RSS Feeds
Use of Prescription Weight Loss Pills among U.S. Adults in 1996–1998
- Laura Kettel Khan, PhD;
- Mary K. Serdula, MD;
- Barbara A. Bowman, PhD; and
- David F. Williamson, PhD
Abstract
Background: Pharmacotherapy is recommended for the treatment of obese persons with a body mass index of 30 kg/m2 or higher or a body mass index of at least 27 kg/m2 plus an obesity-related comorbid condition.
Objective: To estimate the prevalence of use of prescription weight loss pills in the United States in 1996–1998.
Design: 1998 Behavioral Risk Factor Surveillance System, a nationally representative telephone survey.
Setting: United States.
Participants: 139 779 adults 18 years of age and older.
Measurements: Self-reported pill use for 1996–1998, body mass index (current and before pill use), age, sex, and race or ethnicity.
Results: The 2-year prevalence of pill use was 2.5% (95% CI, 2.1% to 2.9%), or 4.6 million U.S. adults. Use was higher in women than in men (4.0% vs. 0.9%, respectively) and highest among Hispanic respondents (3.2%). Of pill users, 25% were not overweight (body mass index < 27 kg/m2 before using pills.
Conclusions: Nearly 5 million U.S. adults used prescription weight loss pills in 1996–1998. However, one quarter of users were not overweight, suggesting that weight loss pills may be inappropriately used, especially among women, white persons, and Hispanic persons.
The prevalence of obesity has increased markedly in the United States (1), as has approval of weight loss drugs by the U.S. Food and Drug Administration (FDA). In 1959, the FDA approved phentermine, the first prescription appetite suppressant used as a single-drug, short-term treatment for obesity. The FDA approved fenfluramine in 1973 for single-drug, short-term use and dexfenfluramine in 1996 as a single-drug, prescription appetite suppressant for longer-term use in obese persons. From 1995 to mid-1997, fenfluramine or dexfenfluramine was widely used in combination with phentermine (“fen-phen”), often for periods longer than a few weeks (2). After numerous reports of cardiac valvulopathy in persons taking fenfluramine or dexfenfluramine (3), the FDA issued a public health advisory on 8 July 1997 (4) that led to the voluntary withdrawal of the drugs from the U.S. market on 15 September 1997.
The U.S. Department of Health and Human Services estimated that between 1995 and 1997, 1.2 to 4.7 million persons in the United States used fenfluramine and dexfenfluramine (3). These figures, however, were not based on population surveys but were indirectly estimated from the number of prescriptions written, with assumptions about the median duration of treatment and mean length of a prescription (5). Thus, the extent to which the population has been exposed to prescription weight loss drugs remains uncertain. In addition, no data were available on the characteristics of persons using the drugs or on whether the drugs were used in accordance with the pharmacotherapy guidelines suggested by the FDA and the 1998 National Heart, Lung, and Blood Institute consensus statement (6, 7).
The purpose of our study was to provide estimates of the 1996–1998 prevalence of the use of prescription weight loss drugs by age, race or ethnicity, and sex by using data from a telephone survey in a sample of U.S. adults.
Methods
We examined data from the 1998 Behavioral Risk Factor Surveillance System (BRFSS), a telephone survey of health practices of adults 18 years of age and older conducted by all U.S. state health departments. Each state, the District of Columbia, and Puerto Rico selected an independent probability sample of noninstitutionalized residents through random-digit dialing; the results were pooled for statistical analyses (Schulman J. Can BRFSS data be pooled for national estimates? Presented at the Sixteenth Annual Behavioral Risk Factor Surveillance System Conference, Minneapolis, Minnesota, 16 May 1999). In 1998, 149 806 persons responded to the BRFSS. (A detailed description of survey methods and quality control indices has been published elsewhere [8].) In 1998, the median upper-bound response rate for individual states (completed interviews divided by completed, refused, and terminated interviews) was 73.4% (range, 45.4% to 95.4%).
Body mass index (BMI) at the time of survey completion (current use) and before use of weight control pills was calculated as the self-reported weight in kilograms divided by height in meters squared and was categorized in five groups: normal weight (<25 kg/m2); preobesity (25 to 29 kg/m2), obesity grade I (30 to 34 kg/m2), obesity grade II (35 to 39 kg/m2), and obesity grade III (≥ 40 kg/m2) (7). According to guidelines of the FDA and National Heart, Lung, and Blood Institute, pharmacotherapy may be considered for persons who are clinically obese (body mass index ≥ 30 kg/m2) or have a body mass index between 27 and 30 kg/m2 plus an obesity-related comorbid condition (for example, hyperlipidemia, hypertension, diabetes, or cardiovascular disease) (6, 7). Because the BRFSS survey did not collect complete information on comorbid conditions, we conservatively defined inappropriate pill use as reported use of weight loss pills in a person with a body mass index less than 27 kg/m2 before pill use.
Of the 149 806 respondents, we excluded those who did not report on pill use (1561 persons), weight before pill use (487 persons), current weight (5229 persons), height (1369 persons), pregnancy status (1780 persons), or sociodemographic characteristics or weight loss behavior (964 persons). We also excluded 88 respondents because the reported weight, height, or body mass index was outside the sex-specific reference values from the Third National Health and Nutrition Examination Survey, 1989–1994 (9). After exclusions, our study sample consisted of 139 779 respondents. To perform statistical analyses, we used SUDAAN software (Research Triangle Institute, Research Triangle Park, North Carolina), which could accommodate our complex sampling design (10). Survey weights were used to produce U.S. population estimates.
Results
In 1996–1998, an estimated 2.5% (95% CI, 2.3% to 2.7%) of U.S. adults used prescription weight loss pills (Table 1). The 2-year prevalence was four times higher among women (4.0%) than men (0.9%) and was one third higher among Hispanic respondents (3.2%) than non-Hispanic white (2.4%) or non-Hispanic black (2.4%) respondents. Among men, pill use was highest in respondents 35 to 64 years of age, and among women, pill use was highest in those aged 25 to 44 years.
Use of prescription weight loss pills increased with current body mass index (Table 1). Among women, the 2-year prevalence of pill use was lowest (1.5% [95% CI, 1.3% to 1.7%]) in those with a current body mass index less than 25 kg/m2 but was substantially higher (17.7% [CI, 15.2% to 20.2%]) in those with a current body mass index of 40 kg/m2 or greater. A similar relationship between pill use and current body mass index was observed among men but with consistently lower values (range, 0.2% to 9.1%). The sex-specific pattern of pill use for current body mass index was similar for all racial and ethnic groups (Table 1). Overall, prescription weight loss pills were used by 3.1% (CI, 2.5% to 3.7%) of obese (body mass index ≥ 30 kg/m2) men and 10.2% (CI, 9.4% to 11.0%) of obese women.
In an analysis restricted to persons who reported use of prescription weight loss pills (n = 3822), we found that 56.1% (CI, 53.5% to 58.7%) of this group were obese before using these pills (Table 2). However, 25.3% (CI, 22.9% to 27.7%) were below the minimum recommended body mass index of 27 kg/m2. Among persons with a body mass index less than 27 kg/m2, pill use was twice as likely among women as men (27.9% [CI, 25.1% to 30.7%] vs. 13.2% [CI, 11.2% to 15.2%]) and almost twice as likely among non-Hispanic white and Hispanic respondents (26.1% [CI, 24.7% to 27.5%] vs. 26.4% [CI, 22.9% to 29.9%]) as non-Hispanic black respondents (15.0% [CI, 12.2% to 17.8%]).
At the time of the survey, 0.5% ([CI, 0.48% to 0.52%]) of respondents were currently using prescription weight loss pills. Paralleling the 2-year prevalence of pill use, the rate for current reported use in women (0.8% [CI, 0.6% to 1.00%]) was fourfold that of men (0.2% [CI, 0.18% to 0.22%]). Except for a peak in current use in October (1.0% [CI, 0.96% to 1.04%]), little seasonal or monthly variation was seen in pill use during 1998 (range, 0.4% to 1.0%).
Discussion
From this population-based study, we estimate that 4.6 million U.S. adults used prescription pills for weight loss in 1996–1998. Women were four times as likely as men to report pill use. Reported use was similar in white and black respondents but was about one third higher among Hispanic respondents than in white and black respondents. Although Hispanic persons have shown a greater inclination toward antibiotic use compared with non-Hispanic persons (11), without more information we cannot speculate on reasons for the higher use of weight loss pills among Hispanic respondents in the current study.
Pill use increased with increasing body mass index and was three times higher among obese women than obese men. Among severely obese respondents (body mass index ≥ 40 kg/m2), the rate of pill use was nearly twice as high in women as men. Reported use of prescription weight loss pills was also substantial among adults who were not obese. In our analysis, we used a conservative definition of appropriate pill use—a body mass index of 27 kg/m2 or greater. We estimate that one eighth of men and more than one fourth of women who used prescription weight loss pills in 1997–1998 had a pre-pill body mass index less than the minimum body mass index of 27 kg/m2 suggested in the FDA and National Heart, Lung, and Blood Institute guidelines. Thus, we estimate that at least 1.2 million adults may have inappropriately used pills in that period. Inappropriate pill use was substantially higher among non-Hispanic white and Hispanic respondents than non-Hispanic black respondents.
Our analysis is limited by the lack of information on the specific prescription drugs used, duration of drug use, dosage, and weight loss during use of the drug. Because the Behavioral Risk Factor Surveillance System did not ask about drugs used concurrently with prescription weight loss pills, no conclusions can be made about the potential for adverse effects or drug interactions. Self-reported weight may also be a concern because many people, especially those who are relatively heavy, may under-report their weight (12, 13).
Although the number of pills prescribed may have decreased immediately before the survey, no data are available for prevalence by month in 1997. However, the prevalence of current use throughout 1998 did not vary by month or season, except in October 1998, which was nearly 1 year after the public advisory on fenfluramine and dexfenfluramine.
Our data indicate that one quarter of U.S. adults who use prescription weight loss pills were not overweight when the pills were prescribed. Other data suggest that more than one third of overweight persons do not meet recommendations for physical activity and dietary practices (12). Furthermore, fewer than half of physicians counsel overweight persons about weight control (14). Taken together, these findings suggest that many overweight persons lack the knowledge or ability to implement lifestyle changes effectively. Patients might be better served if more of them were counseled about long-term weight control and if their physicians adhered more closely to recommendations for initiation of pharmacotherapy. Furthermore, given the increasing prevalence of obesity in the United States, as well as the availability of new prescription weight loss drugs, future investigations of weight-control behaviors and prescription practices are needed.
Article and Author Information
-
Requests for Single Reprints:Laura Kettel Khan, PhD, Division of Nutrition and Physical Activity, Centers for Disease Control and Prevention, 4770 Buford Highway, NE, Mailstop K26, Atlanta, GA 30341; e-mail, LDK7{at}cdc.gov.
-
Current Author Addresses:Drs. Khan, Serdula, and Bowman: Division of Nutrition and Physical Activity, Centers for Disease Control and Prevention, 4770 Buford Highway, NE, Mailstop K26, Atlanta, GA 30341.
-
Dr. Williamson: Division of Diabetes Translation, Centers for Disease Control and Prevention, 4770 Buford Highway, NE, Mailstop K26, Atlanta, GA 30341.
-
Author Contributions:Conception and design: L.K. Khan, M.K. Serdula, B.A. Bowman, D.F. Williamson.
-
Analysis and interpretation of the data: L.K. Khan, M.K. Serdula, B.A. Bowman, D.F. Williamson.
-
Drafting of the article: L.K. Khan, M.K. Serdula, B.A. Bowman, D.F. Williamson.
-
Critical revision of the article for important intellectual content: L.K. Khan, M.K. Serdula, B.A. Bowman, D.F. Williamson.
-
Final approval of the article: L.K. Khan, M.K. Serdula, B.A. Bowman, D.F. Williamson.
-
Provision of study materials or patients: L.K. Khan, M.K. Serdula, B.A. Bowman, D.F. Williamson.
-
Statistical expertise: L.K. Khan.
- Copyright ©2004 by the American College of Physicians
