2021-09-02
Abstract
Experimental evidence and several small studies in humans suggest that HMG-CoA (3-hydroxy 3-methylglutaryl coenzyme A) reductase inhibitors (statins) reduce blood pressure, perhaps through effects on endothelial function or by reducing inflammation. We tested the hypothesis that pravastatin would reduce blood pressure at 3 months and the risk of developing new hypertension over a follow-up period of 5 years. This was a post hoc subgroup analysis of a randomized double-blind placebo-controlled trial of pravastatin 40 mg daily vs placebo in 4159 participants with previous myocardial infarction and total plasma cholesterol <240 mg/dl (6.2 mmol/l). The primary outcome was the unadjusted change in mean arterial pressure (MAP) from baseline to 3 months. We also considered systolic and diastolic blood pressure (SBP and DBP) and pulse pressure. Analysis of covariance was used to calculate the adjusted effect of treatment on change in these outcomes at 3, 6, 12 and 24 months postrandomization, after controlling for potential confounders. Logistic regression was used to calculate the adjusted effect of treatment on incident hypertension (blood pressure ⩾140/90 in those without known hypertension at baseline). This analysis included 4126/4159 (99.2%) participants for whom blood pressure was measured at baseline and during at least one follow-up visit. Median duration of follow-up was 57.8 months. The unadjusted and adjusted change in MAP, SBP, DBP or pulse pressure from baseline was not significantly different for pravastatin or placebo recipients at 3, 6, 12 or 24 months after randomization, or at last follow-up. Pravastatin did not reduce the adjusted risk of incident systolic hypertension (odds ratio 0.99, 95% CI 0.80–1.23), or incident diastolic hypertension (odds ratio 0.97, 95% CI 0.73–1.27). In summary, pravastatin 40 mg daily did not reduce blood pressure in survivors of myocardial infarction without overt hypercholesterolaemia.
Introduction
HMG-CoA (3-hydroxy-3-methylglutaryl coenzyme A) reductase inhibitors (statins) prevent cardiovascular events and reduce mortality in a wide range of patient populations.1, 2, 3, 4 Although these effects appear to be wholly or partially due to reduction in plasma lipids,5 there has been speculation that other mechanisms might also contribute.6 For instance, statins reduce markers of inflammation such as C-reactive protein,7 and may reverse endothelial dysfunction through their effects on nitric oxide synthase8 or by other mechanisms,9 although this is controversial.10 Since these abnormalities are associated with hypertension, it is plausible that statin therapy might also reduce blood pressure.11, 12
Several previous experimental and retrospective studies suggest that statins reduce blood pressure in animals and humans with and without hypertension.13, 14, 15, 16 However, randomized trials have reached conflicting results.17, 18 We analysed blood pressure data from a previously conducted randomized trial, which examined the cardiovascular benefits of pravastatin 40 mg daily compared with placebo. We hypothesized that pravastatin would reduce blood pressure and the risk of developing new hypertension over a follow-up period of approximately 5 years.
Methods
Study design and patients
The design of the Cholesterol and Recurrent Events (CARE) trial has been described in detail elsewhere.5 Briefly, men and postmenopausal women were eligible if they had had an acute myocardial infarction between 3 and 20 months before randomization, were 21–75 years of age, and had plasma total cholesterol levels of <240 mg/dl (6.2 mmol/l), low density lipoprotein (LDL) cholesterol levels of 115–174 mg/dl (3.0–4.5 mmol/l), fasting triglyceride levels of <350 mg/dl (4.0 mmol/l), fasting glucose levels of no more than 220 mg/dl (12.2 mmol/l), left ventricular ejection fractions of no less than 25%, and no symptomatic congestive heart failure. Patients with proteinuria >2+ on routine dipstick or serum creatinine levels >1.5 times the upper limit of normal for the central study laboratory were excluded from the CARE study.
After stratification, according to clinical center, eligible consenting participants were assigned by computer-generated random order to receive either 40 mg of pravastatin (Pravachol, Bristol Myers Squibb) once daily, or placebo. Treatment allocation was concealed using a centrally maintained code. Serum creatinine measurements using the alkaline picrate method of Jaffe were made annually by individuals at a central study laboratory, who were blinded to treatment assignment. These were used to estimate glomerular filtration rate (GFR), as calculated by the modification of diet in renal disease (MDRD) equation: 186 × PCr−1.154* age in years−0.203* 1.210 (if black)* 0.742 (if female), where PCr is plasma creatinine in mg/dl. This formula has been shown to correlate with iothalamate measurements of GFR.19
Measurement of blood pressure
The protocol specified that blood pressure was to be measured every 3 months during the study using manual or automated sphygmomanometers. Venipuncture was performed every 6 months and a full physical examination was performed at annual visits only. We included all individuals in whom blood pressure was measured before randomization and for whom at least one postrandomization blood pressure value was available.
Use of antihypertensive medications
Prescribed antihypertensive medications were recorded at each study visit. We did not assess adherence to antihypertensive regimens. Antihypertensive agents were classified into the following categories: angiotensin-converting enzyme inhibitors, beta-adrenergic blockers, alpha-adrenergic blockers, loop diuretics, thiazide diuretics, direct vasodilators, calcium channel blockers, and other agents.
Definition of incident hypertension
Participants who gave a history of hypertension prior to the baseline visit, or who had systolic blood pressure (SBP)⩾140 mm Hg or diastolic blood pressure (DBP)⩾90 mm Hg at prerandomization visits were considered to have preexisting hypertension.20 Incident systolic or diastolic hypertension was considered to have occurred when two consecutive values of blood pressure during follow-up exceeded 140 mm Hg (SBP) or 90 mm Hg (DBP) in participants without pre-existing hypertension. We also performed analyses in which incident hypertension was defined by either mean follow-up blood pressure ⩾140/90 or an increase in antihypertensive medications during follow-up (regardless of blood pressure).
Study outcomes
The primary outcome was the change in MAP at 3 months from baseline, which was calculated by (SBP+DBP+DBP)/3. Change at 3 months was also considered for SBP, DBP and pulse pressure. Additionally, the following timepoints: 6, 12, 24 months and at last follow-up were measured. The proportion of participants with incident hypertension was also considered.
Statistical analysis
We used analysis of covariance to model change in blood pressure from baseline using treatment group as the independent variable. In adjusted analyses, we used the following covariates: age (per year), race (black vs other), sex, glomerular filtration rate (ml/min/1.73 m2), and change in the number of antihypertensive medications. Additionally, to take advantage of multiple measurements of blood pressure during follow-up, we used a mixed model with a random intercept and random slope to compare changes in blood pressure over time between treatment groups. In these analyses, the number of antihypertensive medications prescribed was treated as a time-dependent covariate, which was updated for each blood pressure measurement, and only blood pressure measurements from annual study visits were included (i.e. only measurements which were made at baseline or after 12, 24, 36, 48 or 60 months of follow-up). Logistic regression analysis was used to examine the proportion of subjects developing hypertension over time. Analyses were performed for the following prospectively defined subgroups: all participants, in participants characterized by baseline renal dysfunction (GFR<60 ml/min/1.73 m2), in those receiving angiotensin-converting enzyme inhibitors, and in those not using any antihypertensive medication at baseline.
All analyses were intention-to-treat, P-values are 2-sided, and P<0.05 was considered statistically significant. Analyses were performed with SAS version 8.2 software (Cary, NC, USA). Values are reported as mean±s.d. or percentages; 95% confidence intervals are provided where appropriate. The ethics review board at the University of Alberta approved the study.
Results
Baseline characteristics
There were 4126/4159 (99.2%) CARE participants for whom blood pressure was measured at baseline and during at least one follow-up visit, and were thus eligible for this analysis. Less than 6% of blood pressure values were missing at all timepoints. Demographic and clinical characteristics of subjects are shown in Table 1. Median duration of follow-up was 57.8 months.
Table 1 Baseline characteristics of participants by treatment group
Use of antihypertensive agents
The mean number of antihypertensive agents was similar in the pravastatin and placebo groups at baseline, 3 and 6 months. The mean number of antihypertensive agents among pravastatin recipients was similar to placebo; 1.1±0.8, 1.1±0.9 and 1.1±0.9 for pravastatin, as compared with 1.0±0.8 at all three time-points for placebo recipients (P 0.08, 0.005 and 0.09, respectively). The median number of antihypertensive agents in each group was also similar at all time points (data not shown).
Change in blood pressure by treatment group
The change in MAP, SBP, DBP or pulse pressure was not significantly different for pravastatin or placebo recipients at any time point (Table 2). Results were similar after adjustment for age, sex, race, baseline glomerular filtration, and change in antihypertensive medications from baseline (data not shown).
Table 2 Unadjusted effect of pravastatin on change in blood pressure
Change in blood pressure in subgroups of participants
We repeated analyses considering only the 639 participants, who were prescribed a constant number of antihypertensive medications during follow-up. There was no evidence that pravastatin reduced MAP, SBP, DBP or pulse pressure in this subgroup (data not shown). Similarly, no significant effect of pravastatin was found in participants with GFR<60 ml/min/1.73 m2 at baseline, or those who were receiving angiotensin-converting enzyme inhibitors at baseline (data not shown).
Change in blood pressure by treatment group – multiple measurements per subject
We repeated analyses using annual blood pressure measurements from all participants, and controlling for number of antihypertensive agents as a time-dependent covariate. In these analyses, pravastatin did not significantly influence postrandomization MAP (0.0 mm Hg higher than control, 95% CI 0.4 lower, 0.5 higher), SBP (0.3 mm Hg higher than control, 95% CI 0.5 lower, 1.0 higher), or DBP (0.2 mm Hg higher than control, 95% CI 0.3 lower, 0.6 higher).
Risk of developing hypertension among participants without hypertension at baseline
During follow-up, 474 (26.3%), 242 (13.4%) and 576 (31.9%) without hypertension at baseline (n=1803) developed systolic, diastolic or either form of hypertension, respectively. In logistic regression analyses, there was no evidence that pravastatin influenced the adjusted risk of incident systolic hypertension (OR 0.99, 95% CI 0.80–1.23), or incident diastolic hypertension (OR 0.97, 95% CI 0.73–1.27). Pravastatin did not reduce the odds of incident hypertension in participants with baseline GFR<60 ml/min/1.73 m2, or those who were not taking any antihypertensive medications at baseline (data not shown). Results were unchanged when the definition of incident hypertension was expanded to include previously normotensive subjects with either mean follow-up blood pressure ⩾140/90 or an increase in antihypertensive medications during follow-up (regardless of blood pressure) (data not shown).
Discussion
In theory, statins might reduce blood pressure through beneficial effects on endothelial function,21 by increasing the availability of nitric oxide,8 by reducing vascular responsiveness to angiotensin,22 or through an anti-inflammatory mechanism, since statins reduce C-reactive protein, which in turn is associated with the development of hypertension.23 Despite these theoretical considerations, we found no evidence that pravastatin reduced blood pressure in these 4126 participants over a 5-year period. MAP, SBP and DBP were all similar in both treatment groups at all times during this study, with no evidence of acute or long-term blood pressure reduction attributable to pravastatin use. Results were not influenced when incident hypertension rather than absolute changes in blood pressure were considered or analyses were restricted to certain subgroups of participants. In particular, despite data suggesting that the beneficial effects of statins on endothelial dysfunction are enhanced when the renin/angiotensin system is interrupted,24 statins did not reduce blood pressure in patients receiving angiotensin converting enzyme inhibitors.
Previous small studies examining the effect of statins on blood pressure have reached conflicting results. For example, Ferrier et al.17 found that SBP pressure in 22 patients with hypertension declined by approximately 6 mm Hg after 3 months of treatment with atorvastatin 80 mg daily, a significantly greater reduction than was observed with placebo. MAP and DBP were also significantly reduced in the atorvastatin arm, although pulse pressure was not affected. A second randomized study compared the impact of dietary modification and angiotensin-converting enzyme inhibitor treatment on blood pressure with and without the addition of statin therapy (lovastatin 20 mg daily or pravastatin 10 mg daily) in 70 hypercholesterolemic subjects.25 These authors found a greater reduction in MAP from baseline among statin recipients (18±2 vs 12±1%, P<0.05), resulting in a 7 mm Hg difference in MAP between groups after 16 weeks despite equivalent MAP at baseline. These data are consistent with experimental data and retrospective studies in humans suggesting that statin use may reduce blood pressure,13, 14, 15, 16, 17, 25, 26 possibly independent of their effects on serum cholesterol.27 However, three previous blinded, randomized studies that included a total of 480 patients found no significant effect of statins on blood pressure,18, 28, 29 although the confidence limits for the individual studies did not exclude a moderate antihypertensive effect of statin treatment.
The current analysis has several important strengths, including its double-blind design, standardized dose of pravastatin, large sample size, comprehensive array of subgroup analyses, and relatively long follow-up. However, this analysis also has potentially significant limitations, which should be considered when interpreting the findings. First, because the CARE study was not designed to investigate the effect of pravastatin on hypertension, blood pressure was not necessarily measured in accordance with accepted guidelines. For example, blood pressure was not measured in duplicate at each visit, participants were not required to rest quietly before measurements were taken, both automated and manual sphygmomanometers were used, and personnel measuring blood pressure were not trained or monitored according to rigorous research methods. This lack of standardization may have reduced our ability to detect an effect of pravastatin on blood pressure. Second, participants were allowed to change antihypertensive regimens during the study, which might lead to bias if placebo recipients were more likely to receive additional agents to offset slightly higher blood pressure. However, the mean number of antihypertensive agents was similar in both treatment groups at baseline and during follow-up, and we adjusted for antihypertensive use. In addition, analyses conducted in patients without any changes in antihypertensive medication did not suggest any reduction in blood pressure among pravastatin recipients. Third, since all participants in the treatment group received pravastatin 40 mg daily, our results do not allow conclusions about the effect of higher doses or other statins on blood pressure. Fourth, participants with overt hypercholesterolaemia were excluded from the CARE study, which together with the other criteria for study entry may have reduced the generalizability of our findings, especially since the majority of subjects had evidence of hypertension at baseline and all had a history of previous myocardial infarction. Fifth, no patients in the CARE study were receiving angiotensin receptor blockers, so we could not test for an interaction between these agents and the effect of statins on blood pressure.24 However, statins did not have an antihypertensive effect among patients receiving angiotensin converting enzyme inhibitors in the current study. Finally, this is a post hoc analysis, the limitations of which are well described.
The recently published ASCOT study (19 257 participants) compared the incidence of cardiovascular outcomes between two commonly used antihypertensive regimens.30 Since blood pressure management was standardized in this study, which also included an atorvastatin vs placebo comparison (using a factorial design),31 ASCOT may permit definitive evaluation of the effect of statins on blood pressure. Preliminary results from ASCOT suggest that atorvastatin may reduce SBP by approximately 1 mm Hg at 6 months, which was nonsignificant by 2 years.32 However, the number of antihypertensive medications was significantly lower in atorvastatin recipients, which may have obscured a larger difference between treatments. Possibilities for the discrepant findings in our study include the standardized blood pressure management used in ASCOT, or differences in the study population or specific statin used. Differences in lipid reduction between studies are unlikely to be responsible, since mean pre- and 6 month posttreatment LDL-C were similar between the two trials (139 and 98 mg/dl for CARE vs 133 and 85 mg/dl for ASCOT). A more detailed discussion of the explanation for the discrepant findings compared with the current study must await full publication from the ASCOT investigators. In the interim, our results argue against the hypothesis that pravastatin exerts a clinically significant influence on blood pressure, at least in the population studied and the range of on-treatment lipid levels achieved.
In summary, we found no evidence that pravastatin 40 mg daily reduced blood pressure in survivors of myocardial infarction without overt hypercholesterolaemia. However, further studies with standardization of blood pressure measurements and treatment would be required to conclusively exclude the possibility that statins modestly reduce blood pressure.
