What is the prevalence of gastroesophageal reflux disease in a population-based study?
G.R. Locke III (Rochester)
What is prevalence?
The term, prevalence, can be defined as the proportion of the population that has a condition at a given point in time. Such data are usually obtained from cross-sectional studies and reported as percentages. Prevalence is a useful measure for chronic diseases and can be used in making health planning decisions; it is less useful for self-limited and rapidly fatal conditions.
Prevalence is often confused with incidence. The word, incidence, is used in two distinct epidemiological measures both of which measure the onset of disease. The incidence rate measures the velocity at which people develop a condition. Like any rate, the denominator contains a unit of time. Cumulative incidence represents the proportion of the population that develops a condition over a given period of time. This is a measure of risk. For rare diseases the cumulative incidence over "n" years is simply the incidence rate multiplied by "n". Both cumulative incidence and incidence rate can be determined by following a group of people over time (cohort study). Incidence is frequently used in describing the epidemiology of malignancies and infectious diseases.
In brief, prevalence is what percentage has the disease today, cumulative incidence is what percentage will develop the disease in the next "n" years and incidence rate is how quickly new cases develop.
What measure should be used to describe the epidemiology of gastroesophageal reflux disease (GERD)? In the past, GERD was considered an acute, self-limited condition characterized by esophagitis. In this case, incidence would be a reasonable measure. However, GERD is now recognized as a chronic condition. By the time a person seeks medical attention, he or she may have had symptoms for years and it is very difficult to determine the timing of the onset of the condition. At any point in time, most people with GERD have had longstanding rather than new onset disease . In this situation, prevalence is more informative.
What is GERD?
The term GERD has been used to describe the full spectrum of disease resulting from acid exposure in the esophagus. People with GERD may report heartburn, acid regurgitation or atypical symptoms such as chest pain, dysphagia, globus sensation, cough, wheezing or hoarseness. At endoscopy, patients with GERD may have esophagitis, strictures, Barrett's esophagus, or even esophageal adenocarcinoma; however, up to 50% of patients with GERD will have no endoscopic findings whatsoever [2, 3]. Ambulatory pH monitoring has been considered the gold standard for the diagnosis of GERD. Yet only a minority of people with GERD will undergo this test. Thus, there is no single test or histologic finding that is required for making the diagnosis of GERD.
This broad concept of GERD works well clinically. It does, however, offer considerable challenges epidemiologically. What is the case definition for GERD? Ideally, ambulatory pH monitoring would be performed on a large random sample of the population. Such a study has never been done and even if tried, a significant number of people would likely refuse the test. Community samples have been studied with endoscopy [4, 5]. However, as noted above, endoscopic studies will underestimate the prevalence of GERD by almost 50% [2, 3]. The symptoms heartburn and acid regurgitation have been considered to be specific for the diagnosis of GERD , but these remain inexact measures. Not everyone with excessive acid reflux will have symptoms. Some people with heartburn will have physiological levels of reflux and it is not clear whether they should be considered to have GERD. Since symptoms can be accurately measured in large samples, many studies have measured the prevalence of heartburn as a surrogate marker for the prevalence of GERD.
What is a population-based study?
For many conditions like GERD, not everyone with the condition seeks medical care. Those people who do seek care may not be representative: their symptoms may be more severe, they may worry more about their symptoms, or they might have better access to care. Thus, clinic-based studies may not be generalizeable to the population at large. Population-based studies avoid many of these problems with selection bias. Techniques are available to generate random samples of the population. These samples can be contacted either by phone or letter. Mailed surveys are low risk and allow for a large number of people to be studied at reasonable cost.
When performing a survey, it is important that the questions used be tested in advance to insure their reliability and validity. Such testing often demonstrates how differently people interpret the same question. As an example, some people believe the word heartburn refers to a cardiac condition and thus they may deny this symptom on a questionnaire simply because they do not have any problems with their heart !
What is the prevalence of GERD in a population-based study?
As noted previously, most studies of the prevalence of GERD have measured the prevalence of heartburn. Technically, the prevalence of heartburn is the proportion of people who are experiencing heartburn at the same time. Such information is not useful. These studies have tended to ask people how often they experience heartburn and then have reported the proportion of people who have episodes on a daily, weekly, monthly, or annual basis.
The most widely quoted figures were derived from a study by Nebel, Fornes and Castell which was published in 1976 . They questioned 385 hospital employees and found that 7% experience heartburn daily, 14% note heartburn weekly, and 15% experience it once a month. Thus 36% had heartburn at least monthly. The actual questions were not formally tested in advance and the subjects were not sampled at random. Hospital employees may not be representative of the community at large. Still, these figures have been confirmed in a number of subsequent studies.
Perhaps the second most widely quoted estimate of prevalence comes from a Gallup poll of 1 000 Americans conducted in 1988 . Monthly heartburn was reported by 44% of respondents. The methodology of this poll was never published and thus its validity is not known.
A number of investigators from Europe and North America have performed surveys to determine the prevalence of reflux symptoms [1, 10-16]. Many of these have tested the validity of the survey instrument in advance and most have been mailed to random samples of the population. The results of these surveys are quite similar (Table I). In summary, approximately 20% of the general population experiences heartburn at least once a week and an additional 20% experience heartburn but to a lesser degree.
Like heartburn, acid regurgitation is considered to be specific for the diagnosis of GERD and its prevalence in the community is quite similar. As expected the overlap between these two symptoms is significant. However, if one asks whether the person has experienced any episodes of heartburn or acid regurgitation, the prevalence is as high as 59% .
The symptoms non-cardiac chest pain, dysphagia, globus, dyspepsia, cough, wheezing and hoarseness, have all been reported to be associated with GERD. Each of these is experienced intermittently by over 5% of the population and 55 to 63% report at least one of these symptoms [1, 10-14]. Overall, two-thirds of people with these atypical symptoms also report episodes of heartburn or acid regurgitation [1, 14].
Symptoms of reflux are common in the community, but what proportion of these symptomatic individuals actually have GERD? Andersen and Jensen surveyed 809 randomly selected residents of Copenhagen and then invited subjects with dysphagia, chest pain or nocturnal heartburn and a sample of asymptomatic controls to undergo diagnostic evaluation (endoscopy, manometry and pH testing) . The response rate for investigation was 66%. They found evidence of esophageal disease in 77% of subjects with multiple symptoms, 57% of subjects with isolated symptoms and 20% of asymptomatic subjects which yields an overall prevalence rate of 34.5%. Johnsen et al. offered endoscopy to a population-based sample of people with dyspepsia and matched asymptomatic controls . The definition of dyspepsia included daily heartburn or acid regurgitation. Esophagitis was identified in 12% of subjects with dyspepsia and in 8.1% of the controls.
What then is the prevalence of GERD in a population-based study? An exact figure is not available. Intermittent reflux symptoms are experienced by 55 to 65% of people. Yet, something experienced by a majority of the population cannot be called a disease! One-fifth of the population reports frequent heartburn and one-third can be found to have abnormalities on diagnostic testing. These findings have been consistent across a number of studies. The data to date are sufficient to show that GERD is an extremely common, chronic condition.