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

Gastrooesophageal reflux and BMI

There is a general understanding of a relationship between weight and increased prevalence of heartburn, or symptoms of gastrointestinal reflux. Indeed, there is a meta-analysis [1] indicating significant increase for those with a BMI of 25 kg/sq m or more compared with those with lower BMI.

However, this is something of a blunt analysis, and does not tell us much about gradations. For instance, is there a gradual increase in risk, or does the risk increase dramatically at any particular BMI? Is there any evidence of a U-shaped relationship, perhaps with higher rates in underweight people? A new study [2] fills in some of the fine details.


Part of the US Nurses study, this survey involved a questionnaire to a random selection of 12,192 nurses, with questions about frequency, severity, and duration of heartburn or acid regurgitation, using validated definitions of both terms. Severity was defined as mild (can be ignored), moderate (cannot be ignored but does not affect lifestyle), severe (affects lifestyle), and very severe (markedly affects lifestyle). Frequent was an episode occurring at least weekly.

Information was collected regarding height, weight (at various ages), drugs, diet, exercise, tobacco and alcohol use, and concurrent disease. Analysis of results used these data to examine confounding variables. Controls were women without symptoms not taking acid suppressing medicines.


The women in the survey had an average age of 66 years, and an average BMI of about 27. Women with symptoms were more likely to have a higher BMI, use medications for asthma or hypertension, or hormone preparations, consumed more calories, and were less active.

Over 10,500 questionnaires were returned, with an 86% response rate. No symptoms of heartburn or acid reflux were reported in 41% (1 in 10 of whom were using proton pump inhibitors), with the remaining 51% reporting symptoms less frequently than monthly, to daily (Figure 1). One woman in five (22%) had symptoms at least weekly. Of those with symptoms, most (95%) were moderate or mild, and only about 5% had symptoms that were severe or very severe, and most (55%) had both heartburn and acid reflux.

Figure 1: Frequency of symptoms

Using those women with frequent (at least weekly) symptoms, and women without symptoms as controls, there was increasing reporting of symptoms of heartburn or acid reflux with increasing BMI (Figure 2), even after adjusting for potential confounders. This was the case for mild, moderate and severe or very severe symptoms. With a BMI ≥25, 60% of the increased risk was accounted for by excess weight.

Figure 2: Odds ratio for heartburn or acid reflux at different BMI levels

Among women who had gained weight during the previous 14 years, there was a dose-dependent increase in the risk of symptoms, with about a threefold increase in those whose BMI increased by 3.5 units. Conversely, there was a reduction in almost 40% in the risk of frequent symptoms in women who reduced BMI by 3.5 units or more.


This nicely captures the relationship between increased risk of heartburn or acid reflux and increased weight. Being or becoming overweight doubles the risk of having these symptoms at least once a week. A back-of-the-envelope calculation gives crude results for the prevalence of moderate or severe heartburn or acid reflux symptoms at least weekly for each band of BMI and shows the gradation (Figure 3). The bottom line is that this is yet another reason to avoid being overweight, along with all the others. If our populations keep growing out as well as in numbers, we will need to step up production of the antacids.

Figure 3: Crude symptom rate and BMI


  1. H Hempel et al. Meta-analysis: obesity and the risk for gastroesophageal reflux disease and its complications. Annals of Internal Medicine 2005 143: 199-211.
  2. BC Jacobson et al. Body-mass index and symptoms of gastroesophageal reflux in women. New England Journal of Medicine 2006 354: 2340-2348.

previous or next story