Sex Differences in Cardiovascular Medication Prescription in Primary Care: A Systematic Review and Meta‐Analysis

Abstract

Background

Sex differences in the management of cardiovascular disease have been reported in secondary care. We conducted a systematic review with meta‐analysis of systematically investigated sex differences in cardiovascular medication prescription among patients at high risk or with established cardiovascular disease in primary care.

Methods and Results

PubMed and Embase were searched between 2000 and 2019 for observational studies reporting on the sex‐specific prevalence of aspirin, statins, and antihypertensive medication prescription, including beta blockers, calcium channel blockers, angiotensin‐converting enzyme inhibitors, and diuretics, in primary care. Random effects meta‐analysis was used to obtain pooled women‐to‐men prevalence ratios for each cardiovascular medication prescription. Metaregression models assessed the impact of age and year on the findings. A total of 43 studies were included, involving 2 264 600 participants (28% women) worldwide. Participants’ mean age ranged from 51 to 76 years. The pooled prevalence of cardiovascular medication prescription for women was 41% for aspirin, 60% for statins, and 68% for any antihypertensive medications. Corresponding rates for men were 56%, 63%, and 69% respectively. The pooled women‐to‐men prevalence ratios were 0.81 (95% CI, 0.72–0.92) for aspirin, 0.90 (95% CI, 0.85–0.95) for statins, and 1.01 (95% CI, 0.95–1.08) for any antihypertensive medications. Women were less likely to be prescribed angiotensin‐converting enzyme inhibitors (0.85; 95% CI, 0.81–0.89) but more likely with diuretics (1.27; 95% CI, 1.17–1.37). Mean age, mean age difference between the sexes, and year of study had no significant impact on findings.

Conclusions

Sex differences in the prescription of cardiovascular medication exist among patients at high risk or with established cardiovascular disease in primary care, with a lower prevalence of aspirin, statins, and angiotensin‐converting enzyme inhibitors prescription in women and a lower prevalence of diuretics prescription in men.

Nonstandard Abbreviations and Acronyms

ACEI

angiotensin‐converting enzyme inhibitors

Antihtn

antihypertensive medications

BB

beta blocker

CCB

calcium channel blocker

CHD

coronary heart disease

CVD

cardiovascular disease

Clinical Perspective

What Is New?

• This systematic review with meta‐analysis shows that there are sex differences in cardiovascular medication prescription among patients at high risk or with established cardiovascular disease in primary care.

• Women were less likely to be prescribed aspirin, statin, or angiotensin‐converting enzyme inhibitor but more likely to have a prescription for diuretics.

What Are the Clinical Implications?

• Sex differences in cardiovascular prescription in primary care need to be addressed in order to optimize the use of cardiovascular medication for both women and men.

Cardiovascular disease (CVD) remains the leading cause of death worldwide, accounting for about a third of all deaths in both women and men.1 Historically, there has been a misperception that CVD predominantly affects men, which may have resulted in suboptimal management and treatment of CVD in women.2, 3 Over recent decades, substantial efforts have been made to characterize CVD in women. As a result, important differences between women and men in the presentation, diagnosis, and medical treatment of CVD have been identified.2, 4

Most studies on sex differences in CVD management have been performed in secondary care.3, 5, 6, 7 For example, among all patients receiving statins after hospitalization for myocardial infarction in the United States, women were less likely than men to receive high‐intensity statins, despite guideline recommendations.6 Also, a study of coronary heart disease patients recruited from routine outpatient cardiology clinics in 11 countries across Europe, Asia, and the Middle East showed that women were less likely than men to reach all treatment targets set by clinical guidelines.3 Whether similar sex differences exist in primary care has not been systematically evaluated. Considering that both patients at high risk and with established CVD attended clinics in primary care to monitor their current CVD treatment, primary care visits are a key stage at which any sex inequities in treatment could and should be investigated. Comprehensive evidence on current sex differences in cardiovascular medication prescription in primary care would help to obtain a better understanding of the utilization of evidence‐based medical treatment for both sexes and encourage all health professionals to strive for sex equity in providing CVD management to their patients.

In this study, we conducted a systematic review and meta‐analysis to determine the prevalence of common cardiovascular medication prescription in women and men in primary care and to evaluate whether prescriptions for guideline‐recommended cardiovascular medications differ between the sexes.

Methods

The authors declare that all supporting data are available within the article and its online supplementary files.

Search Strategy

A systematic search of observational studies was performed in PubMed/MEDLINE and Embase for studies published between 2000 and 2019 using combined text word subject heading terms (Table S1). The reference lists of all related articles were screened for any other potentially relevant studies.

Study Selection and Data Extraction

All observational studies that reported the sex‐specific prevalence of prescriptions of cardiovascular medications (aspirin, statins, and any antihypertensive medication including beta blockers, calcium channel blockers [CCBs], angiotensin‐converting enzyme inhibitors [ACE inhibitors], and diuretics) for patients at high risk or with established CVD (coronary heart disease, stroke, heart failure, and atrial fibrillation) in primary care were included. Studies were excluded if they (1) were published in a language other than English; (2) presented an unrelated study population, outcome, or were not performed in primary care; (3) included <1000 patients; (4) reported cardiovascular medication prescription only for 1 sex; and (5) assessed cardiovascular medication not by prescription (such as self‐report or pharmacy dispensing).

Duplicate records were removed before title and abstract screening. When there were multiple reports from the same study, the report involving the highest number of cases or most explicit participants characteristics and outcome measures was included. Four independent reviewers (M.Z., E.R.C.M., C.C., and K.H.) screened the papers by title and abstract against the inclusion and exclusion criteria. Any disagreement between reviewers was discussed and the full text was reviewed, if necessary. A similar process took place in reviewing the full text of selected papers. A tailor‐made data extraction form was used to collect information on study and participant characteristics and sex‐specific prevalence of prescriptions of cardiovascular medication (Table S2).

Quality Assessment

Study quality was assessed using the modified Newcastle‐Ottawa scale for observational studies. This scale consists of 6 items that assess the quality of participant selection, comparability, and outcome adjudication (Tables S3 and S4).8

Outcomes

The primary outcome was the women‐to‐men prescription prevalence ratio with 95% CI for each cardiovascular medication. The secondary outcomes were the sex‐specific prescription rates of each cardiovascular medication.

Statistical Analysis

In general, the included studies reported unadjusted numbers, rates, or percentages of women and men with cardiovascular medication prescriptions. If a measure of variability was not reported, these were estimated from the rate and the sample size. The women‐to‐men prevalence ratios with 95% CI were pooled across studies using random‐effects meta‐analyses with inverse‐variance weighting for each medication.9 In sensitivity analysis, we pooled the results from studies that had adjusted for age. As different studies reported on different antihypertensive medications, we also restricted the analyses on individual antihypertensive medications to studies that reported on each of the 4 antihypertensive medications. Metaregression analyses were performed to assess the impact of mean age and age difference (women minus men) on our findings. We further investigated whether there was a trend in sex differences in cardiovascular medication prescription over time. In subgroup analysis, we assessed whether the findings differed by CVD status (high risk only, prevalent CVD, and high risk and prevalent CVD combined). P<0.05 were considered statistically significant. Statistical analyses were performed by using the “metafor” package in R version 3.2.2.

Results

Study Characteristics

Of the 10 803 studies identified through the systematic search, 900 studies were reviewed in full text (Figure 1). Of these, 43 studies were included, including a total of 2 264 600 participants, of whom 630 111 (28%) were women. The mean age ranged from 51 to 76 years (where reported). Table shows the key characteristics of the included studies. Of the 43 studies, 18 included information on aspirin,10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27 30 on statins, ^* 14 on any antihypertensive medications, ^† 21 on beta blockers, ^‡ 13 on CCBs, ^§ 21 on ACE inhibitors, ^|| and 14 on diuretics. ^¶ Eight out of 43 studies reported cardiovascular medication prescription for high‐risk patients,17, 32, 38, 47, 48, 49, 52, 53 24 for patients with established CVD, ^# and 11 for both high‐risk and CVD patients. ^**

Figure 1. Flowchart of records screened and included in the systematic review.

ACEI indicates angiotensin‐converting enzyme inhibitor; and CCB, calcium channel blocker.

Table 1.

Key Characteristics of Selected Studies

Study	Year	Country	Prevention Type	Sample Size	Women	Men	Age, y	Cardiovascular Medications
								Aspirin	Statin	Antihtn	BB	CCB	ACEI	Diuretics
Al‐Lawati et al10	2007	Oman	Mixed	2551	1352	1199	54	X		X	X	X	X	X
Alberts et al28	2004	Multiple	Secondary	55 499	18 315	37 184	69		X
Brady et al11	1998	UK	Secondary	24 431	9898	14 533	67	X	X		X	X	X	X
Brady et al20	2002	UK	Secondary	12 045	4457	7588	67	X	X		X		X	X
Bull et al31	2003a	UK	Secondary	13 929	5827	8102	>40		X
Carlsson et al21	2013a	Sweden	Secondary	7408	3330	4078	76	X	X		X	X		X
Carroll et al22	2001	UK	Secondary	6778	2787	3991	NA	X	X		X
Catalán‐Ramos et al32	2009	Spain	Primary	696 073	358 218	337 855	51		X		X	X	X	X
Crilly et al23	2001	UK	Secondary	1162	552	610	69	X	X		X
Dodhia et al33	2013	UK	Secondary	6711	2828	4564	70		X
Dreyer et al34	2007	Australia	Secondary	2005	721	1284	70		X		X	X	X
Driscoll et al24	2007	Australia	Secondary	12 509	5267	7242	73	X		X
Emberson et al25	2001	UK	Mixed	8538	4286	4252	NA	X	X	X	X		X
Forster et al35	2013	UK	Secondary	23 811	4502	4252	NA		X	X
Greving et al49	2000	NL	Primary	7550	4774	2776	63				X	X	X	X
Gulliford et al36	2010a	UK	Secondary	7065	3816	3249	73		X	X
Hawkins et al50	2007	UK	Secondary	13 330	6803	6527	68				X		X
Hendrix et al26	2005a	US	Mixed	72 508	29 208	43 300	NA	X	X		X	X	X	X
Hippisley‐Cox et al27	2001a	UK	Mixed	5891	2783	3108	NA	X			X
Hyun et al37	2012	Australia	Mixed	13 294	6202	7092	61		X	X
Journath et al38	2005	Sweden	Primary	6537	3410	3127	66		X	X	X	X	X	X
Lahoz et al12	2008a	Spain	Secondary	8817	2319	6498	65	X	X		X	X	X	X
Law et al44	2010	Canada	Primary	390	128	262	58		X
Lawlor et al29	2000	UK	Secondary	1314	483	831	NA		X
Lee et al19	2018	Australia	Secondary	130 926	61 142	69 784	67	X	X	X	X		X
Macchia et al13	2012a	Italy	Secondary	21 423	6928	14 495	NA	X	X		X		X
Majeed et al39	1996	UK	Secondary	63 259	34 545	28 714	NA		X
Majeed et al14	2002	UK	Secondary	2129	1224	905	NA	X					X	X
Murphy et al51	2004a	UK	Secondary	2186	1213	973	NA				X		X	X
Nanna et al46	2015	US	Mixed	5693	2460	3233	68		X
Nilsson et al40	2004a	Sweden	Mixed	9375	4293	5082	65		X				X
Nilsson et al52	2007a	Sweden	Primary	1135	714	421	52					X	X	X
Owen et al47	2009a	Australia	Primary	12 499	5896	6603	63			X	X	X	X	X
Paulsen et al48	2011a	Denmark	Primary	5413	3305	2108	66			X	X	X	X	X
Qato et al15	2011	US	Mixed	4136	2233	1903	52	X	X
Saposnik et al30	2004	Canada	Secondary	1094	415	679	67		X	X
Sheppard et al41	2009	UK	Mixed	4699	1937	2762	54		X
Svilaas et al16	2000a	Norway	Secondary	2060	707	1353	69	X
Tabenkin et al17	2004	US	Primary	407	210	197	53	X		X			X
Turnbull et al42	2008	Australia	Mixed	3664	1834	1830	68		X	X
Virani et al43	2011	US	Secondary	972 532	13 371	959 161	71		X
Weler et al18	2003	US	Mixed	3849	1953	1896	65	X		X
Wandell et al45	2007	Sweden	Secondary	7975	3465	4510	NA		X		X	X	X	X

ACEI indicates angiotensin converting enzyme inhibitor; Antihtn, any anti‐hypertensive medication; BB, beta blocker; CCB, calcium channel blocker; EU, Europe; NL, The Netherlands; UK, United Kingdom; and US, United States.

^aYear: study performed year. Studies with asterisk indicate publication year.

Sex Differences in Prevalence of Cardiovascular Medication Prescription

In women, the pooled prevalence of cardiovascular medication prescription was 41% for aspirin, 60% for statins, and 68% for overall antihypertensive medications. The corresponding rates for men were 56%, 63%, and 69%, respectively. The pooled women‐to‐men prevalence ratios were 0.81 (95% CI, 0.72–0.92) for aspirin, 0.90 (95% CI, 0.85–0.95) for statins, and 1.01 (95% CI, 0.95–1.08) for any antihypertensive medications (Figure 2).

Figure 2. Women‐to‐men prevalence ratio of aspirin, statins, and any antihypertensive medications prescription.

For each study, the square is centered on the women‐to‐men prevalence ratio and the horizontal lines show the associated 95% CI. The diamond indicates the pooled summary and its 95% CI.

Figure 3 shows the women‐to‐men prevalence ratios of individual antihypertensive medication prescription. Women were less likely to be prescribed with ACE inhibitors (women‐to‐men prevalence ratio: 0.85; 95% CI, 0.81–0.89) whereas the prevalence of diuretics prescription was higher than in men (women‐to‐men prevalence ratio: 1.27; 95% CI, 1.17–1.37). There were no significant sex differences in the prescription of beta blockers and CCBs. Findings were similar in analyses restricted to studies that reported on all 4 individual antihypertensive medications (Figure S1). Findings were similar in age‐adjusted analyses, available for 31 studies (Tables S5 through S10).

Figure 3. Women‐to‐men prevalence ratio of individual antihypertensive medication prescription.

For each study, the square is centered on the women‐to‐men prevalence ratio and the horizontal lines show the associated 95% CI. The diamond indicates the pooled summary and its 95% CI.

Impact of Age on the Sex Differences in Prevalence of Cardiovascular Medication

Among the 31 studies that reported a sex‐combined mean age of the study population, there was no evidence that the women‐to‐men prevalence ratio varied systematically according to the mean age (Figure S2; P values: 0.57 for aspirin; 0.24 for beta blockers; 0.27 for CCBs; 0.41 for ACE inhibitors; 0.85 for diuretics). The only exception was that in studies with older patients, women were less likely than men to be prescribed statins whereas women had a higher prevalence of statin prescription compared with men in studies including younger patients (P=0.003).

Among the 17 studies that reported sex‐specific mean ages, there was no evidence that the prevalence ratio varied systematically according to the women to men age difference (Figure S3; P values: 0.34 for aspirin; 0.21 for statins; 0.93 for beta blockers; 0.91 for CCBs; 0.89 for ACE inhibitors). The exception was the higher prevalence of diuretics prescription in women increased as the difference between the mean age of women and the mean age of men increased (P=0.006).

Sex Differences in the Prevalence of Cardiovascular Medication Prescription Over Time

The sex differences in prevalence ratio of prescription did not significantly change over time for aspirin (P=0.92), any antihypertensive medications (P=0.99), beta blockers (P=0.43), CCBs (P=0.44), ACE inhibitors (P=0.39), and diuretics (P=0.58) (Figure S4). However, the pattern and magnitude of the sex differences in statin prescription changed over time, with an increased women‐to‐men prevalence ratio (P=0.003).

Sex Differences in Cardiovascular Medication Prescription by CVD Status

Among patients with established CVD, women were less likely to be prescribed with aspirin (0.89, 95% CI, 0.84–0.94), statins (0.85; 95% CI, 0.80–0.90), beta blockers (0.90, 95% CI, 0.85–0.96), and ACE inhibitors (0.88, 95% CI, 0.84–0.93) (Figure S5, Table S11). In contrast, women with established CVD were more likely to be prescribed with diuretics than their male counterparts (1.25; 95% CI, 1.09–1.43). Similar pooled estimates, but with wider CIs, were found when studies included only high‐risk participants, or when studies included both participants at high risk of and with established CVD. Time trends in the women‐to‐men prevalence ratio in medication prescription did not differ materially by CVD status (Figures S6 through S11). However, the women‐to‐men ratio of statin prescription increased over time in studies among high‐risk patients but not in studies including patients with established CVD or in studies including both high‐risk and CVD patients (P for interaction=0.002).

Discussion

In this systematic review and meta‐analysis of 43 studies including over 2 million participants, we found that there were sex differences in cardiovascular medication prescription among patients at high risk or with established CVD in primary care. Compared with men, women were less likely to have a prescription for aspirin, statins, or ACE inhibitors but more likely to have a prescription for diuretics. Sex differences did not vary materially by age, but there was some evidence to suggest that the magnitude of sex differences in statin prescription increased over time.

Previous studies in secondary care have demonstrated that women are generally less likely than men to have a prescription of guideline‐recommended cardiovascular medications after a cardiac event.2, 3, 5, 54 SUrvey of Risk Factors, a clinical audit with over 10 000 patients from 11 countries, indicated that women had a lower prevalence of cardiovascular medication use than men and were less likely to reach treatment targets.3 Similarly, a study of 36 000 patients with established coronary heart disease in the United States, showed that women were less likely than men to be prescribed with aspirin, ACE inhibitors, or statins at both acute and hospital discharge of coronary heart disease.55 A study in the United Kingdom showed that prescription rates for cardiovascular medications were about 10% lower among women than men <55 years for acute myocardial infarction.56 Furthermore, a Dutch population‐based analysis also found persistent sex differences in the use of lipid‐lowering medications for secondary prevention of CVD, particularly in younger patients.5 We did not observe that sex disparities differed between age groups, but we noticed that the sex differences in statin prescription persisted and was even larger in the more recent studies. A recent study in the United States confirmed that women were 9% less likely than men to receive high‐intensity statins, as opposed to other types of statin.57 The present study further expands these findings by showing that sex differences in medication prescription also exist among patients at high cardiovascular risk or with established CVD in a primary care setting. We also demonstrated that women were more likely to be on diuretics but less likely to be on ACE inhibitors, which is in line with other studies.56, 58, 59 Sex differences in progression and presentation of CVD and comorbidities, the efficiency of treatment, and/or adverse drug effects may lead to different requirements on antihypertensive regimens.59, 60 The reasons for the contrasting sex differences within antihypertensive medication classes require further study.

There are several other possible explanations for the lower prescription rates of some cardiovascular medications in women than men. First, the incidence of CVD in women is, typically, about a third that of men in middle age and occurs in men about a decade earlier than women, which might have led to the misperception that CVD is less common in women and does not have to be prevented as intensively as in men.4, 34, 61 Additionally, women may have a lower awareness of the severity of their disease and of appropriate CVD treatment and receive less support from healthy providers, compared with men, resulting in lower health consciousness and less frequent use of healthcare services.5, 62, 63, 64

Although beyond the scope of the current investigation, studies have reported a considerable delay in receiving appropriate medical treatment to reduce the risk of incident or recurrent cardiac event in women.2, 23, 62, 63 Also, women may have less belief than men in the safety and effectiveness of cardiovascular medications and have been reported to have a greater risk of suffering adverse drug reactions, which may lead to a higher discontinuation rate of cardiovascular medications.60, 65, 66, 67 Indeed, studies have shown that women have a poorer adherence to cardiovascular medication than men in primary care.68, 69 These factors would be expected to produce a wider disparity between the usage of cardiovascular medications than our study of prescriptions suggests.

We conducted a large‐scale systematic review with meta‐analyses on sex differences in cardiovascular medication prescription among patients at high risk or with established CVD in a primary care setting. We included all major cardiovascular medications and found that our results were generally robust across patient characteristics. Limitations of this study are inherent to its design and include the differences across studies in design, population, and end point definition.9 We had no information on potential combinations of cardiovascular medications prescribed, nor were we able to adjust our findings to potentially important comorbidities or other characteristics. However, some cardiovascular medications target the same risk factor and the lower use of ACE inhibitors among women, relative to men, could be explained by women's higher use of diuretics. Also, we considered sex differences only in medication prescription and were not able to determine whether those differences, where found, resulted in different levels of risk factor control and event rates. Furthermore, patients with established CVD seen in primary care may also receive treatment from secondary care. Also, it is not clear whether general practitioners or cardiologists would be the main source of prescriptions in any individual case. Finally, as the studies included in this review were conducted in mostly high‐income countries, the generalizability of our findings to low‐ and middle‐income countries needs to be assessed.

In conclusion, this meta‐analysis, summarizing all recent literature, shows that sex differences in cardiovascular medication prescription persist in primary care. Future research is needed to determine the underlying causes of observed sex differences and to develop tailored strategies to optimize the use of evidence‐based cardiovascular medication for both women and men.

Sources of Funding

Zhao is supported by a grant from the Netherlands Organization for Scientific Research (NWO; grant number: 0.22.005.021). Woodward is supported by National Health and Medical Research Council (NHMRC) Australia project grant 632507 and Fellowship APP1080206. Vaartjes is supported by a grant from the Dutch Heart Foundation (grant DHF project “Facts and Figures”). Hyun is supported by National Heart Foundation Australia Postdoctoral Fellowship (102138). Peters is supported by a UK Medical Research Council Skills Development Fellowship (MR/P014550/1).

Disclosures

Woodward is a consultant to Amgen and Kirin. The remaining authors have no disclosures to report.

Supporting information

Tables S1–S11

Figures S1–S11

References 10–52

(J Am Heart Assoc. 2020;9:e014742 DOI: 10.1161/JAHA.119.014742.)