P. Palatini

Heart rate as a predictor of development of sustained hypertension in subjects screened for stage 1 hypertension: the HARVEST Study.

Objective Whether heart rate predicts the development of sustained hypertension in individuals with hypertension is not well known. We carried out a prospective study to investigate whether clinic and ambulatory heart rates assessed at baseline and changes in clinic heart rate during 6 months of follow-up were independent predictors of subsequent blood pressure (BP). Methods The study was conducted in a cohort of 1103 white, stage 1 hypertensive individuals from the HARVEST study, never treated for hypertension and followed-up for an average of 6.4 years. Data were adjusted for baseline BP, age, sex, body fatness, physical activity habits, parental hypertension, duration of hypertension, cigarette smoking, alcohol consumption, and change of body weight from baseline. Results Clinic heart rate and heart rate changes during the first 6 months of follow-up were independent predictors of subsequent systolic blood pressure (SBP) and diastolic blood pressure (DBP) regardless of initial BP and other confounders (all P < 0.01). A significant interaction was found between sex (male) and baseline resting heart rate on final SBP (P = 0.017) and DBP (P < 0.001). The ambulatory heart rate and the heart rate white-coat effect did not add prognostic information to that provided by the clinic heart rate. Patients whose heart rate was persistently elevated during the study had a doubled fully adjusted risk (95% confidence interval 1.4–2.9) of developing sustained hypertension in comparison with subjects with a normal heart rate. Conclusions Baseline clinic heart rate and heart rate changes during the first few months of follow-up are independent predictors of the development of sustained hypertension in young persons screened for stage 1 hypertension.

RECTANGULAR CUFFS OVERESTIMATE BLOOD PRESSURE IN OBESE PEOPLE WITH VERY LARGE ARMS

Objective: Rectangular cuffs and bladders are currently used for blood pressure (BP) measurement at the upper arm. However, large arms always have a tronco-conical shape. Aim of this study was to ascertain whether rectangular and tronco-conical cuffs provide different readings in obese subjects with very large arms. Design and method: In 33 subjects with morbid obesity (BMI, 45 ± 5.0 Kg/m2.16 men) aged 51 ± 12 years, with arm mid-circumference ranging from 42 to 55 cm, and 33 subjects of control with standard arm circumference (range, 22–31 cm) the upper-arm frustum slant angle was measured. Two different rectangular and two different tronco-conical bladders of appropriate size and shape were constructed, on the basis of previous anthropometric measures obtained in our laboratory. In each subject, BP was measured in triplicate by two observers using the two cuffs in a random order. In addition, in the obese participants, the pressure under the two cuffs was measured at five pressure levels (60, 90, 120, 150 and 180 mmHg) using a paper-thin pressure sensor attached to the central point of the cuffs. Results: In all obese participants the upper arm shape was tronco-conical with slant angles ranging from 80.4 to 87.6° (mean 84.1 ± 1.4°). Systolic BP (SBP) and diastolic BP (DBP) differences between the troncoconical and the rectangular cuff were −4.8 ± 4.0 and −3.0 ± 4.3 mmHg, respectively, whereas they were negligible in the controls (SBP, p < 0.001 and DBP, p = 0.01). In the subjects of the top SBP quintile (SBP > = 150 mmHg), the between-cuff SBP difference was −9.1 ± 5.1 mmHg. Arm slant angle was an independent predictor of the between-cuff SBP discrepancy (p = 0.003). Measurement with the pressure sensor showed a higher pressure under the rectangular compared to the conical cuff at any pressure level. The mean difference was −10.2 ± 5.2 mmHg and the difference progressively increased with increasing level of pressure applied to the cuffs. Conclusions: In obese people, the upper arm has a pronounced tronco-conical shape and rectangular cuffs may overestimate BP. Tronco-conical cuffs should be used for BP measurement in subjects with very large arms.

[BP.11.02] ALCOHOL INTAKE MORE THAN DOUBLES THE RISK OF EARLY CARDIOVASCULAR EVENTS IN YOUNG HYPERTENSIVE SMOKERS

Objective: Little information is available on the joint effect of smoking and alcohol on major adverse cardiovascular and renal events (MACE) in hypertension. Aim of the study was to quantify the interactive effect of smoking and alcohol intake on MACE in young subjects screened for stage 1 hypertension. Design and method: A total of 1204 untreated patients aged from 18 to 45 years (mean 33.1 ± 8.5) were included in this prospective cohort study. Subjects were classified into 4 classes of smoking and 3 classes of alcohol use. Main outcome variable was risk for MACE. A parsimonious set of covariates was selected using sequential backward elimination of the least significant variables. Subsequently, a final model was developed in which lifestyle factor category was entered as the first variable and then was adjusted for other risk factors and several confounders. Results: During a 12.6-year follow-up there were 74 fatal and nonfatal MACE. In multivariable Cox models there was a linear relationship of smoking and alcohol drinking with risk of MACE. Hazard ratio was 1.65 (95% CI, 1.35–2.02, p < 0.0001) for smoking and was 2.05 (95%CI, 1.22–3.43, p = 0.007) for alcohol use. In addition, a strong interactive effect was found between smoking and alcohol on risk of MACE (p < 0.0001). Among the 142 smokers who also drank alcoholic beverages the risk of MACE (4.02, 95%CI, 1.98–8.15) was more than doubled compared to the 112 smokers who abstained from drinking (1.64, 95%CI, 0.63–4.27). In the group of heavy smokers who also were alcohol drinkers (n = 51) the risk of MACE was even quadrupled. In this group, urinary catecholamines (mean ± SEM, 104.5 ± 13.3 mcg/g versus 70.9 ± 4.1 mcg/g, p = 0.01) and 24 h heart rate (mean ± SEM, 75.7 ± 1.1 versus 72.8 ± 0.2 bpm, p = 0.008) were increased compared with the rest of the cohort. Conclusions: Alcohol use strongly potentiates the deleterious cardiovascular effects of heavy smoking in stage 1 hypertensive subjects younger than 45 years. Increased sympatho-adrenergic activity may be a contributing factor to the synergistic effect of smoking and alcohol on MACE. These results call for prompt intervention addressed to improve unhealthy behaviours in these subjects.

[BP.11.05] URIC ACID AND HYPERTENSION. ROLE OF RENIN ANGIOTENSIN AND SYMPATHETIC NERVOUS SYSTEMS ACTIVITY

Objective: A number of studies have documented an association of serum uric acid (SUA) with risk of hypertension. We did a prospective study to investigate whether SUA was an independent predictor of risk of hypertension in the young and whether plasma renin activity (PRA) and office and ambulatory heart rates differed according to SUA tertile. Design and method: The study was conducted in a cohort of 1156 young-to-middle-age subjects screened for stage 1 hypertension and followed for a median of 11.4 years. Mean age was 33.0 ± 8.6 years and mean office blood pressure at screening was 145.3 ± 10.6/93.4 ± 5.8 mmHg. PRA was measured in 799 participants. In multivariable Cox analyses, data were adjusted for age, gender, body mass index, parental hypertension, 24-hour systolic and diastolic BP, serum creatinine, and lifestyle factors. Results: At follow-up end, 63.3% of study participants developed hypertension needing treatment. SUA was an independent predictor of future hypertension (p = 0.005). Participants with SUA >5.60 mg/dL (top tertile) had a 31% (95%CI, 6%-60%) increase in risk compared to those of the bottom tertile (SUA <4.32 mg/dL). Office heart rate, average 24-hour heart rate, and night-time heart rate significantly increased across SUA tertiles (p = 0.008, p = 0.033, and p = 0.007, respectively, for ANCOVA after adjusting for age, sex, and lifestyle factors). PRA adjusted for age, sex, smoking, coffee and alcohol use increased on going from the first to the third SUA tertile (mean ± SEM, 3.1 ± 0.5, 3.8 ± 0.4, and 4.9 ± 0.5 ng/ml/h, respectively, p = 0.038). Inclusion of physical activity in the model slightly attenuated the relationship (p = 0.044). PRA was lower in active than in sedentary participants (3.2 ± 0.7 versus 4.8 ± 0.4 ng/ml/h, p = 0.050) as were office and ambulatory heart rates (both p < 0.001). No association between SUA and risk of hypertension was found among the active participants (p = 0.71). Conclusions: In keeping with our previous results, these data confirm that SUA is an independent predictor of hypertension in young-to-middle-age subjects. This may be due to the unfavourable effect of SUA on several factors including sympathetic tone and renin-angiotensin-system activity. Regular physical activity can counteract these pathophysiological mechanisms and prevent the development of hypertension in hyperuricemic subjects.

[OP.8A.05] REGULAR PHYSICAL ACTIVITY ATTENUATES THE ASSOCIATION BETWEEN COFFEE CONSUMPTION AND CARDIOVASCULAR EVENTS IN HYPERTENSION. THE HARVEST.

Objective: Coffee consumption has been found to predict the risk of cardiovascular events (CVE) in hypertension. Aim of the present study was to ascertain whether the strength of the coffee-CVE association is modulated by regular physical activity. Design and method: The analysis was made in 1204 participants from the HARVEST, a prospective cohort study of non-diabetic subjects aged 18–45 years, screened for stage 1 hypertension. Among the participants, 26.2% were abstainers, 62.8% were moderate coffee drinkers (1–3 cups/day) and 10.0% were heavy coffee drinkers (>3 cups/day). For physical activity level, 61.1% were categorized as sedentary (class 0), 14.9% as mild exercisers (class 1), 14.9% as amateurs (class 2), and 9.1% as competitive athletes (class 3). Results: The frequency of coffee drinkers was higher among the sedentary (77.4%) than the active (67.9%) subjects (p < 0.001). During a 12.6 year follow-up 70 participants developed a CVE. The rate of CVE was higher among the coffee drinkers (7.1%) than the abstainers (2.2%, p = 0.001) and was higher in the sedentary (7.2%) than the active (3.6%) subjects (p = 0.01). In a 2x2 sub-table the coffee-related difference in CVE was significant among the sedentary (p = 0.006) but not the active subjects (p = 0.22) with a Mantel-Haenszel Chi-Square = 8.3 (p = 0.004). In a multivariable Cox analysis, including age, sex, parental CVE, smoking, alcohol use, total cholesterol, follow-up hypertension, and a coffee-physical activity interaction term, coffee consumption was a significant predictor of CVE in heavy coffee drinkers (Hazard ratio, 95%CI; 3.3, 1.2–8.9) and was a borderline predictor in moderate drinkers (HR, 95%CI; 2.0, 0.9–4.5). In addition, a borderline negative interaction was found between coffee consumption and physical activity habits on risk of CVE (p = 0.05). Conclusions: These data confirm that coffee use is associated with increased risk of CVE in young-to-middle age subjects with stage 1 hypertension. However, the effects of coffee on risk of CVE is modulated by the individual level of physical activity. Regular physical activity seems to counteract the long-term deleterious effect of caffeinated coffee on the cardiovascular system.

PREMENOPAUSAL WOMEN ARE AT HIGHER RISK OF HYPERTENSIVE COMPLICATIONS THAN MEN: PP.35.455

Objective: Little is known about whether hypertension has a different impact on target organs in young to middle-age women compared to men. The purpose of this study was to describe sex-specific differences in target organ involvement in a cohort of never treated hypertensive subjects followed for a median of 7 years. Design and Methods: Participants were 626 adults (451 men) aged 18 to 45 years screened for stage 1 hypertension. Ambulatory blood pressure (BP) at entry was 127.5 ± 12.5/83.7 ± 8.2 mmHg in women and 131.9 ± 10.3/81.0 ± 7.9 mmHg in men. Patients were seen every six months for BP and global risk assessment until they needed drug therapy according to current guidelines. Ambulatory BP, albuminuria, and echocardiographic data (n = 470) were obtained at entry, every 5 years, and/or before starting treatment. Data were adjusted for age, body mass, BP, physical activity, parental hypertension, smoking, coffee and alcohol use. Results: Female gender was a significant predictor of urinary albumin (p = 0.002) and left ventricular mass indexed to height (LVMI,p = 0.002) at final assessment. At follow-up end, microalbuminuria was more common among women than men (13.7% versus 6.2%, adjusted p = 0.001) as was left ventricular hypertrophy (LVH, 26.4% versus 8.8%, p < 0.0001). These differences remained significant also when adjusted for baseline urinary albumin or LVMI. In a multivariable Cox analysis, female gender was a significant predictor of time to development of microalbuminuria (p = 0.002) with a HR(95%CL) of 2.6(1.4–4.7), and of LVH (p = 0.01) with a HR of 2.1(1.2–3.8). After inclusion of BP changes over time in the models, HRs were 3.2(1.7–5.9) and 2.7(1–5–5.1), respectively. When baseline urinary albumin or LVMI were taken into account, the associations remained highly significant with HRs of 2.7(1.4–5.1) and 2.5(1.3–4.6), respectively. Conclusions: These data show that in young-to-middle-age hypertensive subjects the risk of target organ damage is much greater among women than men irrespective of the BP changes over time. This raises the question about whether early antihypertensive treatment should be considered for premenopausal women.