physical activity and alcohol use.5 It is estimated that up to 80% of heart diseases, stroke and type 2 diabetes and over a third of cancers could be prevented by eliminating those shared risk factors.6

One-fifth of UK adults regularly smoke cigarettes2 and more than a fifth of men and around 15% of women regularly exceed government benchmarks for heavy drinking.7 Self-reported physical activity levels in England and Scotland have slowly increased over the last 10 years; however, accelerometry data suggest that actual physical activity uptake may be far lower.8 Progress towards the target consumption level of five portions of fruit and vegetables a day has been slow and the intake of sugar and saturated fat remains higher than recommended levels.9

Clear estimates of the economic burden of these risk factors are important to help prioritize public health strategies. Recent work has established a method of producing comparable results for the economic burden to the NHS in the UK of disease associated with poor diet, smoking, alcohol intake, physical inactivity and overweight/obesity.10–14 The methods used to develop these results rely on an estimate of the percentage of the total budget of the NHS that is spent on different diseases. Until recently, the most up-to-date breakdown of the NHS budget by disease category was produced in 1996, and referred to the budget of the NHS in 1992–93.15 Therefore, the estimates of the economic cost of poor diet, physical inactivity, smoking, alcohol and overweight/obesity were based on the assumption that the contribution of different disease categories to the overall NHS budget has not substantially changed since 1992–93. Recently, a new assessment of NHS costs by disease category has been produced, based on the 2006–07 NHS budget for England. This paper reports on an update of the estimates of the economic burden to the NHS of poor diet, physical inactivity, smoking, alcohol and overweight/obesity to 2006–07 prices, using the updated data on cost by disease category.

Methods

Estimates were developed and tested using these steps:10–14 Cost data (Step 2) were updated using 2006–07 NHS costs by disease category from the NHS Programme Budgeting estimates for England from 2006 to 07.16 Previously these estimates were taken from 1992 to 93 cause-specific cost estimates15 and inflated to 2002 costs for each risk factor listed in Step 1. The disease-specific cost estimates were collected from all PCTs in England, who were asked to categorize all spending in primary and secondary care services into 23 broad and 57 more detailed disease categories, based on the tenth revision of the International Classification of Disease. The more detailed categorization was used for the analyses reported here. The programme budgeting database only provides estimates for costs for the NHS in England. For each disease, we applied the percentage of all costs in England to the total budget of the NHS in the UK (under the assumption that the distribution of NHS costs between diseases in England will be broadly representative of the situation in the UK). The total budget of the NHS in the UK for 2006–07 was provided by combining budgets for the NHS in England,16 Scotland,17 Wales18 and Northern Ireland.19

  • Identification of diseases where poor diet, physical inactivity, smoking, alcohol or overweight/obesity are risk factors.

  • Identification of the total economic cost to the NHS in the UK for these diseases.

  • Identification of the population attributable fractions (PAFs) relating to the risk factors for each disease.

  • Application of these PAFs to economic cost data, to calculate the direct burden of those risk factors.

  • Sensitivity analysis for each estimate.

In some cases, the disease categorizations provided in the 1996 burden of disease report15 and the 2006–07 programme budgeting database had changed. Table 1 provides a comparison of the disease categories in the 1996 and 2008 reports.

Table 1

Comparison of disease categorizations used in World Health Report5, published estimates of economic costs related to risk factors and updated disease categorizations provided in the NHS 2006–07 Programme Budgeting Manual16

Risk factor Associated condition (World Health Report) Disease categorization (1992–93 cost categories)a Disease categorization (2006–07 cost categories) 
Overweight/obesity Diabetes mellitus Diabetes mellitus (250) Diabetes 
Ischaemic heart disease Ischaemic heart disease (410–414) Coronary heart disease 
Ischaemic stroke Stroke (430–438) Cerebrovascular disease 
Hypertensive disease Hypertension (401–405) Problems of circulation (other)b 
Breast cancer Female breast cancer (174) Cancer, breast 
Colon/rectum cancer Colorectal cancer (153–154) Cancer, lower gastrointestinal system 
Corpus uteri cancer Genito-urinary cancers, female (179, 181–184) Cancer, gynaecological 
Osteoarthritis Osteoarthritis (715) Problems of musculoskeletal system 
Poor diet N/A Cardiovascular disease (390–459) Problems of circulation 
N/A Cancer (140–239) Cancers and tumours 
N/A Diabetes and other disease of the hormone and immune system (240–279) Endocrine, nutritional and metabolic disorders 
N/A Dental caries and other diseases of the digestive system (520–571) Dental problems 
Smoking Chronic obstructive pulmonary disease Chronic obstructive pulmonary disease (490–492, 494–496) Obstructive airways disease 
Mouth and oropharynx cancers Mouth cancers (140–149) Cancer, head and neck 
Trachea/bronchus/lung cancers Larynx, lung and other respiratory cancers (161–165) Cancer, lung 
Other cancers Other cancers (150–159, 170–239) Other cancersc 
Other medical conditions Ulcers and respiratory conditions excluding chronic obstructive pulmonary disease (460–487, 493, 500–519, 531–533) Problems of the respiratory system (other)d, problems of the gastrointestinal system 
CVD CVD (390–459) Problems of circulation 
Alcohole Cirrhosis of the liver Chronic liver disease and cirrhosis (571) Hepatobiliary 
Epilepsy Epilepsy (345) Neurological (other)f 
Strokeg Stroke (430–438) Cerebrovascular disease 
Ischaemic heart disease Ischaemic heart disease (410–414) Coronary heart disease 
Unipolar depressive disorders Other non-organic psychoses (296–299) Mental health disorders, otherh 
Liver cancer Kidney and other urinary cancers (189) Cancer, urological 
Mouth and oropharynx cancers Mouth cancers (140–149) Cancer, head and neck 
Oesophagus cancers Oesophagus cancer (150) Cancer, upper gastrointestinal system 
Other cancers Other cancers (151–188, 190–239) Other cancersi 
Physical inactivity Ischaemic heart disease Ischaemic heart disease (410–414) Coronary heart disease 
Ischaemic stroke Stroke (430–438) Cerebrovascular disease 
Breast cancer Female breast cancer (174) Cancer, breast 
Colon/rectum cancer Colorectal cancer (153–154) Cancer, lower gastrointestinal system 
Diabetes mellitus Diabetes mellitus (250) Diabetes 
Risk factor Associated condition (World Health Report) Disease categorization (1992–93 cost categories)a Disease categorization (2006–07 cost categories) 
Overweight/obesity Diabetes mellitus Diabetes mellitus (250) Diabetes 
Ischaemic heart disease Ischaemic heart disease (410–414) Coronary heart disease 
Ischaemic stroke Stroke (430–438) Cerebrovascular disease 
Hypertensive disease Hypertension (401–405) Problems of circulation (other)b 
Breast cancer Female breast cancer (174) Cancer, breast 
Colon/rectum cancer Colorectal cancer (153–154) Cancer, lower gastrointestinal system 
Corpus uteri cancer Genito-urinary cancers, female (179, 181–184) Cancer, gynaecological 
Osteoarthritis Osteoarthritis (715) Problems of musculoskeletal system 
Poor diet N/A Cardiovascular disease (390–459) Problems of circulation 
N/A Cancer (140–239) Cancers and tumours 
N/A Diabetes and other disease of the hormone and immune system (240–279) Endocrine, nutritional and metabolic disorders 
N/A Dental caries and other diseases of the digestive system (520–571) Dental problems 
Smoking Chronic obstructive pulmonary disease Chronic obstructive pulmonary disease (490–492, 494–496) Obstructive airways disease 
Mouth and oropharynx cancers Mouth cancers (140–149) Cancer, head and neck 
Trachea/bronchus/lung cancers Larynx, lung and other respiratory cancers (161–165) Cancer, lung 
Other cancers Other cancers (150–159, 170–239) Other cancersc 
Other medical conditions Ulcers and respiratory conditions excluding chronic obstructive pulmonary disease (460–487, 493, 500–519, 531–533) Problems of the respiratory system (other)d, problems of the gastrointestinal system 
CVD CVD (390–459) Problems of circulation 
Alcohole Cirrhosis of the liver Chronic liver disease and cirrhosis (571) Hepatobiliary 
Epilepsy Epilepsy (345) Neurological (other)f 
Strokeg Stroke (430–438) Cerebrovascular disease 
Ischaemic heart disease Ischaemic heart disease (410–414) Coronary heart disease 
Unipolar depressive disorders Other non-organic psychoses (296–299) Mental health disorders, otherh 
Liver cancer Kidney and other urinary cancers (189) Cancer, urological 
Mouth and oropharynx cancers Mouth cancers (140–149) Cancer, head and neck 
Oesophagus cancers Oesophagus cancer (150) Cancer, upper gastrointestinal system 
Other cancers Other cancers (151–188, 190–239) Other cancersi 
Physical inactivity Ischaemic heart disease Ischaemic heart disease (410–414) Coronary heart disease 
Ischaemic stroke Stroke (430–438) Cerebrovascular disease 
Breast cancer Female breast cancer (174) Cancer, breast 
Colon/rectum cancer Colorectal cancer (153–154) Cancer, lower gastrointestinal system 
Diabetes mellitus Diabetes mellitus (250) Diabetes 

aFurther conditions that fall under ‘trauma and injuries’ are covered in Table 2.

bIncludes all diseases of the circulatory system except ‘coronary heart disease’, ‘cerebrovascular disease' and ‘problems with rhythm’.

cAll cancers other than ‘cancer, head and neck’ and ‘cancer, lung’.

dAll problems of the respiratory system excluding ‘obstructive airways disease’.

eICD-9 codes in brackets

fAll neurological conditions excluding ‘chronic pain’.

gThe World Health Report provides separate PAFs for haemorrhagic stroke and ischaemic stroke. Since cost data are not available by stroke subtype, we have averaged the PAFs using the weighting of 2:1 in favour of ischaemic stroke, in accordance with the estimated ratio of mortalities by stroke subtype in the UK.41

hAll mental health disorders excluding ‘substance misuse’, ‘organic mental disorders’, ‘psychotic disorders' and ‘child and adolescent mental health disorders’.

iAll cancers other than ‘cancer, urological’, ‘cancer, head and neck’ and ‘cancer, upper gastrointestinal system’.

For the calculation of the burden of alcohol-related ill health, special attention was made to the 2006–07 disease categorization of ‘trauma and injuries’. This broad category needed to be further broken down to drowning, falls, motor vehicle accidents, self-inflicted injuries, and other unintentional injuries in order to apply the necessary PAFs. This further stratification was performed by applying the proportion of deaths from each sub category (compared with all deaths from trauma and injury) to the cost of all trauma and injury (see Table 2). The mortality data were taken from the World Health Organization's (WHO) estimates of mortality for WHO member states in 2004.20 These estimates were derived by the WHO using disease-specific death certification data from the UK for an earlier year, and then projecting estimates to 2004.

Table 2

Further stratification of ‘trauma and injuries’ disease category for calculation of the economic costs to the NHS of alcohol-related conditions

World Health Report Category Number of deaths, 2004 Percentage of all trauma and injury deaths Estimated percentage of NHS costs 
Drowning 2000 10 0.4 
Falls 4000 19 0.8 
Motor vehicle accidents 3700 18 0.8 
Self-inflicted injuries 4700 23 1.0 
Other unintentional injuries 5200 25 1.1 
All trauma and injuries 20 770 100 4.3 
World Health Report Category Number of deaths, 2004 Percentage of all trauma and injury deaths Estimated percentage of NHS costs 
Drowning 2000 10 0.4 
Falls 4000 19 0.8 
Motor vehicle accidents 3700 18 0.8 
Self-inflicted injuries 4700 23 1.0 
Other unintentional injuries 5200 25 1.1 
All trauma and injuries 20 770 100 4.3 

For physical inactivity, smoking, alcohol consumption and overweight/obesity, the counterfactuals associated with the estimates of economic burden are drawn from the WHO global burden of disease project, and relate to the theoretical minimum within a population. The counterfactuals were therefore zero smoking, zero alcohol consumption and everyone within the population achieving at least some activity at work, home, for transport or during discretionary time. Because of the J-shaped relationship between alcohol intake and many health outcomes, this counterfactual allows for the increase in risk for some individuals and decrease in risk for others.21 For overweight/obesity, the theoretical minimum was taken to be the part of the distribution associated with the lowest risk, i.e. everyone in the population with a body mass index of 21 kg/m2.21

The counterfactual for poor diet is more complicated as diets are multi-factorial. The global burden of disease project calculated PAFs for individual risk factors and warns that the PAF for combined risk factors for any specific disease may be less than the sum of PAFs for individual risk factors. In order to calculate a combined PAF for diet, we triangulated estimates of the percentage of disability adjusted life years (DALYs) that were due to diet-related risk factors in the WHO European region,22 the EU23 and Australia,24 and conservatively estimated that 10% of DALYs were due to diet.14 Since 30% of DALYs in the WHO European region were due to CVD and cancer, we assumed a PAF of 0.33 for diet. Using this method, it should be assumed that the counterfactual for poor diet is a combination of all counterfactuals for diet-related risk factors in the global burden of disease project: everyone consuming 600 g per day of fruit and vegetables, having total cholesterol levels of 3.8 mmol/l and a body mass index of 21 kg/m2.

A sensitivity analysis was conducted to assess how sensitive the results were to the assumption that PAFs generated for the WHO EUR-A region were suitable for use in the UK. Prevalence rates used to generate the PAFs for each risk factor were compared with equivalent rates drawn from the Health Survey for England 200625 and the General Household Survey 2006.26 The ratio between the UK prevalence rate and the EUR-A prevalence rate was applied to the modelled cost estimates, and the range for the sensitivity analysis was between the estimates adjusted for UK prevalence rates and the original estimates using the EUR-A prevalence rates. The cost estimates for poor diet were not included in the sensitivity analysis because of the difficulties in defining PAFs for poor diet, as described already.

Results

In 2006–07, 46% of total NHS costs (over £43 billion) were due to diseases related to poor diet, physical inactivity, smoking, alcohol and overweight/obesity (although not all of these costs were due to the risk factors, since not all of the incident cases of these conditions were due to the risk factors) (Table 3). Of the behavioural risk factors, £5.8 billion was spent on poor diet-related ill health, £3.3 billion on alcohol-related ill health, £3.3 billion on smoking-related ill health and £0.9 billion on physical inactivity-related ill health. Over £5 billion was spent on overweight/obesity-related ill health, although this includes costs from poor diet and physical inactivity.

Table 3

Original and updated economic costs of poor diet, physical inactivity, smoking, alcohol and overweight/obesity to the NHS

 % of total NHS costs
 
PAF Total NHS costs (£millions)
 
 1992–93 2006–07  Original costs (2002) Updated costs (2006–07) 
Physical inactivity 
 Ischaemic heart disease 2.6 2.9 23 526 542 
 Ischaemic stroke 4.1 1.2 12 347 117 
Breast cancer 0.3 0.6 11 26 54 
 Colon/rectum cancer 0.5 0.5 16 61 65 
 Diabetes mellitus 1.0 1.3 15 101 158 
 Total 8.5 6.5  1062 936 
Overweight and obesity    Original costs (2002) Updated costs (2006–07) 
 Ischaemic heart disease 2.6 2.9 34 778 801 
 Ischaemic stroke 4.1 1.2 34 983 332 
 Breast cancer 0.3 0.6 12 29 59 
 Colon/rectum cancer 0.5 0.5 16 61 65 
 Hypertensive disease 1.4 4.5 58 576 2121 
 Corpus uteri cancer 0.1 0.2 49 41 80 
 Osteoarthritis 1.6 5.0 21 229 853 
 Diabetes mellitus 1.0 1.3 79 533 835 
 Total 11.6 16.2  3231 5146 
Poor diet    Original costs (2002) Updated costs (2006–07) 
 CVD 12.0 9.2 33 2780 2468 
 Diabetes mellitus 2.0 2.8 33 463 751 
 Cancer 4.0 6.2 33 927 1663 
 Dental caries 8.0 3.4 33 1853 912 
 Total 26.0 21.6  6023 5793 
Smoking    Original costs (2006) Updated costs (2006–07) 
 Trachea/bronchus/lung cancers 0.4 0.3 85 235 207 
 Other cancers 3.7 5.7 10 306 463 
 Chronic obstructive pulmonary disease 2.2 0.8 68 1188 442 
 CVD 12.1 9.2 22 2134 1645 
 Other medical conditions 5.8 5.7 11 536 510 
 Total 24.3 18.9  4399 3267 
Alcohol    Original costs (2006) Updated costs (2006–07) 
 Cirrhosis of the liver 0.7 0.9 56 319 410 
 Epilepsy 0.7 2.7 41 243 900 
 Unipolar depressive disorders 0.0 7.7 313 
 Ischaemic heart disease 2.6 2.9 −1 −21 −24 
 Stroke 3.2 1.2 0.7 0.7 
 Mouth and oropharynx cancers 0.0 0.2 35 57 
 Oesophagus cancer 0.1 0.3 41 31 100 
 Liver cancer 0.0 0.6 32 10 156 
 Other cancer 3.9 5.1 10 291 415 
 Drowning 0.0 0.4 34 115 
 Falls 0.9 0.8 15 120 101 
 Poisoning 0.1 0.2 39 35 63 
 Motor vehicle accidents 0.9 0.8 38 284 238 
 Self-inflicted injuries 1.0 1.0 20 181 159 
 Other intentional injuries 0.0 0.0 26 
 Other unintentional injuries 0.7 1.1 24 147 211 
 Total 14.9 25.9  1742 3228 
 % of total NHS costs
 
PAF Total NHS costs (£millions)
 
 1992–93 2006–07  Original costs (2002) Updated costs (2006–07) 
Physical inactivity 
 Ischaemic heart disease 2.6 2.9 23 526 542 
 Ischaemic stroke 4.1 1.2 12 347 117 
Breast cancer 0.3 0.6 11 26 54 
 Colon/rectum cancer 0.5 0.5 16 61 65 
 Diabetes mellitus 1.0 1.3 15 101 158 
 Total 8.5 6.5  1062 936 
Overweight and obesity    Original costs (2002) Updated costs (2006–07) 
 Ischaemic heart disease 2.6 2.9 34 778 801 
 Ischaemic stroke 4.1 1.2 34 983 332 
 Breast cancer 0.3 0.6 12 29 59 
 Colon/rectum cancer 0.5 0.5 16 61 65 
 Hypertensive disease 1.4 4.5 58 576 2121 
 Corpus uteri cancer 0.1 0.2 49 41 80 
 Osteoarthritis 1.6 5.0 21 229 853 
 Diabetes mellitus 1.0 1.3 79 533 835 
 Total 11.6 16.2  3231 5146 
Poor diet    Original costs (2002) Updated costs (2006–07) 
 CVD 12.0 9.2 33 2780 2468 
 Diabetes mellitus 2.0 2.8 33 463 751 
 Cancer 4.0 6.2 33 927 1663 
 Dental caries 8.0 3.4 33 1853 912 
 Total 26.0 21.6  6023 5793 
Smoking    Original costs (2006) Updated costs (2006–07) 
 Trachea/bronchus/lung cancers 0.4 0.3 85 235 207 
 Other cancers 3.7 5.7 10 306 463 
 Chronic obstructive pulmonary disease 2.2 0.8 68 1188 442 
 CVD 12.1 9.2 22 2134 1645 
 Other medical conditions 5.8 5.7 11 536 510 
 Total 24.3 18.9  4399 3267 
Alcohol    Original costs (2006) Updated costs (2006–07) 
 Cirrhosis of the liver 0.7 0.9 56 319 410 
 Epilepsy 0.7 2.7 41 243 900 
 Unipolar depressive disorders 0.0 7.7 313 
 Ischaemic heart disease 2.6 2.9 −1 −21 −24 
 Stroke 3.2 1.2 0.7 0.7 
 Mouth and oropharynx cancers 0.0 0.2 35 57 
 Oesophagus cancer 0.1 0.3 41 31 100 
 Liver cancer 0.0 0.6 32 10 156 
 Other cancer 3.9 5.1 10 291 415 
 Drowning 0.0 0.4 34 115 
 Falls 0.9 0.8 15 120 101 
 Poisoning 0.1 0.2 39 35 63 
 Motor vehicle accidents 0.9 0.8 38 284 238 
 Self-inflicted injuries 1.0 1.0 20 181 159 
 Other intentional injuries 0.0 0.0 26 
 Other unintentional injuries 0.7 1.1 24 147 211 
 Total 14.9 25.9  1742 3228 

Total NHS budget 2002: £70.2 billion; 2006: £94.2 billion; 2006–07: £81.3 billion.

Proportional NHS costs due within disease categories were similar in 1992–93 and 2006–07 for most of the diseases with some notable exceptions. For example, the percentage of NHS costs due to cerebrovascular disease fell from 4.1% in 1992–93 to 1.2% in 2006–07, and costs due to osteoarthritis increased from 1.6% in 1992–93 to 5.0% in 2006–07. A combination of the changes in costs by disease, the differences in disease categorization displayed in Table 1 and the change from 2002 or 2006 prices to 2006–07 prices explains the difference between the original estimates and the updated estimates presented here. The original estimates of the cost of poor diet, physical inactivity and overweight/obesity were based on 2002 prices; so it would be expected that updating to 2006–07 prices would increase the estimates. That has been the case for the costs of overweight/obesity, which increased from £3.2 to £5.1 billion. However, the costs for both physical inactivity and poor diet have slightly reduced in the updated estimates, because of a downwards revision in the total percentage of NHS costs of related conditions between 1992–93 and 2006–07. The original estimates for the cost of smoking and alcohol were based on 2006 prices, where the estimate used for the total budget for the NHS (£94.2 billion) was higher than that used in 2006–07 (£81.3 billion). This difference explains some of the reduction in the estimate of smoking-related ill health, from £4.4 to £3.3 billion. The estimate for the cost of alcohol-related ill health increased, due to an increase in the percentage of NHS costs due to alcohol-related conditions.

Table 4 shows the results of the sensitivity analysis. For physical inactivity, alcohol consumption, and overweight and obesity, the prevalence estimates for the WHO EUR-A region were very similar to equivalent estimates for the UK in 2006, and as such the sensitivity range for the costs associated with these risk factors is narrow (ranging from £5.1 to £5.2 billion for overweight and obesity). This was not the case for smoking, where the UK prevalence of smoking is considerably lower than the prevalence in the EUR-A region. The sensitivity range for smoking was £2.3 to £3.3 billion.

Table 4

Sensitivity analysis of the impact of using EUR-A PAFs to generate cost estimates for the UK

Risk factor EUR-A prevalence rate UK prevalence rate Adjustment factora Sensitivity range of total costs to NHS 
Physical inactivityb 17 18c 1.05 £0.9–£1.0 billion 
Overweight and obesityd 26.7 27.2c 1.01 £5.1–£5.2 billion 
Smokinge 32 22f 0.69 £2.3–£3.3 billion 
Alcohol intakeg 87 86c 0.99 £3.2–£3.2 billion 
Risk factor EUR-A prevalence rate UK prevalence rate Adjustment factora Sensitivity range of total costs to NHS 
Physical inactivityb 17 18c 1.05 £0.9–£1.0 billion 
Overweight and obesityd 26.7 27.2c 1.01 £5.1–£5.2 billion 
Smokinge 32 22f 0.69 £2.3–£3.3 billion 
Alcohol intakeg 87 86c 0.99 £3.2–£3.2 billion 

aAdjustment factor = ratio of UK prevalence rate to EUR-A prevalence rate.

b‘Prevalence rate’ refers to percentage of adults who do no or very little physical activity.

cEstimate taken from 2006 Health Survey for England.

d‘Prevalence rate’ refers to mean BMI.

e‘Prevalence rate’ refers to percentage of adult current smokers.

fEstimate taken from 2006 General Household Survey.

g‘Prevalence rate’ refers to percentage of adults who consume no alcohol at all.

Discussion

One of the aims of deriving comparable estimates of the cost of risk factors to the NHS is to provide evidence to allow policy-makers to prioritize interventions aimed at preventing chronic disease. The update of cost estimates using more recent data on cost by disease category is therefore important to ensure that the evidence is up-to-date and reliable. The updated estimates reinforce that the largest economic burden to the NHS is due to poor diet and that much of this food-related burden is due to overweight and obesity. Physical inactivity asserts a considerable burden on the NHS, but not as high as other behavioural risk factors.

The main difference between the results presented here and the original estimates is that the economic burden of alcohol is now similar to the burden due to smoking. It is important to remark that this change in priority is not directly due to changes in the prevalence of smoking or binge drinking in the UK (since the PAFs used to calculate both sets of estimates were the same). Rather, it is due to changes in the disease-specific costs associated with alcohol and smoking. In turn, these differences in costs may be due to changes in the incidence of these diseases (which in turn could be influenced by the prevalence of smoking or binge drinking) or changes in the cost of treatment of these diseases. For example, the impact of unipolar depressive disorders (which are associated with alcohol intake) on the NHS budget has changed considerably between the 1991–9215 and 2006–07 estimates27—due in part to differences in coding these conditions in the two reports (see Table 1). Although changes in coding between the reports hinder comparisons between the two sets of estimates, use of the updated 2006–07 costs ensure a more realistic estimate of the burden of risk factors to the NHS.

The analysis for this paper required a comparison of two sets of cause-specific NHS costs (for the years 1992–93 15 and 2006–7,16 which revealed some interesting changes in relative cost of conditions to the NHS. Such comparisons can be problematic: percentages may change due to a change in spending on a specific condition, or a change in spending on other conditions; changes in spending on a specific condition could reflect changes in incidence, prevalence or treatment methods for that condition. Cerebrovascular disease was responsible for 4.1% of NHS costs in 1992–93 and only 1.2% of NHS costs in 2006–07, which seems at odds with the ageing population of the UK. While incidence rates for stroke are falling in the UK, the prevalence of stroke is increasing.28 It seems likely, therefore, that this change in relative burden to the NHS is either a reflection of considerable uncertainty in the cause-specific NHS cost estimates, or is due to changes in treatment methods for stroke since the early 1990s that have reduced the financial burden to the NHS.

The National Audit Office (NAO) estimated that the direct cost of overweight and obesity to the NHS in 2002 was £1.1 billion.29 This figure provided a baseline value for simulations of obesity costs in the Foresight project.30 It is considerably lower than the equivalent reported in this paper, mainly due to the counterfactuals used in the two analyses (BMI = 25 for the NAO report, BMI = 21 here). The cabinet office estimated that population meeting national nutritional guidelines would provide health benefits of £19.9 billion each year in quality-adjusted life years (QALYs).31 Our estimate is much lower but does not put an economic value on QALYs. The Game Plan32 estimated that the total cost of physical inactivity is £1.9 billion per annum (£0.3 billion NHS, £0.8 billion work absence and £0.8 billion early mortality)—direct cost to the NHS was higher in our estimate of £0.9 billion, although the authors of Game Plan suggested that the cost to the NHS may be as high as £1.7 billion per annum. Previous estimates of the cost of alcohol consumption to the NHS33–35 ranged from £1.4 to £1.7 billion (2.5% of total NHS cost) for England and £ 95.6 million for Scotland (9% of total NHS cost). Our cost estimate is between these two estimates when judged as a proportion of NHS costs (4%). Our updated estimates of the cost of smoking was £3.3 billion while Parrott et al.,36 estimated the cost to be around £1.4 billion in the UK in 1991, and Buck et al.,37 estimated it to be £1.4–1.7 billion in the UK in the same year, but comparisons with our estimate are hampered by a 15 year gap between the studies.

The methods used to develop the estimates in this paper rely on the PAFs calculated by the WHO.5,38,39 These PAFs are based on broad WHO regions (specifically the EUR-A region of developed European countries with very low child and adult mortality) and as such they may not accurately represent the picture in the UK. PAFs must take account of the underlying prevalence of a risk factor within a population (since, for example, in a population with zero smokers, none of the ischaemic heart disease could be attributable to smoking), and the use of WHO regional PAFs will therefore affect the accuracy of these estimates. This limitation has been addressed by a sensitivity analysis, which suggests that using EUR-A PAFs for the UK has very little impact on the estimated costs for alcohol consumption, physical inactivity and overweight/obesity, but the results for smoking are more sensitive to this assumption. This is almost entirely due to differences in male smoking rates—the WHO estimates the prevalence of male smoking in the EUR-A region to be 37% compared with only 23% in the UK.25 This suggests that the cost of alcohol consumption in the UK may be higher than the cost of smoking, and this should be considered when using these results to prioritize public health issues in the UK. A separate but related issue is the uncertainty that accompanies these estimates of the cost of risk factors in the UK regarding the PAFs and the use of cause-specific NHS cost estimates. The WHO suggests that the uncertainty associated with the PAFs for physical inactivity, smoking and alcohol consumption is ±10%, whereas the uncertainty for the overweight/obesity PAF is about ±3%.21 The uncertainty of cause-specific NHS cost estimates provides a significant limitation of the estimates presented here.

A further limitation is that the range of diseases associated with each risk factor was taken to be those for which the WHO has calculated a PAF, although there are possibly other additional conditions associated with these risk factors (e.g. physical activity and depression.40 The possible overlap between risk factors (such as overweight and obesity) was not addressed here but should be considered when calculating the total economic burden of these risk factors.

This study developed a well-established methodology10–14 by updating cost data by disease category from 1992–93 to 2006–07. The resultant estimates of the cost of risk factors are directly comparable, based on recent financial estimates and can easily be repeated in future years. This updated comparison of costs should help policy-makers to prioritize resource allocation and make effective use of limited NHS resources.

Funding

Peter Scarborough, Prachi Bhatanagar, Kremlin K. Wickramasinghe and Steven Allender were supported by the British Heart Foundation, grant number PPC/Jul05/5b. Mike Rayner and Charlie Foster were supported by the British Heart Foundation, grant number PPC/JAN05/6biii. The funders had no role in study design, collection, analysis and interpretation of data, writing of the report, or decision to submit the paper for publication.

References

1
Yach
D
Hawkes
C
Gould
CL
, et al.  . 
The global burden of chronic diseases: overcoming impediments to prevention and control
JAMA
 , 
2004
, vol. 
291
 
21
(pg. 
2616
-
22
)
2
Allender
S
Peto
V
Scarborough
P
, et al.  . 
Coronary heart disease statistics
 , 
2008
London
British Heart Foundation
3
Luengo-Fernandez
R
Leal
J
Gray
A
, et al.  . 
Cost of cardiovascular diseases in the United Kingdom
Heart
 , 
2006
, vol. 
92
 
10
(pg. 
1384
-
89
)
4
NICE
Prevention of Cardiovascular Disease at Population Level
 , 
2010
London
National Institute for Health and Clinical Excellence
5
WHO
World Health Report 2002
2003
Geneva
World Health Organization
6
WHO
2008–2013 action plan for global stratedy for the prevention and control of noncommunicable diseases
2008
Geneva
World Health Organization
7
General Lifestyle Survey-Smoking and Drinking Among Adults, 2008
 , 
2008
London
Office for National Statistics
8
Health Survey for England 2008—Physical Activity and Fitness
 , 
2009
Leeds
The Health and Social Care Information Centre
9
National Diet and Nutrition Survey
 , 
2010
London
Food Standards Agency
10
Allender
S
Foster
C
Scarborough
P
, et al.  . 
The burden of physical activity-related ill health in the UK
J Epidemiol Community Health
 , 
2007
, vol. 
61
 
4
(pg. 
344
-
48
)
11
Allender
S
Balakrishnan
R
Scarborough
P
, et al.  . 
The burden of smoking-related ill health in the UK
Tob Control
 , 
2009
, vol. 
18
 
4
(pg. 
262
-
67
)
12
Allender
S
Rayner
M
The burden of overweight and obesity-related ill health in the UK
Obes Rev
 , 
2007
, vol. 
8
 
5
(pg. 
467
-
73
)
13
Balakrishnan
R
Allender
S
Scarborough
P
, et al.  . 
The burden of alcohol-related ill health in the United Kingdom
J Public Health (Oxf)
 , 
2009
, vol. 
31
 
3
(pg. 
366
-
73
)
14
Rayner
M
Scarborough
P
The burden of food related ill health in the UK
J Epidemiol Community Health
 , 
2005
, vol. 
59
 
12
(pg. 
1054
-
57
)
15
Department of Health
Burden of Disease
 , 
1996
Wetherby
Department of Health
 
A Discussion Document
16
Programme Budgeting Guidance 2006–07
 , 
2008
London
Department of Health
17
Scottish Government Health Department
Scottish Health Service Costs, Year Ended 31st March 2007
 , 
2008
Edinburgh Information Services Division
 
Edinburgh
18
Welsh Assembly Government
NHS Expenditure Programme Budgets: 2003–04 to 2006–07
 , 
2008
Cardiff
19
Department of Health, Social Service, Public Safety
Summary of HPSS Expenditure in Northern Ireland
 , 
2008
Belfast
Northern Ireland Executive
20
WHO
Disease and injury country estimates
2009
World Health Organization
21
Ezzati
M
Lopez
AD
Rodgers
A
, et al.  . 
WHO
Comparative quantification of health risks [CD ROM]
2004
Geneva
22
Ezzati
M
Lopez
AD
Rodgers
A
, et al.  . 
Comparative Quantification of Health Risks
 , 
2004
Geneva
World Health Organization
23
NIPH
Determinants of the Burden of Disease in the European Union
 , 
1997
Stockholm
National Institute of Public Health
24
Mathers
CD
Vos
ET
Stevenson
CE
, et al.  . 
The burden of disease and injury in Australia
Bull World Health Organ
 , 
2001
, vol. 
79
 
11
(pg. 
1076
-
84
)
25
Health Survey for England 2006—Cardiovascular Disease and Risk Factors in Adults
 , 
2008
Leeds
The Information Center
26
General Household Survey 2006—Smoking and Drinking among Adults
 , 
2008
London
Office for National Statistics
27
Department of Health SS, Public Safety. Resource Accounts 2005–06.
 , 
2007
London
Stationery Office
 
HC 1680: November 2006
28
Scarborough
P
Peto
V
Bhatnagar
P
, et al.  . 
Stroke Statistics
 , 
2009
London
British Heart Foundation and Stroke Association
29
Commitee
HoCH: Obesity: Third Report of Session 2003–04/Great Britain
2004
, vol. 
Volume 1
 
London
Parliament
30
Butland
B
Jebb
S
Kopelman
P
, et al.  . 
Foresight. Tackling Obesities: Future Choices
 , 
2007
London
Government Office for Sciences
31
Office C
Food: An Analysis of the Issues
 , 
2008
London
The Strategy Unit
32
Unit DoCMaSS
Game Plan: A Strategy for Delivering Government's Sport and Physical Activity Objectives
 , 
2002
London
Cabinet Office
33
Varney
SJ
Guest
JF
The annual societal cost of alcohol misuse in Scotland
Pharmacoeconomics
 , 
2002
, vol. 
20
 
13
(pg. 
891
-
907
)
34
London RCoPo
ALCOHOL—can the NHS afford it?
2001
London
Royal College of Physicians
 
A Report of a working Party of the Royal College of Physicians
35
Office C
Alcohol Misuse: How Much Does it Cost?
 , 
2001
London
Cabinet Office
36
Parrott
S
Godfrey
C
Raw
M
, et al.  . 
Guidance for commissioners on the cost effectiveness of smoking cessation interventions. Health Educational Authority
Thorax
 , 
1998
, vol. 
53
 
Suppl. 5 Pt 2
(pg. 
S1
-
38
)
37
Buck
D
Godfrey
C
Parrott
S
Cost Effectiveness of Smoking Cessation Interventions
 , 
1997
London
Health Education Authority
38
Murray
CJ
Ezzati
M
Lopez
AD
, et al.  . 
Comparative quantification of health risks conceptual framework and methodological issues
Popul Health Metr
 , 
2003
, vol. 
1
 
1
pg. 
1
 
39
Murray
CJ
Lopez
AD
Global mortality, disability, and the contribution of risk factors: Global Burden of Disease Study
Lancet
 , 
1997
, vol. 
349
 
9063
(pg. 
1436
-
42
)
40
Services USDoHaH
Effects of Physical Activity on Health and Disease: A Report from the Surgeon General
 , 
1996
Atlanta
Centers for Disease Control and Prevention
41
Lawlor
DA
Smith
GD
Leon
DA
, et al.  . 
Secular trends in mortality by stroke subtype in the 20th century: a retrospective analysis
Lancet
 , 
2002
, vol. 
360
 
9348
(pg. 
1818
-
23
)