Impact of blood pressure lowering on cardiovascular outcomes in normal weight, overweight and obese individuals: the PROGRESS Trial

Abstract

There is considerable uncertainty regarding the efficacy of blood pressure-lowering therapy in reducing cardiovascular risk in obese people. In this report we examine the effects of blood pressure lowering according to baseline body mass index (kg/m²) in the Perindopril Protection Against Recurrent Stroke Study (PROGRESS). A total of 6105 participants with cerebrovascular disease were randomized to perindopril-based blood pressure-lowering therapy or placebo. The overall mean difference in systolic/diastolic blood pressure between participants assigned active therapy or placebo was 9/4 (SE, 0.5/0.3) mmHg, with no difference by body mass index quarters: < 23.1, 23.1-25.3, 25.4-27.8 and ≥ 27.9 kg/m². A consistent treatment benefit was demonstrated for protection against major vascular events across quarters with the following hazard ratios (95% confidence intervals): 0.80 (0.62-1.02), 0.78 (0.61-1.01), 0.67 (0.53-0.86), 0.69 (0.54-0.88) and 0.74 (0.66-0.84) (p for heterogeneity = 0.16). Similar results were apparent for stroke and stroke subtypes (all p for heterogeneity ≥ 0.07) or with the standard definitions of overweight and obesity (< 25, 25 to 29 and ≥ 30 kg/m) (all p for heterogeneity ≥ 0.28). The absolute effects of treatment were, however, more than twice that in the highest compared with the lowest body mass index quartile. Across increasing quarters of body mass index over five years, active therapy prevented one major vascular event among every 28, 23, 13 and 13 patients treated. In conclusion, blood pressure-lowering therapy produced comparable risk reductions in vascular disease across the whole range of body mass index in participants with a history of stroke. However, the greater baseline level of cardiovascular risk in those with higher body mass index meant that these patients obtained the greatest benefit.

Introduction

Overweight and obesity are common, affecting more than 1.1 billion individuals worldwide.¹ In several industrialised countries, approximately two-thirds of the adult population is classified as overweight or obese on the basis of having a body mass index (BMI) in excess of 25 kg/m².² In addition to reducing life expectancy, excess weight is an independent risk factor for a wide spectrum of chronic disorders, in particular, type-2 diabetes, cardiovascular disease (CVD) and some site-specific cancers.³

Studies have shown that a gain in BMI of 2.1/2.7 kg/m² (men/women) is associated with a 2.2 mmHg increment in systolic blood pressure,⁴ and likewise that a weight loss of 1 kg results in a 1 mmHg reduction in systolic blood pressure.⁵ There is some indication that the magnitude of the association between blood pressure and subsequent CVD or stroke is stronger in obese, compare to lean individuals.⁶ Furthermore, as indicated in North American and European current guidelines for the management of hypertension, because available trials in hypertensive obese are scarce, there are no specific recommendations for high blood pressure management in patients with excess weight,^7;8.^9-15

In this report, we describe the results of new analyses from the PROGRESS study, a large placebo-controlled trial of a perindopril-based BP lowering regimen in people with prior cerebrovascular disease. The primary aim of this analysis was to assess the effects of blood pressure lowering on major cardiovascular events according to BMI at baseline. A secondary aim was to address the same question for the outcomes of stroke and its subtypes, ischemic and hemorrhagic.

Materials and Methods

Main study

The design of the PROGRESS study has been described in detail previously.¹⁶ In summary, 6105 individuals with a history of cerebrovascular disease (ischemic, hemorrhagic or transient ischemic attack, but not subarachnoid hemorrhage) within the previous 5 years, and no clear indication for, or contraindication to, treatment with an angiotensin-converting enzyme inhibitors (ACEI) were recruited to the study from 172 centers in 10 countries.

Eligible participants received perindopril (2mg for two weeks, followed by 4mg for 2 weeks) during a 4 week open label active run-in period. Participants who tolerated and adhered to this treatment were subsequently randomly allocated to active therapy or matching placebo. Active treatment comprised a flexible treatment regimen based on perindopril (4 mg daily) in all participants, with the addition of indapamide (2.5 mg daily; or 2 mg daily in Japan) in those for whom the responsible study physician felt that there was no specific indication for, nor contraindication to, the use of a diuretic. Those participants assigned placebo received one or two tablets identical in appearance to the active agent(s). ‘Combination therapy’ (perindopril and indapamide or double placebo), rather than ‘single drug therapy’ (perindopril or single placebo), was used wherever possible, in order to maximize the reduction in blood pressure. However, as many investigators had concerns about the safety of blood pressure lowering in patients with stroke (particularly in those with average or below average levels of blood pressure), it was necessary to provide some flexibility with respect to the intensity of treatment. All other aspects of medical care of the patients were left to the discretion of the responsible physician.

Body mass index

Height and weight were measured during the run-in period. BMI (weight, kg/height, m²) was categorized into quarters of the BMI distribution (<23.1, 23.1-25.3, 25.4-27.8 and ≥ 27.9 kg/m²). Supplementary analyses were also performed according to World Health Organization (WHO) cut points used to define normal, overweight and obese status: <25, 25-29 and ≥ 30 kg/m².¹⁷

Outcomes

The predefined primary outcome for this analysis was ‘major cardiovascular events’ defined as the composite of non-fatal stroke, non-fatal myocardial infarction and cardiovascular death. Secondary outcomes were total stroke, ischemic stroke and hemorrhagic stroke¹⁸. Stroke was defined as a neurological deficit lasting at least 24 hours and thought to be due to cerebral ischemia or hemorrhage.

Data analysis

The incidence of events was calculated using the person-year method. All analyses of treatment effect were performed on an intention to treat basis. The level of blood pressure reduction across BMI categories was tested by adding an interaction term to the relevant linear mixed model after adjustment for combination therapy status. A test for trend across these categories was computed. The effects of randomized treatment on events were calculated using univariate Cox proportional hazards model. Consistency of treatment effect across BMI quarters, and usual WHO cut points, was examined using tests of homogeneity that were performed by adding an interaction term to Cox models. Consistency of treatment was also tested by fitting BMI as a continuous variable and including a “treatment*BMI” interaction term in the model. Mean difference in blood pressure over time between randomized groups was calculated from linear mixed model for each BMI group by subtracting the values for the placebo group from the values for the active group. Additionally, because the overall effect of treatment was greater among participants treated with combination therapy (perindopril and indapamide vs. double placebo) than those treated with single-drug therapy (perindopril vs. single placebo), treatment effects in each BMI group were standardized for all outcomes for the proportions of the study population for whom combination (58%) or single-drug therapy (42%) was prescribed, by taking weighted averages of the estimates obtained for the two therapies.¹⁹ The absolute risk reduction (95% CI) over 5 years was calculated as the difference in incidence of events between the placebo and the active therapy group. Numbers needed to treat were calculated as reciprocals of the absolute risk differences. A p-value below 0.05 was considered statistically significant. All analyses were performed using SAS (SAS Institute, Inc, Cary, NC).

Results

A total of 6105 participants were recruited from 172 centers in 10 countries. During a mean follow-up of 4 years there were 1062 major primary and secondary vascular events including 724 strokes (565 ischemic and 111 hemorrhagic). At study entry, 2545 (41.9%) participants were overweight and 739 (12.2%) obese. Baseline characteristics of participants according to baseline BMI quarters are shown in Table 1. Median BMI across quarters ranged from 21.6 to 29.9 kg/m². Overall, participants in the highest quartile were, less likely to be Asian or smokers, but more likely to have preexisting coronary disease, diabetes and to have higher systolic and higher diastolic blood pressure levels. Baseline calcium antagonist use was similar across BMI quarters (p=0.7149), whereas beta-blocker and diuretic use increased across quarters (p<0.0001). Furthermore, the proportion of baseline users of two or more BP-lowering agent increased in quarters: 8, 12, 12 and 17%, respectively (p<0.0001). The proportion of participants receiving study treatment combination drug therapy also increased across BMI quarters (p for trend < 0.001).

Table 1.

Baseline characteristics according to body mass index (BMI) categories in PROGRESS

Variables		BMI categories (kg/m2)
		Q1 (n=1516)	Q2 (n=1547)	Q3 (n=1522)	Q4 (n=1520)	p for trend
Body mass index, kg/m2	median (range)	21.6 (13.7-23.1)	24.2 (23.1-25.4)	26.4 (25.4-27.9)	29.9 (27.9-48.4)
Age, years	mean (SD)	65 (10)	64 (10)	64 (9)	63 (9)	<0.0001
Women	%	37	26	26	33	0.03
Asian*	%	52	41	37	23	<0.0001
Systolic blood pressure, mmHg	mean (SD)	145 (19)	147 (19)	146 (19)	149 (19)	<0.0001
Diastolic blood pressure, mmHg	mean (SD)	84 (11)	86 (11)	86 (11)	87 (11)	<0.0001
Medical history
Ischemic stroke	%	70	72	70	70	0.63
Hemorrhagic stoke	%	14	10	11	8	<0.0001
Stroke of unknown type	%	5	5	3	6	0.32
Transient ischemic attack	%	20	21	24	25	<0.0001
Coronary heart disease †	%	12	16	18	19	<0.0001
Diabetes	%	9	11	12	18	<0.0001
Current smoker	%	25	21	17	17	<0.0001
Medication
Any antihypertensive therapy	%	43	50	53	54	<0.0001
Ca-antagonist	%	38	41	41	39	0.7149
Beta-blocker	%	10	16	18	24	<0.0001
Diuretics	%	8	10	12	17	<0.0001
Use of ≥2 antihypertensive agent	%	8	12	12	17	<0.0001
Antiplatelet therapy	%	69	71	75	75	<0.0001
Oral anticoagulants	%	8	9	9	10	0.049
Lipid-lowering therapy	%	11	14	16	15	0.0002
Study treatment regimen
Active therapy	%	52	48	50	50	0.50
Combination therapy or double placebos	%	50	56	61	66	<0.0001

^*Participants recruited from People’s Republic of China or Japan,

^†History of myocardial infarction or coronary revascularization, or angina (supported by documented electrocardiographic or angiographic evidence).

Effects of perindopril based therapy on cardiovascular events according to baseline BMI

During follow-up, the overall mean difference in SBP/DBP between participants assigned active therapy or placebo was 9/4 (SE, 0.5/0.3) mmHg, with no difference between BMI quarters (p for trend = 0.98 and 0.57, respectively for SBP and DBP). The perindopril-based blood pressure lowering regimen produced similar reductions in the risk of major vascular events at all levels of BMI (p for homogeneity = 0.16) (Figure 1). A similar pattern was observed for total stroke (Figure 1), although there was some evidence of a trend towards greater benefit with increasing BMI (p for homogeneity = 0.07 for total stroke and p = 0.10 for stroke subtypes). There was no difference in treatment effect at all levels of BMI when the population was separated in those receiving the combination active therapy (p for homogeneity = 0.44 for major vascular events and p = 0.40 for total stroke) or single (Perindopril) therapy (p for homogeneity = 0.60 for major vascular events and p = 0.26 for total stroke).

Figure 1.

Effects of active therapy compared with placebo on the risk of major vascular events, total stroke and stroke subtypes according to baseline body mass index quartiles (<23.1, 23.1-25.3, 25.4-27.8 and ≥ 27.9 kg/m²)

Treatment effects in subgroups are standardized for the proportions of the study population receiving combination (58%) or single-drug therapy (42%). Mean difference in blood pressure over time between randomized groups was calculated by subtracting the values for the placebo group from those of the active group in each BMI category.

Employing the usual WHO cut points for normal weight, overweight and obesity produced the same conclusions (all p for homogeneity ≥ 0.44, Figure 2). Likewise, fitting BMI as a continuous variable, provided no evidence that the effect of treatment varied by baseline level of BMI (all p for homogeneity ≥ 0.16).

Figure 2.

Effects of active therapy compared with placebo on the risk of major vascular events and total stroke according to baseline body mass index in three categories (BMI < 25, 25 to 29 and ≥ 30 kg/m²)

Treatment effects in subgroups are standardized for the proportions of the study population receiving combination (58%) or single-drug therapy (42%). Mean difference in blood pressure over time between randomized groups was calculated by subtracting the values for the placebo group from those of the active group in each BMI category.

Compared to participants in the lowest BMI quartile, absolute risk reductions for both major cardiovascular events and stroke were more than twice as great in participants in the highest BMI quartile (Table 2). The number needed to treat (NNT) for 5 years was accordingly smaller in the highest quartile, with one major vascular event prevented among every 13 patients treated over five years compared to 28 patients in the lowest quartile. The corresponding figures for total stroke were 13 and 42 patients treated over the same duration, respectively.

Table 2.

Incidence rate and absolute risk reduction over 5 years for major vascular events and total stroke according to BMI category (kg/m²) in PROGRESS

Events	5-year cumulative incidence (%)*		NNT for 5 yr
	Active	Placebo
Major vascular events
<23.1	20.1	23.7	28
23.1-25.3	17.9	22.2	23
25.4-27.8	18.4	25.9	13
≥ 27.9	18.8	26.7	13
Total stroke
<23.1	14.6	17.0	42
23.1-25.3	12.2	15.8	28
25.4-27.8	12.7	17.4	21
≥ 27.9	11.3	19.0	13

^*Cumulative incidences were corrected from 3.9 years to 5 years, assuming constant risk per year. NNT: number needed to treat.

Discussion

In this large randomized trial of participants with a history of cerebrovascular disease, BP-lowering treatment produced similar reductions in the risks of major CV events and stroke in normal weight, overweight and obese participants. However, the absolute risk reduction was almost twice as great in the obese group with a correspondingly lower number of patients needing to be treated to prevent one event.

The observed greater absolute risk reduction in those with the highest BMI is unsurprising given that they exhibited the highest prevalence of antihypertensive drug use, but also the highest blood pressure levels at baseline. For instance, data from the Framingham study and other cohorts^20;21 showed that, in study members with hypertension, obese participants were more likely to receive antihypertensive treatment than individuals with normal weight but hypertension was not better controlled.

These findings have important implications in the management of hypertension in the obese since it suggests that, for a given blood pressure reduction, the NNT will be smaller in preventing cardiovascular disease in this group. Our results, based on a population recruited in 1995-1997 with a rather low prevalence of obesity (12.2%), also provide further impetus for the allocation of blood pressure lowering agents on the basis of overall vascular risk,^22;23 rather than blood pressure level alone, and identify the obese as another population that warrants specific attention.

Current guidelines for the management of hypertension do not provide specific recommendations for the pharmacological treatment of high blood pressure in overweight or obese patients.^7;8 This reflects the limited data available about the effects of treatment in the hypertensive obese and the few prior analyses designed to address this question.^9-15;24;25 Indeed, even if the obvious treatment for the management of obesity-associated hypertension is weight-loss, the expected drop in blood pressure for every kilogram loss, is about 1 mm Hg;⁵ or even less in studies with a follow-up of two years or more.²⁶ Data from the Swedish Obese Study (SOS)²⁷ also indicated that after 10 years of follow-up, no difference was observed in the incidence of hypertension between patients who underwent bariatric surgery, compare to those enrolled to a conventional nutritional management. While weight-loss drugs, such as sibutramine, were developed for the management of weight in obese patients, they tend to increase blood pressure by 1 to 2 mm Hg, and present an additional challenge to the management of obesity-related hypertension.^28;29

Perspectives

Results of our study, involving patients with a history of cerebrovascular disease have shown that patients allocated to receive a perindopril-based BP lowering therapy experienced an overall risk reduction of 26% in major vascular events and of 28% in total stroke, with consistent reductions in risk across the full range of BMI, in the absence of evidence to support an interaction between BMI and the effect of blood pressure lowering. The greater absolute effect of treatment in those with higher body mass index appears to be a consequence of their greater baseline level of risk. These findings support the use of blood pressure lowering in the obese but there remain outstanding uncertainties. For example, while all commonly-used BP lowering drugs have been shown to protect against the risk of major cardiovascular events in the general population,³⁰ it remains unclear whether the effects of all drug classes are the same in the obese and non-obese individuals.^8;31 A recent statement of the Working Group on Obesity of the European Society of Hypertension has highlighted the need for more data to address this issue.³²