Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Obesity and health

Childhood obesity and sweetened drinks
Obesity and cancer
Low fat diets produce weight loss

Bandolier was overcome with a sinking feeling when the newspapers in the UK highlighted a National Audit Office (NAO) report on obesity. It is a big problem (no pun intended), we are aware that obesity is associated with impaired health, and it is just about one of the most difficult topics to get to grips with. On the one hand it is just eating too much and exercising too little, but on the other it is associated with almost every aspect of our lives, from transport, to education, our jobs and our leisure.

Much too big a problem. And yet Bandolier's 10 tips for health living ( Bandolier 78 ) produced a huge response from professionals and the public, even being reproduced in a national newspaper.

So we visited the NAO Internet site ( ) and downloaded the report [1]. It is a terrific document, being of a reasonable length (74 pages), readable, informative and comprehensive. It may not have actual diet sheets or guaranteed ways to lose weight painlessly, but for those dealing with patients or making policy it comes into the must read category. It is particularly valuable because it examines the issues across all government activities, and because the case studies show how imaginatively the complicated issues about health promotion can be tackled.

Bandolier thought it worthwhile, therefore, to pick out some of the more useful items from the report, and to examine a few studies relating to obesity that have swum into our ken recently.


Overweight is a body mass index (weight in kg divided by height in metres squared) of 25 to 30. A BMI between 30 and 40 is obese and above 40 is morbidly or severely obese. Someone 5 feet 6 inches tall (1.68 metres) becomes obese at 13 stone 4 pounds (84 kg) and morbidly obese at 17 stones 7 pounds (110 kg) (Note that the Bandolier healthy living website has a neat coloured chart in feet, inches, metres, lbs, kg and stones and pounds, or any combination of these)

In England 1 in 5 adults are obese, and that proportion has trebled over the last 20 years. In adults over 45 years, two-thirds are overweight or obese. Obesity is more common in women in lower socioeconomic groups, and obesity is more common in some ethnic groups.

Obesity predisposes us to higher risks of associated diseases (Table 1). There is a high human cost in diabetes, hypertension and other disorders, including cancer. There is a big cost to the NHS (Figure 1; estimated by the NAO to be £500,000,000 a year in 1998, though this may be a low estimate), and a big cost to society through lost work time and economic output (estimated by the NAO to be around £2,000,000,000 a year).

Table 1: Relative risk of different diseases in obese versus nonobese people

  Relative risk  
Disease Women Men Working days lost
Type 2 diabetes 12.7 5.2 5,960,000
Hypertension 4.2 2.6 5,160,000
Heart attack 3.2 1.5 1,230,000
Colon cancer 2.7 3  
Angina 1.8 1.8 2,390,000
Gall bladder disease 1.8 1.8 20
Ovarian cancer 1.7    
Osteoarthritis 1.4 1.9 950,000
Stroke 1.3 1.3 440,000
Cancers     970,000
Working days lost are certificated absence from secondary diseases attributable to obesity

Figure 1: Annual cost of obesity-related disease in England

Childhood obesity and sweetened drinks

Sugar-sweetened soft drinks are consumed daily by about 70% of American adolescents. Excessive weight is now the most common paediatric medical problem in the USA. Is there any relationship between these two statements? A study on over 500 US children [2] suggests that there is.

The study was observational, involving comprehensive data collection on 548 children in Boston in October 1995 and May 1997. The data collected included the amount of sweetened and diet soft drinks consumed at baseline and at the follow up, and diet and energy intake, and BMI and triceps skinfold thickness, and activity analysis. It also involved some considerable analytical expertise to construct a number of different models involving possible confounders relating consumption of sweetened soft drinks to development of obesity.

There were 548 children in total, of whom 398 were not obese at baseline, but 37 of these 398 became obese at follow up. Sixty percent of the children reported drinking more sweetened soft drinks between baseline and follow up, and 25% were drinking one more serving a day.

The odds ratio of becoming obese increased by 1.6 times for each additional can or glass of sugar sweetened drink consumed every day. Consumption of diet-soda was negatively associated with obesity incidence.

Obesity and cancer

A new analysis [3] tells us more about obesity and cancer. Articles relating body weight with cancer incidence for a range of different cancers were sought and subjected to individual meta-analyses to explore the dose-response between increasing BMI and cancer risk. This information was then combined with prevalence of persons overweight (BMI 25 to 30) and obese (BMI above 30) in each of the 15 EU countries to obtain estimates of the number of cancers attributable to overweight, and the percentage of cancers attributable to overweight (for about 1995).

Information is given for each country individually, and the whole EU. The figures for the UK are given for the percentage and number of cancers in Figures 2 and 3. The total number of cancers attributable to overweight in the UK was 9,000 (it was 70,000 for the EU), and the percentage of cancers was 2.7% for men and 4.9% for women.

Figure 2: Percent of cancers related to obesity (UK)

Figure 3: Number of cancers related to obesity (UK)

Low fat diets produce weight loss

A meta-analysis is highly supportive of the role of low fat ad libitum diets [4] for weight loss in the obese. The review sought studies of at least three months duration of ad libitum diets comparing low fat with normal diet, and which reported weight change as an outcome. Fifteen publications of 16 studies were found, 13 of which were randomised. The total number of patients was 1728, of whom 1074 were women. The mean BMI was 21-29, and the studies lasted between nine weeks and 12 months.

Low fat diets produced reductions in percentage of energy from fat from 3.5% to 24%, and with mean weight losses in individual trials of up to 10 kg. The overall mean weight loss associated with a low fat diet was about 2.5 kg (results for randomised studies in Figure 4). The amount of weight loss was related to the degree of fat reduction in control and intervention groups (Figure 5, randomised studies). With no reduction in dietary fat, no weight loss occurred. For every 1% reduction in dietary fat there was a weight loss of 0.4 kg.

Figure 4: Weight change in ad libitum low fat versus normal diets

Figure 5: Relation between fat reduction and weight loss (all groups)

Weight loss was least in those with body weights in the range of 60-72 kg. Weight loss increased progressively with increased initial body weight. Extrapolated to an initial weight of 88 kg and a 10% reduction in fat, the predicted weight loss would be 4.4 kg.


The explosion in overweight and obese adults and children is a real problem with major implications for the health of individuals and for organisations delivering health care. It is a worldwide problem, and it is predicted to get much worse.

Put simply, we eat too much and do too little, and what we eat is often injurious of itself. Health promotion is clearly advocated, though this is a complex matter, and there are few beacons of success. Bandolier has had difficulty finding good quality evidence about successful delivery of health promotion messages, but that may be because we are looking in the wrong place. We'd love to hear of examples in the literature, or examples that could be written up for ImpAct .


  1. Tackling obesity in England. London: The Stationary Office, 2001.
  2. DS Ludwig et al. Relation between consumption of sugar-sweetened drinks and childhood obesity: a prospective, observational analysis. Lancet 2001 357: 505-508.
  3. A Bergstöm et al. Overweight as an avoidable cause of cancer in Europe. International Journal of Cancer 2001 91: 421-430.
  4. A Astrup et al. The role of dietary fat in body fatness: evidence from a preliminary meta-analysis of ad libitum low-fat dietary intervention studies. British Journal of Nutrition 2000 83 Suppl 1: S25-S32.
previous or next story in this issue