Abstract
After randomized trials failed to support the use of hormone replacement therapy (HRT) for preventing cardiovascular disease (CVD), HRT use for postmenopausal women declined. Our analysis of 1999–2000 and 2003–2004 National Health and Nutrition Surveys (NHANES) shows that HRT use decreased 19% (from 27.6 to 8.4%; P < .001) among women with CVD versus 3% (from 19.8 to 16.8%; P = .68) among low-risk women, suggesting that most of the drop in HRT use may be among women prescribed HRT as an unproven treatment to prevent CVD.
In the early 1990s, on the basis of findings from observational studies, hormone replacement therapy (HRT) was widely promoted to decrease cardiovascular disease (CVD) in postmenopausal women.1,2 Subsequently, large randomized trials found that HRT did not prevent CVD and that, instead, HRT was associated with an increased risk for stroke and combined estrogen-progestin HRT was associated with a possible increased risk for heart disease.3–6 Following broad dissemination of these results,7,8 HRT use declined markedly.1,9 We report on HRT use in the years 1999–2000 and 2003–2004 by CVD risk, socioeconomic status, and race/ethnicity. We hypothesized that reductions in HRT would be greatest among women with CVD or CVD risk factors.
METHODS
We analyzed data from the 1999–2000 and 2003–2004 US National Health and Nutrition Examination Surveys (NHANES).10–12 Our analysis focused on women aged 45 to 74 years who described themselves as postmenopausal.
Study Variables
In our study population, HRT use and history of CVD (heart disease or stroke) were based on self-report. We considered women as having CVD risk factors if they reported being a smoker; being aged 55 years or older; having diabetes mellitus, hypertension, hyperlipidemia, or family history of CVD; or taking medications for diabetes mellitus, hypertension, or hyperlipidemia. Among study participants for whom physiological data were available, those with a fasting glucose level of 126 mg/dL or higher, mean systolic blood pressure of 140 mm Hg or higher or diastolic blood pressure of 90 mm Hg or higher, and non–high-density lipoprotein cholesterol of 130 mg/dL or higher were considered as having diabetes, hypertension, or hyperlipidemia, respectively. Obesity (body mass index [BMI; defined as weight in kilograms divided by height in meters squared] ≥ 30 kg/m2) was also a CVD risk factor.
Statistical Analysis
The proportion of women using HRT was estimated for the whole sample and stratified by history of CVD or CVD risk factors and by sociodemographic categories. We used t statistics to determine statistical significance for the change in HRT use between the 2 surveys. Sample weights were used in the analysis to account for differential probabilities of selection into sample, nonresponse, and noncoverage (populations not sampled by NHANES) categories. Analyses were performed with SAS version 9.1 (SAS Institute Inc, Cary, North Carolina) and SUDAAN version 9.0 (Research Triangle Institute, Research Triangle Park, North Carolina).
RESULTS
In the period 1999–2000, 26% of US women aged 45 to 74 years (10.9 million) reported using HRT. By the period 2003–2004, HRT use fell to 6.0 million, or 14% of the population. For postmenopausal women aged 45 to 74 years, declines in HRT use were significant for women with or without individual CVD risk factors, with 2 exceptions: women with diabetes (among whom usage was low to begin with) and women without dyslipidemia (Table 1). Among women with CVD and with 2 or more CVD risk factors, there was an absolute decrease in HRT use of 19.2 percentage points (P < .001; 70% relative decrease) and 13.1 percentage points (P < .01; 48% relative decrease), respectively (Figure 1). By contrast, among those with fewer than 2 risk factors, there was no significant decrease (3%; P = .68; 15% relative decrease).
TABLE 1.
Postmenopausal Hormone Replacement Therapy (HRT) Use Among US Women With or Without Cardiovascular Disease (CVD) Risk Factors: US National Health and Nutrition Surveys (NHANES), 1999–2000 and 2003–2004
| 1999–2000 |
2003–2004 |
Change in No. Using HRT, % Points | ||||
| No.a | Using HRT, % (95% CI) | No.a | Using HRT,% (95% CI) | Pb | ||
| Presence of CVD or CVD Risk Factors | ||||||
| History of CVD | ||||||
| Yes | 118 | 27.6 (18.0, 37.2) | 151 | 8.4 (2.4, 14.4) | 19.2 | <.01 |
| No | 955 | 25.4 (20.3, 30.5) | 906 | 14.6 (11.1, 18.1) | 10.8 | <.01 |
| Age, y | ||||||
| < 55 | 670 | 24.6 (19.2, 30.0) | 650 | 11.5 (7.8, 15.2) | 13.1 | <.001 |
| ≥ 55 | 403 | 26.5 (22.2, 30.8) | 407 | 16.4 (12.8, 20.0) | 10.1 | <.001 |
| Hypertension | ||||||
| Yes | 595 | 26.4 (20.4, 32.4) | 620 | 15.6 (11.3, 19.9) | 10.8 | <.01 |
| No | 478 | 24.9 (19.2, 30.6) | 437 | 12.2 (7.0, 17.4) | 12.7 | <.01 |
| Dyslipidemia | ||||||
| Yes | 801 | 28.1 (23.3, 23.9) | 811 | 14.1 (11.4, 16.8) | 14.0 | <.001 |
| No | 272 | 17.6 (11.1, 24.1) | 246 | 13.3 (6.4, 20.2) | 4.3 | .38 |
| Diabetes | ||||||
| Yes | 198 | 12.4 (6.8, 18.0) | 209 | 9.9 (5.2, 14.6) | 2.5 | .5 |
| No | 875 | 27.5 (22.8, 32.2) | 848 | 14.6 (11.2, 18.0) | 12.9 | <.001 |
| Obesity | ||||||
| Yes | 423 | 23.2 (17.4, 29.0) | 415 | 10.9 (7.6, 14.2) | 12.3 | <.001 |
| No | 650 | 26.9 (21.4, 32.4) | 642 | 15.6 (11.0, 20.2) | 11.3 | <.01 |
| Current smoking | ||||||
| Yes | 159 | 18.6 (10.3, 26.9) | 176 | 5.8 (1.1, 10.5) | 12.8 | .01 |
| No | 914 | 27.2 (22.0, 32.4) | 881 | 15.7 (11.8, 19.6) | 11.5 | .001 |
| Family history of coronary heart disease | ||||||
| Yes | 160 | 30.7 (22.7, 38.7) | 173 | 14.3 (8.1, 20.5) | 16.4 | <.01 |
| No | 913 | 24.4 (20.2, 28.6) | 884 | 13.9 (10.5, 17.3) | 10.5 | <.001 |
| Sociodemographic Factors | ||||||
| Education Level | ||||||
| < High school | 482 | 15.9 (10.3, 21.5) | 317 | 7.3 (4.7, 9.9) | 8.6 | .01 |
| High school | 246 | 26.1 (19.3, 32.9) | 297 | 13 (7.8, 18.2) | 13.1 | <.01 |
| College and beyond | 341 | 31.7 (25.6, 37.8) | 441 | 16.6 (13.0, 20.2) | 15.1 | <.001 |
| Annual household income, % of federal poverty levelc | ||||||
| ≤ 100 | 362 | 13.8 (9.4, 18.2) | 229 | 6.8 (2.5, 11.1) | 7.0 | .03 |
| 101–299 | 370 | 20.5 (13.5, 27.5) | 427 | 10.4 (6.7, 14.1) | 10.1 | .02 |
| ≥ 300 | 341 | 36.1 (30.6, 41.8) | 401 | 18.6 (13.7, 23.5) | 17.5 | <.001 |
| Race/ethnicity | ||||||
| Non-Hispanic Black | 211 | 10.8 (7.1, 14.6) | 220 | 12.2 (8.1, 16.3) | 1.4 | .60 |
| Hispanic | 310 | 16.8 (16.8, 20.6) | 215 | 10 (5.4, 14.6) | 6.8 | .56 |
| Non-Hispanic White | 446 | 31.7 (27.8, 35.5) | 555 | 14.4 (10.6, 18.2) | 17.3 | <.001 |
| Other | 106 | 7.0 (3.3, 10.7) | 67 | 11.7 (0.8, 22.6) | 4.7 | .06 |
| Total | 1073 | 27.6 (21.5, 29.7) | 1057 | 13.9 (10.6, 17.2) | 11.7 | <.001 |
Note. CI = confidence interval.
Number of participants in the original NHANES sample for that category.
P is for absolute change in proportion between surveys, obtained through use of a t statistic.
Census bureau determination of poverty threshold for the most recent survey year (1999 and 2004).
FIGURE 1.
Change in hormone replacement therapy (HRT) use by number of cardiovascular disease (CVD) risk factors: US National Health and Nutrition Examination Surveys (NHANES), 1999–2000 and 2003–2004.
Note. P is for absolute change in proportion between surveys, determined by use of a t statistic.
HRT use declined significantly among non-Hispanic White women, the racial/ethnic group with the highest rate of HRT use. There was no drop in HRT use among non-Hispanic Black women and a nonsignificant drop among Hispanic women, both of whom started off with much lower rates of HRT use. HRT use declined within all education and income groups, but the largest absolute decline was among women of higher education and income (Table 1).
DISCUSSION
As hypothesized, we found that from 1999–2004, HRT use for postmenopausal women decreased significantly among those at highest CVD risk, whereas women at lowest CVD risk had no significant change in HRT use. Our results add to past findings by suggesting that the largest drop in HRT use may be among women who were most likely to have been prescribed HRT for prevention of CVD events. Although physicians and patients are often faulted for not rapidly adopting effective new treatments, in this case, dissemination of HRT clinical trial results in the scientific and lay press7,9 likely contributed to rapidly diminished use of this therapy. Nonetheless, although HRT use declined significantly among women with CVD risk, in 2004, 8.4% of women with CVD and 14% of those with 2 or more CVD risk factors were still taking HRT. It is not clear how many of these women were taking HRT for reasons unrelated to CVD (such as severe perimenopausal symptoms), nor is the duration of treatment.
Studies have documented persistent racial differences in the rates of new medical technologies such as prescription drugs and medical procedures.13,14 Our finding that, compared with non-Hispanic White women and women of higher socioeconomic status, minority women and those of low socioeconomic status had much lower baseline rates of HRT use and smaller or no reductions in use over the study period suggests that, at least in this case, an unequal health care system helped the most vulnerable to avoid receiving this ineffective and potentially dangerous therapy.
This study's strength was that it used data from a nationally representative sample to identify a large number of postmenopausal women using HRT in 2 time periods. A limitation was that it was based on self-reported data, potentially resulting in misclassification of HRT and CVD risk factor categories, which would dilute the power to discern significant changes. In conclusion, we have shown that steep declines in HRT use were found among women with CVD or with 2 or more CVD risk factors, but no significant change was seen among women at low cardiovascular risk. Although without chart review we cannot be certain of the reason women were prescribed HRT, these findings suggest that these women may have received HRT for the prevention of CVD, a therapy that was never proven for this indication and ultimately found to be ineffective.
Human Participant Protection
All data came from publicly available data sets that contain no individually identifiable private information.
References
- 1.Hersh AL, Stefanick ML, Stafford RS. National use of postmenopausal hormone therapy: annual trends and response to recent evidence. JAMA 2004;291(1):47–53 [DOI] [PubMed] [Google Scholar]
- 2.Majumdar SR, Almasi EA, Stafford RS. Promotion and prescribing of hormone therapy after report of harm by the Women's Health Initiative. JAMA 2004;292(16):1983–1988 [DOI] [PubMed] [Google Scholar]
- 3.Anderson GL, Limacher M, Assaf AR, et al. Effects of conjugated equine estrogen in postmenopausal women with hysterectomy: the Women's Health Initiative randomized controlled trial. JAMA 2004;291(14):1701–1712 [DOI] [PubMed] [Google Scholar]
- 4.Hulley S, Grady D, Bush T, et al. Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. Heart and Estrogen/progestin Replacement Study (HERS) Research Group [see comment]. JAMA 1998;280(7):605–613 [DOI] [PubMed] [Google Scholar]
- 5.Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women's Health Initiative randomized controlled trial [see comment]. JAMA 2002;288(3):321–333 [DOI] [PubMed] [Google Scholar]
- 6.Wassertheil-Smoller S, Hendrix SL, Limacher M, et al. Effect of estrogen plus progestin on stroke in postmenopausal women: the Women's Health Initiative: a randomized trial. JAMA 2003;289(20):2673–2684 [DOI] [PubMed] [Google Scholar]
- 7.Haas JS, Geller B, Miglioretti DL, et al. Changes in newspaper coverage about hormone therapy with the release of new medical evidence. J Gen Intern Med 2006;21(4):304–309 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Haas JS, Miglioretti DL, Geller B, et al. Average household exposure to newspaper coverage about the harmful effects of hormone therapy and population-based declines in hormone therapy use. J Gen Intern Med 2007;22(1):68–73 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Haas JS, Kaplan CP, Gerstenberger EP, Kerlikowske K. Changes in the use of postmenopausal hormone therapy after the publication of clinical trial results [see comment]. Ann Intern Med 2004;140(3):184–188 [DOI] [PubMed] [Google Scholar]
- 10.National Health and Nutrition Examination Survey Questionnaire (or Examination Protocol, or Laboratory Protocol) Hyattsville, MD: Centers for Disease Control and Prevention, National Center for Health Statistics; 2003–2004 [Google Scholar]
- 11.National Health and Nutrition Examination Survey Questionnaire (or Examination Protocol, or Laboratory Protocol) Hyattsville, MD: Centers for Disease Control and Prevention, National Center for Health Statistics; 2001–2002 [Google Scholar]
- 12.National Health and Nutrition Examination Survey Questionnaire (or Examination Protocol, or Laboratory Protocol) Hyattsville, MD: Centers for Disease Control and Prevention, National Center for Health Statistics; 1999–2000 [Google Scholar]
- 13.National Healthcare Disparities Report Rockville, MD: Agency for Healthcare Research and Quality; 2006 [Google Scholar]
- 14.Smedley BSA, Nelson A. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care Washington, DC: National Academy Press; 2003 [PubMed] [Google Scholar]