In a separate analysis of patients enrolled in the Losartan Intervention For Endpoint reduction in hypertension (LIFE) study, thin patients and those with moderate-to-severe obesity were at increased risk for cardiovascular events. These results indicate that patients at the low anda high ends of the body size spectrum should be treated aggressively to lower the risk of cardiovascular events.
As body weight increases, so does hypertension, resulting from greater systemic resistance and blood volume due to hemodynamic and metabolic dysfunction. Obesity amplifies cardiovascular risk because multiple cardiovascular risk factors coexist in obese individuals. At the opposite end of range of body size, decreased body weight also carries greater cardiovascular risk because it is often linked to serious disease.
Because it was not known if increased weight raises cardiovascular risk in patients with target organ damage resulting from hypertension independently of other factors, we evaluated the influence of body mass index (BMI) on outcomes in the Losartan Intervention For Endpoint reduction in hypertension (LIFE) study, which included patients with hypertension and established left ventricular hypertrophy (LVH).1,2
Patients and methods
Details of the protocol, patient characteristics, and main results of the LIFE study have been previously reported.1,2 All participants had LVH as shown by electrocardiographic (ECG) criteria and a blood pressure of ≥ 160-200/90-115 mm Hg after a 2-week placebo run-in period. Patients were randomly assigned to receive the angiotensin receptor blocker (ARB) losartan (Cozaar) or the beta blocker atenolol (Tenormin), to which hydrochlorothiazide or other antihypertensive therapies (excluding angiotensin-converting enzyme inhibitors, ARBs, and beta blockers) could be added with the goal of achieving a target blood pressure of < 140/90 mm Hg. Patients were followed for a mean of 4.8 years.
The present analysis was carried out in 9079 patients from the original sample of 9193 participants for whom baseline BMI data were available. Patients were divided into the following 6 groups according to BMI, based on 1998 National Institutes of Health guidelines: thin (< 20 kg/m2, 2% of patients); normal weight (20-24.9 kg/m2, 24% of patients); overweight (25-29.9 kg/m2, 45% of patients); and obese (class I: 30-34.9 kg/m2, 21% of patients; class II: 35-39.9 kg/m2, 6% of patients; class III: > 40 kg/m2, 2% of patients). We evaluated the baseline patient characteristics according to BMI category and the effect of BMI category on the primary composite end point of cardiovascular death, stroke, and myocardial infarction (MI; n = 1081), as well as cardiovascular death (n = 432) and new-onset diabetes (n = 554).
Being overweight existed among races in the following distribution: whites, 45%; blacks, 37%; and other ethnic groups, 47%. Among those with class I obesity, 21% were white, 25% were black, and 17% were in other ethnic groups. For those with class II obesity, 5% were white, 10% were black, and 8% were in other ethnic groups. For those with class III obesity, 2% were white, 6% were black, and 4% were in other ethnic groups. Forty-eight percent of overweight participants were women. The distribution among obese patients according to sex was as follows: class I, 59% women, 41% men; class II, 69% women, 31% men; and class III, 79% women, 21% men.
Obese patients had higher diastolic blood pressures than the other groups, and thin patients had lower diastolic blood pressures, although mean systolic blood pressures were similar for all groups (all < .001). In addition, pulse pressure was highest in thin patients < .001). The highest heart rates were also present in thin patients and in those with class II and III obesity (all < .001).
At the start of the study, patients with class III obesity had the lowest prevalence of ischemic heart disease (20%), followed by those with class I obesity (22%), overweight (25%), and class II obesity (25%); thin patients had the highest prevalence of ischemic heart disease (33%; < .001). Thin patients were also older than the other patients, with a mean age of 70.2 years, and those with class II and III obesity were younger than the other groups (all < .001).
The prevalence of diabetes increased according to an increase in weight. At the start of the study, 5% of thin patients, 7% of those with normal weight, 12% of overweight patients, 15% of those with class I obesity, 24% of those with class II obesity, and 34% of those with III obesity had diabetes ( < .001 for trend).
Thin and normal-weight patients had the highest levels of high-density lipoprotein (HDL) cholesterol, and patients with class II and III obesity had lower total and HDL cholesterol levels. Thin and normal-weight pa­tients also had the lowest levels of plasma glucose, which increased in a stepwise fashion according to weight (all < .001).
Thin patients and those who were in the highest obesity classes had the lowest creatinine clearance ( < .001). Thin patients and those with class III obesity also had significantly higher urinary albumin-to-creatinine ratios ( < .001).
Only 11% of those with class II obesity smoked compared with 12% of patients with class III obesity, 13% of patients with class I obesity, 15% of overweight patients, 24% of normal-weight patients, and 43% of thin patients ( < .001).
The Cornell voltage-duration product was lowest in thin patients and increased according to weight ( < .001). There was an inverse relationship between the Sokolow-Lyon voltage and BMI.
For all groups of body build, the event rates for the primary end point were similar: thin patients, 17%; normal-weight patients, 12%; overweight patients, 12%; patients with class I obesity, 10%; patients with class II obesity, 12%; and patients with class III obesity, 11%.
Compared with normal-weight patients, overweight patients had a 17% greater risk of the primary end point, after adjustment for covariates. The combined class II and III obesity groups had a 35% greater primary event rate. In addition, there was a correlation between prevalent cardiovascular disease, older age, current smoking, ECG indices of LVH, diabetes, and male sex (
Thin patients had the greatest risk of cardiovascular death (10%), compared with the other groups (5% in normal-weight patients, overweight patients, and those with class II obesity; 4% in patients with class I obesity; and 2% in patients with class III obesity; = .09). After adjusting for covariates, the risk of cardiovascular death was higher among thin patients ( < .05) and patients with class II and III obesity ( < .004) than among those with normal body build (
). There was also an association between cardiovascular death and ECG indices of LVH, black race, male sex, prevalent cardiovascular disease, current smoking, older age, and diabetes.
Drug treatment did not significantly alter the greater risks related to body build. Independent of body build and baseline characteristics, losartan decreased the risk of the composite of cardiovascular mortality, stroke, and MI by 15% compared with atenolol.
The proportion of patients who developed diabetes over the course of the study (n = 7899) was as follows: thin patients, 3%; normal-weight patients, < 1%; overweight patients, 6%; class I obesity patients, 12%; class II obesity patients, 15%; and class III obesity patients, 23% ( = .001 for trend). Among patients of all weights, the prevalence of new-onset diabetes was somewhat lower for those treated with losartan than for those treated with atenolol (
Obesity often carries numerous major risk factors for adverse cardiovascular events.3
It is also associated with cardiovascular disease independently of coexisting risk factors.4 The potential mechanisms for increased cardiovascular risk in patients with obesity are under investigation. Obesity contributes to increased cardiac workload and therefore to the development of LVH.5 It is also correlated with higher fibrinogen levels,6 greater vascular thromboxane receptor gene expression,7 and the triggering of inflammation mechanisms.8 Central fat distribution, which is almost constant in class II and III obesity,9 further increases cardiovascular risk and is related to more target organ damage than peripheral body fat distribution.10
In the present study, there was no correlation between greater cardiovascular risk and class I obesity. It is likely that the potential adverse effects of overweight and mild obesity in the context of already established hypertensive target organ damage depend on the associated risk factors, which is consistent with the findings of epidemiologic studies in unselected populations.11 In more severe obesity, biologic mechanisms associated with central adiposity increase cardiovascular risk beyond that predictable from cardiovascular risk factors or the presence of LVH. In our analysis, thin patients had a high risk of cardiovascular death that was similar to patients with severe obesity. Thin patients had more risk factors and prevalent cardiovascular disease, and their low BMI was probably a marker of the severity of cardiovascular impairment. Other studies have shown a correlation between cardiovascular events and low body weight.11
Renal dysfunction was a characteristic common to the lowest and highest BMI groups, whereas renal function was not substantially different among the other body size strata. Because thin patients had as much risk of cardiovascular death as patients with moderate to severe obesity, renal impairment may be either the marker or the cause of more accelerated arterial dysfunction.12
Stratification for BMI did not significantly affect differences between drug treatments. The superiority of treatment with losartan compared with atenolol for the primary end point2 and new-onset diabetes13 seen in the main studies was also preserved across the body size strata.
The risk of cardiovascular events in patients who were underweight or moderately to severely obese was shown to be increased among LIFE study participants. The body build stratification did not interfere with the benefits of losartan therapy. The risk of new-onset diabetes was also shown to be lower in patients treated with losartan, independently of body size stratification. Risk factors in patients at the low and high ends of the body size spectrum should be treated aggressively to decrease the incidence of cardiovascular death.