Heartburn Risk Factors, Knowledge, and Prevention Strategies:  A Population-Based Survey of Individuals With Heartburn FREE

TWENTY-FIVE MILLION adults experience heartburn daily, and more than one third of adults experience heartburn at least once a month in the United States.^1- 4 Heartburn, a burning sensation behind the breast bone, is the most common symptom of gastroesophageal reflux disease.⁴ Although not life threatening, gastroesophageal reflux can cause complications, including esophagitis, ulceration, and stricture formation. Moreover, gastroesophageal reflux is a risk factor for Barrett esophagus, a premalignant condition of the esophagus.^5- 6 The high prevalence and associated long-term discomfort of heartburn affects quality of life and is costly to society. In 1996, more than $1 billion was spent on nonprescription heartburn remedies, including hydrogen₂ receptor antagonists and antacids (AC Nielsen marketing company, unpublished data, September 1996).

Heartburn occurs when the esophagus is exposed to gastric acid. Ordinarily, the lower esophageal sphincter, which separates the esophagus from the stomach, prevents acid from entering the esophagus. At inappropriate times, transient lower esophageal sphincter relaxation occurs, resulting in gastroesophageal reflux and heartburn.⁷ Heartburn results when sensory nerve endings are stimulated by acid reflux or esophageal distention.⁸ Consumption of certain foods and beverages has been reported to induce heartburn.^6,9 Body position, including recumbency, bending over, and the postprandial state, has also been associated with heartburn.^4,6 Lifestyle factors, such as stress, cigarette smoking, alcohol consumption, obesity, medication use, and certain exercises, may also precipitate heartburn.^6,10- 13 Specific factors may be responsible for heartburn, but it appears as though these factors are unique for each individual. For instance, some patients report that coffee consumption causes their heartburn, while other patients are not affected by the consumption of this beverage. Furthermore, it is unclear whether individuals can identify the factors that cause their heartburn symptoms. Recommendations are generally made by physicians for dietary modification, weight loss, lifestyle behaviors, and body position, even though they may not know the specific cause of heartburn in an individual patient.^5- 6,14

Characteristics of individuals with heartburn, causes of heartburn, patient knowledge, and prevention strategies have not been described previously in the general population. Recently, there has been an increase in the number of nonprescription medications available to prevent heartburn. To target individuals for prevention programs that include dietary modification, lifestyle and behavior modification, and treatment remedies, individuals and the associated causes of their heartburn must first be identified. The assessment of an individual's knowledge is crucial to successfully implement prevention strategies. We conducted a population-based survey to describe the sociodemographic characteristics, frequency of symptoms, self-reported causes of the condition, and knowledge of heartburn risk factors in individuals with heartburn. We also collected information on the different approaches used by individuals with heartburn to prevent the onset of symptoms.

A cross-sectional survey of 2000 individuals with heartburn was conducted. The study population included adults aged 18 years or older in the United States who reported having heartburn in the past 6 months. All eligible study participants were required to have telephone service. A survey instrument was developed and pilot tested and included a screening question to identify individuals with heartburn as well as questions related to sociodemographic characteristics, frequency of symptoms, self-reported causes of heartburn, knowledge of factors contributing to heartburn, and prevention strategies. The survey fieldwork was done by the MEDSTAT Group, Ann Arbor, Mich. Random digit dialing was used to select households stratified by state to ensure that the study population was geographically representative. Persons aged 18 years or older who answered the telephone were identified as individuals with heartburn if they had experienced heartburn during the past 6 months. Heartburn was defined as "a burning sensation in the chest behind the breast bone, which may or may not be associated with the sensation of food coming back into the mouth or an acid or bitter taste." All eligible individuals were recruited for the study. Trained interviewers administered the standardized survey using computer-assisted telephone interviewing. Recruitment was ongoing until 2000 individuals successfully completed the entire survey.

Random calls (N=10,559) were made to complete 2000 surveys. About one fourth (or 2377) of the telephone numbers were ineligible because they were the wrong number, a cell or business number, or a computer or fax number. Seven hundred twenty-four respondents were ineligible for the study: 467 did not speak English or were deaf, and 257 did not have heartburn. Eligibility was unknown for 5130 potential survey respondents who either could not be reached by telephone (n=1160) or refused to answer the screening question (n=3970). Of the 2328 eligible subjects, 328 (14.1%) refused to complete the survey.

Descriptive statistics, including means, SDs, and percentages, were calculated to characterize the survey population for sociodemographic variables, frequency of heartburn symptoms, causes of heartburn, and knowledge of risk factors. Logistic regression modeling was used to identify predictors of frequent heartburn and informing a physician about heartburn symptoms. A commercially available statistical program¹⁵ was used for all analyses. Odds ratios (ORs), confidence intervals (CIs), χ² test statistics, and P values (2 tailed) are presented.

Sociodemographic characteristics of the 2000 individuals with heartburn representing all 50 states are presented in Table 1. The age range of the survey respondents was 18 to 91 years. Women made up 59% of the study population. Most survey respondents were white (87%) and married or living with someone (68%). About half of the men and women were college educated, and more than 70% of survey respondents lived in small cities or rural towns.

The self-reported frequency of heartburn was as follows: daily (21.5%), a few times a week (24.4%), once a week (11.3%), a few times a month (16.5%), once a month (11.5%), and less than once a month (13.8%). We classified individuals with heartburn as persons with frequent (daily or a few times a week) vs infrequent (once a week, a few times a month, once a month, or less than once a month) heartburn. Forty-six percent of the survey respondents had frequent heartburn. Approximately 38% of the survey respondents reported that the frequency of their heartburn was increasing with age, while 18% reported a decrease of heartburn frequency with age. The trend of increasing heartburn with age is presented in Figure 1.

Table 2 presents the prevalence of heartburn symptoms by sociodemographic characteristics. Women were somewhat more likely to report daily heartburn than men. Women reported the onset of heartburn about 5 years later than men (mean ± SD age, 34.7 ± 16.6 vs 29.9 ± 14.2 years; P<.001). Survey respondents who were not college educated were more likely to report frequent symptoms compared with individuals with heartburn who were college educated (OR, 1.45; 95% CI, 1.21-1.73) (P=.001).

We asked survey respondents to report when they experienced heartburn using a checklist of potential trigger situations (Table 3). Respondents were allowed to respond to all questions that applied. Heartburn was experienced after or during a broad range of activities and events. Men and women reported heavy meals and high-fat foods, acidic foods, and lying down as causes of heartburn; 23% to 52% of the respondents attributed their heartburn to these situations. Women reported heartburn caused by a hectic day at home, stressful family situations, eating a heavy meal, and lying down more often than men. Women were 70% more likely than men to report stressful family situations (OR, 1.70; 95% CI, 1.40-2.07) and 55% more likely than men to report a hectic day at home (OR, 1.55; 95% CI, 1.25-1.92) as causes of heartburn. In contrast, men were 24% more likely than women to report a week of long work hours (OR, 1.24; 95% CI, 0.99-1.55) and 50% more likely than women to report business travel (OR, 1.50; 95% CI, 1.08-2.07) as causes of heartburn. We asked specifically about the frequency of heartburn after eating, during sedentary activities, and during sports and exercise. Men and women reported similar frequencies of exercise-related heartburn, although about 75% of those surveyed did not experience heartburn associated with exercise. As was expected, specific foods and beverages were reported to cause heartburn (Table 4). Men were 64% more likely than women to report drinking alcoholic beverages as a cause of heartburn (OR, 1.64; 95% CI, 1.29-2.09). Women reported eating a higher percentage of heartburn-promoting foods, including fatty foods, chocolate, peppermint, citrus fruits or juices, and tomato products. No other predominant male-female differences were observed.

In preliminary analyses, we explored the relationship between age groups, frequent vs infrequent heartburn status, and the factors listed in Table 3 and Table 4. Statistical testing was not accomplished for age comparisons because of the small numbers available after stratification by age groups, but was done for analyses comparing frequent vs infrequent heartburn status. A hectic day at work, business travel, a week of long work hours, and a week with numerous deadlines were reported less by the older age groups as causes of heartburn. Heartburn before, during, and immediately after exercise or sports was reported more frequently as a cause of heartburn by the younger age groups. Furthermore, smoking and drinking alcoholic beverages were reported less by the older age groups as causes of heartburn. No other important trends were observed. Consumption of fast foods, caffeinated beverages, and alcoholic beverages were reported less by the older age groups as causes of heartburn. Those with frequent heartburn reported a higher frequency for all the causes of heartburn listed in Table 3 and Table 4. Nearly all the comparisons reached statistical significance (P≤.05).

Obesity was associated with an increased frequency of self-reported heartburn. We calculated body mass index (calculated as weight in kilograms divided by height in meters squared) and categorized those with heartburn into quartiles. Individuals in the highest quartiles of body mass index were more likely to report daily heartburn compared with those in the lowest quartile (referent) of body mass index; ORs ranged from 1.30 to 2.19. A dose-response relationship was apparent, with an overall trend test that was statistically significant (P=.02).

Most survey respondents were unable to identify known causes of heartburn. Table 5 presents the respondents' knowledge of heartburn risk factors. To obtain valid responses to the questions on heartburn knowledge, we included questions on factors that are not known to be associated with heartburn. Although the survey population was composed of 2000 individuals with heartburn, most of them were unaware of the risk factors for heartburn. The findings revealed that men and women differ in their knowledge about risk factors for heartburn, which may reflect sex-dependent differences in lifestyle. For example, men were more likely than women to identify the following risk factors for heartburn: work schedule or conditions, alcohol consumption, smoking, and exercise. Conversely, women were more likely than men to identify the following factors as risk factors for heartburn: age, citrus juice consumption, weight, stress, and wearing tight clothes around the stomach.

Although all of the respondents surveyed had heartburn, with almost 50% having weekly symptoms, only 55% of men and 66% of women informed their physician. Survey respondents with frequent heartburn were 4.14 times more likely (95% CI, 3.41-5.02) to have told their physician that they had symptoms compared with survey respondents with infrequent heartburn (P<.001). Of the 916 respondents with frequent heartburn, 716 (78.2%) informed their physician vs 199 (21.7%) who did not inform their physician; of the 1061 respondents with infrequent heartburn, 492 (46.4%) informed their physician vs 566 (53.3%) who did not inform their physician. Women were more likely to have told their physician about their heartburn symptoms compared with men. Regression modeling identified increasing age, female sex, higher level of education, and frequent vs infrequent heartburn as significant predictors of whether patients told their physician about their heartburn symptoms (Table 6).

Approaches used to prevent and treat heartburn are presented in Table 7 and Table 8. Forty-five percent of the survey respondents used a nonprescription medication to prevent the onset of heartburn symptoms. Other approaches to heartburn prevention included taking prescription medication and avoiding spicy foods. Women were more likely to take a prescription medication and avoid fatty foods and citrus juices, while men were more likely to do nothing to prevent heartburn. More than 75% of male and female respondents reported taking a nonprescription medication to treat heartburn. Although certain foods and lifestyles have been identified as potential causes of heartburn, most individuals did not report that they modified these behaviors to avoid or prevent heartburn. Table 9 presents the results of multivariate analyses. Increasing age, higher body mass index, and reduced level of education were statistically significant predictors of frequent vs infrequent heartburn in this study population.

This is the first population-based study to characterize individuals with heartburn for cause, knowledge of risk factors, and prevention strategies. Lifestyle and work habits, in addition to certain food and beverage consumption, were associated with heartburn. Most survey respondents were unaware of the risk factors for heartburn, and sex-dependent differences in knowledge were apparent. Women reported the onset of heartburn symptoms about 5 years later than men, which is likely to reflect differences in lifestyle. Obesity was associated with an increased frequency of self-reported heartburn, and a dose-response relationship was evident. Many survey respondents did not inform their physician about their symptoms, and nonprescription medication was the approach used most often to prevent or treat heartburn. College educated individuals were less likely to report heartburn symptoms compared with those without a college education. This may reflect differences in access to care and knowledge about prevention and treatment strategies. Thus, those respondents with less than a college education were about 20% less likely to use a prescription or nonprescription medication to prevent heartburn. Increasing age was an important predictor of frequent heartburn. It is unclear, however, whether heartburn was associated with aging because of a change in esophageal physiological characteristics, lifestyle, or increased use of medications that predispose to gastroesophageal reflux.

This study used a telephone survey and, thus, excluded individuals who had no telephone service or who did not answer the telephone. Fifty-nine percent of the respondents were women, which probably reflects the greater likelihood of women to answer the telephone. Although we used random digit dialing (conducted during various times of the day), stratified by state, to ensure geographic representativeness, our study population was predominantly white individuals (87%) who lived in small cities or rural towns (72%), which may limit the generalizability of the findings. Another limitation of the study was our inability to collect information on use of medications or medical conditions, such as pregnancy, diabetes mellitus, and hiatal hernia, that predispose to heartburn.⁶

Physicians often recommend lifestyle changes in addition to medications to prevent heartburn. Decreasing consumption of spicy foods and caffeinated beverages, for example, has been suggested to improve symptoms. Weight loss is recommended for obese patients. Avoiding the supine position after eating may decrease gastroesophageal reflux and thereby prevent heartburn. The results of this study support these recommendations and show a relationship between certain lifestyle factors and heartburn. Future studies are needed, however, to confirm that behavior modification is useful in preventing and treating heartburn.

The findings of this study are important for the development of a heartburn prevention program, which incorporates lifestyle modification. It is apparent that most individuals with heartburn have little insight into the factors that cause heartburn. Lifestyle modification strategies necessarily must first begin with education so that dietary modification, behavior modification, and treatment remedies can be implemented. This is especially true for individuals who are reluctant to use medications. The results of this study need to be considered in the context of medications that are used to prevent or treat heartburn. Whether the need for medical intervention can be decreased by simple changes in lifestyle needs to be tested in the general population. Alternatively, many individuals are unlikely to be willing to modify their lifestyles because of easy access to safe and effective therapy.

Reprints: Susan A. Oliveria, ScD, MPH, Memorial Sloan-Kettering Cancer Center, 1275 York Ave, Box 99, New York, NY 10021 (e-mail: oliveri1@mskcc.org).

Accepted for publication November 2, 1998.

This study was supported by Whitehall-Robins Healthcare, Madison, NJ.

We thank Deborah Plutzer and Bridgette Bayliss for their assistance with manuscript preparation.