Katharine H. Hendrix, PhD, Medical University of South Carolina, Department of Medicine, Hypertension Section, 96 Jonathan Lucas Street, 826 CSB, Charleston, SC 29425 E-mail:email@example.com
Dyslipidemic, hypertensive patients (N=48,863) were stratified by gender, age, and angina (n=2502) vs. nonangina (n=46,358) status. Comparing 95% confidence intervals yielded significant differences in treatment and cardiovascular risk factor control between subgroups. More men than women bad low-density lipoprotein cholesterol (LDL-C) <100 mg/dL (angina, 43.94-43.96 vs. 34.42-34.50; nonangina, 32.43-32.43 vs. 17.25-17.25) and 100-129 mg/dL (angina, 32.12-32.14 vs. 35.10-35.18; nonangina, 53.86-53.86 vs. 32.44-32.44). More women than men had LDL-C >130 mg/dL (angina, 27.68-27.72 vs. 23.91-23.93; nonangina, 38.70-38.70 vs. 35.38-35.39). Women were less likely than men to receive statins (angina, 69.95-69.99 vs. 82.11-82.13; nonangina, 59.80-59.80 vs. 63.72-63.72), any antilipidemic medication at all (angina, 25.93-25.97 vs. 13.48-13.48; nonangina, 36.73-36.73 vs. 30.73-30.73), or to have current cholesterol measurements (angina, 56.82-56.88 vs. 34.54-34.56; nonangina, 45.77-45.77 vs. 39.75-39.75). Primary care providers treat high-risk patients relatively aggressively; however, opportunities to forestall cardiovascular disease may be missed in hypertensive, dyslipidemic women whose LDL-C is often not measured and controlled.
Coronary heart disease (CHD) is the single leading cause of death in both men and women in the United States, accounting for more than one out of every five deaths in 2001.1 Hypertension is one of the most significant risk factors for development of CHD, affecting over half of all people who have suffered a first heart attack, and preceding development of congestive heart failure in 91% of cases.2,3 While the health problems presented by CHD and hypertension have long been recognized in men and older patients, the magnitude of the burden of these diseases in women and younger patients has been largely overlooked until recent decades.4–6 While women do have some health advantages over men regarding the development of heart disease, these advantages are demonstrably not universal—and once CHD is manifest, women lose prognostic superiority over men.7,8
Elevated low-density lipoprotein cholesterol (LDL-C) levels (dyslipidemia) are well documented as a major contributor to the development of atherosclerosis and subsequent coronary disease.9 Hypertensive patients with dyslipidemia are at increased risk for formation of atherosclerotic plaque and all related cardiovascular (CV) events and are classified as high-risk patients.6
Angina and preinfarction angina or acute coronary syndrome are the most common manifestations of CHD, affecting an estimated 6.8 million American adults.1 Patients with angina, compared with an age-matched cohort, have been found to be at a three-fold increased risk of myocardial infarction or sudden cardiac death.10,11 Research indicates that angina is more common in women than men (4.3% vs. 2.7% (1,4,12 but that chest pain complaints among women may not be fully investigated by primary care physicians, leaving angina undiagnosed for longer periods of time than in men.4,13 Furthermore, case identification in the primary care setting can be difficult as, on examination, many patients presenting with chest pain do not meet diagnostic criteria for ischemic heart disease.14 Failure to adequately diagnose angina among women is particularly serious, since chest pain more often precedes myocardial infarction in women than men, and fatality rates for acute myocardial infarction are greater in women than men (32% vs. 27%).4,13 Consistent with these data, another large study found that when controlling for age, race, and other risk factors, dyslipidemia and hypertension are associated with similar increased risk of death for both women and men.15
Treatment and risk factor control rates for hypertensive, dyslipidemic patients in primary care have been found to differ by patient gender and age.5,16 Studies have also found demographic differences in the diagnosis and treatment of patients with CHD and angina.13,17–21 One analysis found gender to be an independent variable influencing primary care physician management of patients presenting for chest pain—with women receiving the fewest treatments and interventions.21
The undertreatment of dyslipidemia in patients with hypertension and angina is widely recognized as an important barrier to improving patient outcomes and preventing future coronary events.22–24 The importance of aggressively treating high cholesterol in young and middle-aged patients (especially those with multiple CV risk factors) is also clear.6,16,25 Atherogenesis is a progressive disease, and cholesterol levels during young adulthood have been found to strongly predict the risk of CHD and related mortality in later years of life.5 Furthermore, it has been found that the total cholesterol level in youth determines the age at which critical levels of atherosclerosis will be reached.5 The ability to reduce mortality by controlling the combined CV risk factors of hypertension and dyslipidemia in younger patients is underscored by a large study of patients aged 18-55, which found that both genders experience substantially increased mortality associated with these uncontrolled risk factors.6
The prevalence and treatment of hypertensive patients with dyslipidemia and angina and the control of CV risk factors among these high-risk patients in primary care is not well researched. The purpose of the current study is to address this gap in the literature and to examine the potential impact of patient gender and age on CV risk factor treatment and control.
We analyzed a sample (N=48,863) of high-risk (i.e., dyslipidemic) hypertensive patients from the Hypertension Initiative primary care database. The Hypertension Initiative database contains information on 104,031 hypertensive patients receiving health care at approximately 90 primary care practices from more than 300 providers in the southeastern United States.26,27 Records were entered into the database via electronic medical record download or, in sites without electronic medical records, via manual data entry of report cards completed at hypertensive patients' appointments (approximately 20% of records). No records from paper card entries were included in this analysis, as diagnosis data for angina were not available from this source.
All data monitoring and review procedures were approved by the Office for Research Protection and Integrity at the Medical University of South Carolina to ensure that appropriate patient confidentiality safeguards were in place and that the study complied with the Health Insurance Portability and Accountability Act.
The sample of hypertensive patients with documented diagnoses of angina pectoris and/or preinfarction angina and/or acute coronary syndrome was identified by querying the database for records with International Classification of Diseases, Ninth Revision (ICD-9) codes 413, 413.9, 411.1, 413.1, and 413.0 and capturing records for analysis.
Patients with hypertension were defined as those with a diagnosis of hypertension documented in the medical record, excluding gestational hypertension. Patient diagnosis was searched using methods similar to those described above, i.e., searching ICD-9 codes.
Dyslipidemic patients were defined as those with a diagnosis of hyperlipidemia and/or dyslipidemia and/or hypercholesterolemia documented in the primary care medical record. Patient diagnoses were searched by ICD-9 codes and captured for analysis.
Treatment modalities among dyslipidemic, hypertensive patients with and without an angina diagnosis were compared by searching patient records for prescriptions in 12 medication classes of interest: angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor blockers (ARBs), β blockers, a/p blockers, dihydropyridine calcium channel blockers (CCB-Ds), nondihydropyridine calcium channel blockers (CCB-NDs), diuretics, statins, other antilipidemics (ALs), vasodilators, nitrates, and aspirin. Medication text field lists were searched for the year before the last visit to establish whether or not a patient was currently prescribed a medication treatment.
Laboratory and test measurements were also captured from the medical record by searching code and text fields for blood pressure (BP) and LDL-C measurements. BP used for analysis was an average of the last two readings. All patients had at least two BP readings within the last year. LDL-C measurements were taken from the year before the last visit. A patient without an LDL-C recorded within a year of his or her last visit was counted as having no current LDL-C measurement in accordance with national treatment guidelines. Because primary care providers are expected to follow national treatment guidelines and because at the time of this study these guidelines focused on LDL-C control, high-density lipoprotein cholesterol (HDL-C) measurements were not captured for analysis. This does not imply that HDL-C is not an important component in treatment decisions, especially among women; however, women with hypertension are known to have lower HDL-C than their normotensive counterparts.28 Therefore, in this population of women at CV risk (i.e., with documented hypertension and dyslipidemia) and high CV risk (i.e., with documented hypertension, dyslipidemia, and angina), LDL-C control may be of greater clinical significance than among women at lower CV risk.
The analytic dataset of dyslipidemic hypertensive patients in primary care (N=48,863) was extracted from the overall database and exported to SPSS (SPSS Inc., Chicago, IL) or STATA (StataCorp LP, College Station, TX) for analysis. The analytic purpose was to describe BP and LDL-C control and patterns of medication prescriptions in hypertensive, dyslipidemic patients with and without diagnosed angina. Subgroups were stratified and compared by diagnosis (angina vs. nonangina), gender (men vs. women), and age (younger than 60 years vs. 60 years or older) to illuminate differences in diagnosis rates, treatment patterns, and CV risk factor control.
Analyses compared proportions among the sub-samples of interest using 95% confidence interval (CI) as the basis of comparisons due to the large sample and to accommodate differences in subgroup sizes. Significant differences (p<0.05) were interpreted as the failure of CIs to overlap, i.e., to include the mean of one group within the mean range of the other. Differences in key study variables were described as CI vs. CI, mean ± SD, or n (%) for all patients with and without angina, and for subgroups by gender, diagnoses, and age group.
Table I sets forth the results of the data analyses described above. Of all 104,031 hypertensive patients being seen at primary care sites participating in the Hypertension Initiative, 48,863 (47%) were also diagnosed with dyslipidemia. Five percent (2502) of these dyslipidemic, hypertensive patients had also been diagnosed with angina. More hypertensive men than women were diagnosed with both dyslipidemia and angina. Mean age did not differ substantially between hypertensive, dyslipidemic men and women with or without angina, either within or between diagnostic groups; however, proportions of hypertensive, dyslipidemic men and women diagnosed with angina were substantially different. It should be emphasized that all diagnoses for hypertension, dyslipidemia, and angina were drawn directly from the primary care provider's own diagnostic codes as documented in the patient's medical record. No diagnoses were assigned by the research team. As noted, the diagnosis of dyslipidemia only considered the total cholesterol and LDL-C levels, and not HDL-C levels. Medication prescription rates differed among patients by gender, diagnosis, and age group. As would be expected, older patients (60 years and older) with hypertension and comorbid dyslipidemia received consistently more prescriptions than similarly diagnosed patients under 60 years of age in both the angina and nonangina groups. The mean number of medications (± SD) among men was higher than among women in both diagnosis groups, with the angina group, overall, receiving substantially more medications than the nonangina group (angina men, 4.98±2.20; angina women, 4.43±2.97 vs. nonangina men, 3.15±1.96; nonangina women, 3.14±2.14). An age of 60 years or older was also associated with a higher mean number of medications in both diagnosis groups.
Table I. Demographics and Medications by Age, Gender, and Presence of Angina (or Acute Coronary Syndrome) Diagnosis
Age <60 yr
Age <60 yr
Total (n [%])
Men (M) (n [%])
Age (yr) (mean ± SD)
Women (W) (n [%])
Age (yr) (mean ± SD)
(mean ± SD)
Angiotensin receptor blocker
ACE=angiotensin-converting enzyme; CCB=calcium channel blocker; D=dihydropyridine; *percentage of patients receiving each, shown as 95% confidence intervals (CIs); **significant differences between CIs, p<0.05
Among women both with and without angina, the difference in prescription rates between older and younger patients was significant for all but one of the 12 medication classes. Older patients received substantially more medications than their younger, same-gender counterparts in the same diagnosis group. This pattern was also found among men both with and without angina, but the trend of increasing medications with increasing age was not as consistent as was observed among women.
Differences in prescription rates were found when subpopulations were compared by gender (Table I). Among all patients in both diagnosis groups, men received more prescriptions than women for eight of the 12 medication classes of interest—all except ARBs, CCB-Ds, CCB-NDs, and diuretics. Among hypertensive, dyslipidemic patients with angina, men received significantly more prescriptions than women for seven of the 12 medications of interest: ACE inhibitors (53.61-53.67 vs. 63.86-63.88), β blockers (55.07-55.13 vs. 67.52-67.54), statins (69.95-69.99 vs. 82.11-82.13), all other ALs (12.82-12.84 vs. 25.60-25.62), vasodilators (1.75-1.75 vs. 1.99-1.99), aspirin (47.78-47.78 vs. 60.34-60.36), and nitrates (47.20-47.26 vs. 69.51-69.53). Men without angina also received significantly more prescriptions than women for seven of the 12 medications of interest: ACE inhibitors (44.93-44.93 vs. 50.77-50.77), β blockers (33.64-33.64 vs. 34.20-34.20), a/p blockers (1.86-1.86 vs. 2.13-2.13), statins (59.80-59.80 vs. 63.72-63.72), all other ALs (10.37-10.37 vs. 16.09-16.09), aspirin (22.73-22.73 vs. 31.68-31.68), and nitrates (7.86-7.86 vs. 13.43-13.43).
Women with or without angina received more medication than men in three classes: ARBs (angina 20.97-21.01 vs. 12.78-12.78; nonangina 26.59-26.59 vs. 17.18-17.18), CCBs (CCB-Ds: angina 37.59-37.63 vs. 31.81-31.83; nonangina 29.13-29.13 vs. 25.07-25.07; CCB-NDs: angina 14.08-14.08 vs. 13.16-13.16; nonangina 25.34-25.38 vs. 21.94-21.96), and diuretics (angina 67.62-67.66 vs. 57.70-57.72; nonangina 62.49-62.49 vs. 47.10-47.10). These findings are consistent with prior analyses of other large samples drawn from the Hypertension Initiative primary care database.16,29
Given that this was a population at risk (hypertension and dyslipidemia) and high risk (hypertension, dyslipidemia, and angina) for CV events, the finding of substantial gender and age group differences in treatment for dyslipidemia was especially striking. As might be expected, there were significant differences when comparing younger same-gender patients to older same-gender patients in both the angina and nonangina groups; however, there were also significant gender differences within the age groups. When older women were compared with older men in each diagnostic category, significant medication differences persisted. These significant gender- and age-group differences were also found when the analysis focused separately on statins and all other AL medications. Even more striking was the fact that women in both diagnostic categories and age groups were more likely than men to receive no AL medication at all. While increased age and increased risk (diagnosis of angina) did narrow these medication gaps between genders and age groups, significant differences persisted. Regardless of HDL-C status, the evidence is abundant that hypertensive, dyslipidemic women—whether with or without angina—benefit from controlling LDL-C to levels recommended by national treatment guidelines.30,31 Hence, these findings represent an important treatment gap.
These particular disparities in AL medications appeared to be reflected in measurements of CV risk factor control. In all age and diagnosis groups, women were significantly less likely than men to have had a documented cholesterol measurement within the past year (no LDL-C: women, 46%; men, 39%). Among those with a cholesterol measurement within the past year, men were significantly more likely than women to have LDL-C <100 mg/dL (angina, 43.94-43.96 vs. 34.42-34.50; nonangina, 32.43-32.43 vs. 17.25-17.25) (Table II). Women were significantly more likely than men to have LDL-C >130 mg/dL (angina, 27.68-27.72 vs. 23.91-23.93; nonangina, 38.70-38.70 vs. 35.39-35.39). In the intermediate range between these benchmarks (i.e., 100-129 mg/dL) the gender difference was split by diagnosis. Women with angina more often fell into the intermediate range than men (35.10-35.18 vs. 32.12-32.14) but men without angina fell into the intermediate range more often than women (53.86-53.86 vs. 32.44-32.44).
Table II. Blood Pressure (BP) Among All Patients and Low-Density Lipoprotein Cholesterol (LDL-C) Control Among Patients With a Current Measurement, by Age, Gender, and Presence of Angina Diagnosis
Age <60 yr
Age <60 yr
Men (M) (n)
Women (W) (n)
BP Controland LDL-C Monitoring Among All Patients
BP <140/90 mm Hg
No current LDL-C
Ldl-C Control Among Patients Witha Current Ldl Measurement
LDL-C <100 mg/dL
LDL-C 100–129 mg/dL
LDL-C ≥130 mg/dL
Data shown are number or percentage among subsamples, shown as 95% confidence interval (CI). Significant differences are determined by comparisons between CIs. *p<0.05
Among women with an LDL-C measurement on record, those on statin medication exhibited the highest proportion of LDL-C <100 mg/dL, those on other AL medications (but not a statin) had the second highest proportion of LDL-C <100 mg/dL, and those on no AL medication at all had the lowest proportion with an LDL-C <100 mg/dL (Figure). A prescription for statins was also associated with a lower proportion having LDL-C >130 mg/dL among women either with or without angina (Figure).
Gender-based differences in BP were split by diagnosis group. Women without angina demonstrated BP <140/90 mm Hg more often than men (55.75-55.75 vs. 54.27-54.27); however, among patients with an angina diagnosis, more men had their BP controlled (63.86-63.88 vs. 59.74-59.80) (Table II).
While prescription medication treatments were used regularly among these hypertensive, dyslipidemic patients in primary care, men and older patients were treated substantially more aggressively than women and younger patients—despite similar diagnostic status. Specifically, several first-line medications, such as ACE inhibitors, aspirin, nitrates, and statins, which substantial clinical evidence indicates are highly effective in reducing CV morbidity and mortality, were not being used in accordance with national guidelines—especially in younger patients and women.
An angina diagnosis was associated with substantially higher rates of prescription medication treatments, as one would expect. This indicates that primary care physicians in this study responded to increased CV diagnoses with a more aggressive prescription treatment regimen.
However, it is notable that while an angina diagnosis was associated with increased prescription medications in all classes except ARBs, there were significant and consistent differences between men and women in both age groups across all other medication classes. This suggests that treatment decisions may be influenced more by patient demographics (i.e., age and gender) than by treatment guidelines.
The most important finding of this analysis was that dyslipidemic, hypertensive women in all age groups—both with and without an angina diagnosis—had substantially poorer LDL-C control than men. In addition, they were less likely to receive a statin or any other AL medication to control their cholesterol levels. It is highly likely that the finding of poorer LDL-C control is directly associated with the significantly lower rate of statin prescriptions among these women. This finding is particularly relevant as research indicates that small decreases in LDL-C can produce exponential benefits over time in terms of risk reduction for CV morbidity and mortality.30,31 Future studies on the impact of HDL-C on the CV benefits of LDL-C reduction in this high-risk population would be of interest.
This study concurs with previous research finding that medications and cholesterol control rates of hypertensive, dyslipidemic patients in primary care differ by patient gender and age.5,16,29 It also further supports several studies concluding that dyslipidemia is often undertreated, resulting in unnecessary CV compromise.24,25
This study was limited by the fact that the data were exclusively drawn from electronic medical records and cannot indicate patient compliance with treatment regimens prescribed. In addition, cell sizes among women with angina and a current LDL-C measurement were quite small due to the fact that fewer women receive an angina diagnosis, and a high proportion of women did not have a current LDL-C on record. Regardless, the positive association between AL prescriptions and LDL-C control was mirrored in the larger group of women without angina and was, intuitively, what one would expect to find. We feel it is unlikely that a larger sample of women with angina and current LDL-C measurements would reverse this trend, but larger studies are warranted. Another limitation might be the lack of HDL-C measurements in women, since some women with high HDL-C levels (>70 mg/dL) may not have been considered dyslipidemic even with total cholesterol levels >250 mg/dL or LDL-C levels >130 mg/dL.
Despite limitations, this analysis provided a useful picture of real-world, primary-care practice patterns in a very high-risk population. The gender-and age-related differences in cholesterol screening and treatment among hypertensive, dyslipidemic patients who are at risk for—or already have—angina represent significant opportunities to reduce CV morbidity and mortality, especially among women and younger patients.
Acknowledgment and disclosure: This study was supported by funding from CV Therapeutics Inc., the Agency for Health Research and Quality (Project EXCEED), the Duke Endowment, the US Centers for Disease Control and Prevention (through the South Carolina Department of Health and Environmental Control), the US Department of Health and Human Services (Stroke Belt Elimination Initiative), and the National Heart, Lung, and Blood Institute.