The prognostic significance, interdependence, and hierarchy of cardiovascular risk factors could evolve with advancing age. Our study reports on the interdependence among blood pressure (BP), other metabolic syndrome components, and high-sensitivity C-reactive protein according to age in hypertensive subjects. A total of 5,712 nondiabetic patients (50.1% men, age range 40 to 95 years) evaluated in outpatient hypertension clinics were included and divided into 5 age groups (age 40 to 49, 50 to 59, 60 to 69, 70 to 79, and >80 years). BP, evaluated by both office and 24-hour ambulatory BP monitoring, and the metabolic and inflammation parameters were determined after a ≥2-week drug washout period. The prevalence of the metabolic syndrome (Adult Treatment Panel III definition) remained stable across the age groups. We observed a stable or increased association between waist circumference and insulin resistance (Homeostasis Model of Assessment–Insulin Resistance index) and fasting plasma glucose. However, the association between waist circumference and ambulatory BP monitoring systolic BP ( r 2 decrease from 9.9% to 1.0%, p <0.001), high-density lipoprotein cholesterol ( r 2 decreased from 21% to 4.9%, p = 0.002), and triglyceride levels ( r 2 decreased from 17.5% to 1.9%, p <0.001) decreased with age. High-sensitivity C-reactive protein correlated with all metabolic syndrome components in all age groups (p <0.001 for all). It became the strongest determinant of ambulatory BP monitoring systolic BP (p <0.001) and high-density lipoprotein cholesterol (p <0.05) in patients >80 years old. In contrast, its association with waist circumference markedly decreased. In conclusion, hypertension and dyslipidemia, but not fasting plasma glucose, dissociate from central obesity with advancing age. They are increasingly determined by low-grade inflammation, independently of central obesity. These changing associations might underlie the weakening of obesity as a cardiovascular risk factor in older persons.
The prognostic significance, interdependence, and hierarchy of cardiovascular risk factors might evolve with advancing age. The study of the interdependence of these risk factors, in particular the metabolic syndrome (MS) components, is compounded because they are usually considered as binary variables and defined by specific cutoffs. Although this approach is convenient in clinical practice, it fails to take into account that the MS components, not only coexist, but also correlate in severity. Studies in which the MS components were analyzed as continuous variables have suggested that the blood pressure (BP), fasting plasma glucose, high-density lipoprotein (HDL) cholesterol, and triglycerides levels correlate with the body mass index or insulin resistance. However, the associations between these risk factors may be overestimated (and the lessening of these associations underestimated) by analyses of the binary variables, in particular, when low cutoffs (using the recent MS definitions) are used. Another important issue is the stability and reproducibility of the risk factor measurements. This is particularly problematic with BP. Office BP values are much more variable and higher than ambulatory BP monitoring (ABPM) values. However, the latter are more strongly associated with the risk of cardiovascular events and should therefore be used to study the interdependence of the MS components (whether analyzed as binary or continuous variables). The aim of the present study was to determine the associations among BP, other MS components, and low-grade inflammation and their evolution with age in a large cohort of patients with office hypertension.
Methods
The hypertension outpatient clinics of the “Hippokration” and “Onassis” Hospitals (Athens, Greece) offer comprehensive hemodynamic and metabolic evaluations to hypertensive patients, usually referred by their general practitioners. The detailed evaluation procedures used have been previously reported. At the initial visit, a medical history and physical examination (including height, weight, and waist and hip circumference measurements) were performed in all subjects. All patients receiving therapy for arterial hypertension or dyslipidemia underwent a complete washout period of ≥15 days.
The office BP, measured with mercury sphygmomanometers, was recorded at the second visit as the mean of 3 values, taken with the patient seated, 1 minute apart. Arterial hypertension was defined as the elevation of either systolic or diastolic BP >140/90 mm Hg or the use of antihypertensive drugs for ≥6 months. All patients underwent 24-hour ABPM using the nondominant arm with a Spacelabs 90207 device (SpaceLabs, Redmond, Washington). The recorder was set to take readings at 20-minute intervals from 6:00 a.m. to 10:00 p.m. and every 30 minutes from 10:00 p.m. to 6:00 a.m. The recording was analyzed to obtain the 24-hour average systolic and diastolic BP values. The total, low-density lipoprotein, and HDL cholesterol, triglycerides, fasting plasma glucose, insulin levels, and high-sensitivity C-reactive protein (hs-CRP) were measured using blood serum or plasma collected after the patient had fasted overnight. Insulin resistance was estimated using the Homeostasis Model of Assessment–Insulin Resistance index. The MS was defined according to the Adult Treatment Panel III criteria.
For the present study, only nondiabetic patients with office hypertension (office systolic BP >140 mm Hg and/or diastolic BP >90 mm Hg), for whom an ABPM, hs-CRP, fasting plasma glucose, and insulin levels were available, were included. The population was divided into 5 groups by age decade (40 to 49, 50 to 59, 60 to 69, 70 to 79, and ≥80 years). The baseline parameters were compared among the age groups using analysis of variance or chi-square tests, as appropriate. The normality of the parameters was assessed by histograms and Q–Q plots; in the case of significant deviation from a normal distribution, a log transformation was applied.
The age-related progression of the associations between the ABPM systolic BP, waist circumference, and hs-CRP with the MS components was assessed by multiple linear regression models, including the age variable (categorical data age groups), the independent variable tested, and their interactions (each time-independent variable tested × age group interaction). The independent determinants of the MS components were identified and compared in patients <60 years old and >80 years old by stepwise backward multiple linear regression models, including the following variables: ABPM systolic BP, waist circumference, HDL cholesterol, log_triglycerides, gender, fasting plasma glucose, log_hs-CRP, and smoking. Each was tested with the 7 others as independent variables. To allow comparisons between the relative effects of each determinant on the dependent variable, the results of the multiple regression analyses are expressed equally for a change from the 5th to the 95th percentile. Significance was assumed at a level of p <0.05. The results for ABPM-systolic BP and HDL cholesterol are shown. All analyses were performed using the Statistical Package for Social Sciences, version 13.0 (SPSS, Chicago, Illinois).
Results
The final study population included 5,712 nondiabetic patients with office hypertension (50.1% men). Their general characteristics according to age group are listed in Table 1 . As expected, the ABPM mean values were lower than the office BP values, with an age-related increase in the systolic BP and decrease in the diastolic BP for both office and ABPM measurements. We observed a progressive decrease in the body mass index and minor (although statistically significant) changes in the waist circumference and lipid levels. Fasting plasma glucose, but not the Homeostasis Model of Assessment–Insulin Resistance values, significantly increased with age. The hs-CRP values also significantly increased with age. Overall, the prevalence of the MS remained quite stable with advancing age ( Table 2 ). The age-related increase in the prevalence of a high waist circumference and fasting plasma glucose was counterbalanced by a decreasing prevalence of dyslipidemia.
| Variable | Age Group (years) | p Value | ||||
|---|---|---|---|---|---|---|
| 40–49 | 50–59 | 60–69 | 70–79 | ≥80 | ||
| Patients (n) | 1,119 | 1862 | 1559 | 895 | 277 | |
| Men | 56.7% | 50.9% | 47.9% | 46.7% | 40.8% | <0.001 |
| Office blood pressure (mm Hg) | ||||||
| Systolic | 161 ± 11 | 165 ± 10 | 168 ± 8 | 171 ± 7 | 177 ± 8 | <0.001 |
| Diastolic | 105 ± 5 | 102 ± 6 | 98 ± 7 | 92 ± 7 | 86 ± 7 | <0.001 |
| Ambulatory blood pressure monitoring (mm Hg) | ||||||
| Systolic | 134 ± 12 | 134 ± 11 | 135 ± 10 | 137 ± 10 | 140 ± 11 | <0.001 |
| Diastolic | 86 ± 8 | 84 ± 7 | 81 ± 7 | 77 ± 6 | 75 ± 6 | <0.001 |
| Never treated patients | 68.5% | 59.9% | 47.9% | 41.1% | 40.8% | <0.001 |
| Waist circumference (cm) | 95.0 ± 14.2 | 95.5 ± 12.8 | 94.8 ± 13.1 | 94.0 ± 11.8 | 94.0 ± 11.9 | 0.037 |
| Body mass index (kg/m 2 ) | 28.5 ± 4.9 | 28.7 ± 4.4 | 28.3 ± 4.5 | 27.6 ± 4.2 | 27.3 ± 4.5 | <0.001 |
| Body mass index ≥30 kg/m 2 | 31.5% | 35.1% | 32.9% | 25.7% | 22.0% | <0.001 |
| Total cholesterol (mg/dl) | 213 ± 38 | 215 ± 37 | 216 ± 38 | 210 ± 38 | 205 ± 34 | <0.001 |
| High-density lipoprotein cholesterol (mg/dl) | 48.6 ± 12.1 | 50.8 ± 12.8 | 52.6 ± 13.4 | 53.5 ± 13.9 | 51.9 ± 13.6 | <0.001 |
| Low-density lipoprotein cholesterol (mg/dl) | 139.1 ± 34.4 | 139.5 ± 33.2 | 139.4 ± 33.9 | 133.8 ± 34.9 | 128.3 ± 31.1 | <0.001 |
| Triglycerides (mg/dl) | 126.4 ± 65.6 | 124.1 ± 61.9 | 119.3 ± 51.3 | 115.5 ± 52.0 | 124.0 ± 57.3 | 0.006 |
| Fasting plasma glucose (mg/dl) | 94.5 ± 12.3 | 97.0 ± 12.3 | 99.2 ± 13.2 | 99.0 ± 13.7 | 100.5 ± 13.8 | <0.001 |
| Homeostasis Model of Assessment–Insulin Resistance | 2.31 ± 1.12 | 2.34 ± 1.07 | 2.37 ± 1.11 | 2.32 ± 0.94 | 2.37 ± 0.97 | 0.6 |
| High-sensitivity C-reactive protein (mg/dl) | 1.33 ± 1.08 | 1.43 ± 1.14 | 1.44 ± 1.10 | 1.49 ± 1.07 | 1.75 ± 1.13 | <0.001 |
| Smokers | 49.1% | 40.2% | 31.6% | 22.3% | 18.4% | <0.001 |
| Variable | Age Group (years) | p Value | ||||
|---|---|---|---|---|---|---|
| 40–49 | 50–59 | 60–69 | 70–79 | ≥80 | ||
| High blood pressure ⁎ | 100% | 100% | 100% | 100% | 100% | |
| High waist circumference ⁎ | 41.6% | 47.1% | 47.3% | 45.0% | 49.1% | 0.016 |
| Low high-density lipoprotein cholesterol ⁎ | 35.8% | 32.4% | 28.0% | 28.2% | 30.7% | <0.001 |
| High triglycerides ⁎ | 29.1% | 26.7% | 25.1% | 23.9% | 26.0% | 0.071 |
| High fasting plasma glucose ⁎ | 11.7% | 15.4% | 21.4% | 21.9% | 26.0% | <0.001 |
| Metabolic syndrome components (n) | 2.18 ± 1.13 | 2.22 ± 1.13 | 2.22 ± 1.15 | 2.19 ± 1.16 | 2.32 ± 1.17 | 0.476 |
| Metabolic syndrome | 35.1% | 36.2% | 35.9% | 34.7% | 38.6% | 0.782 |
⁎ Defined according to Adult Treatment Panel III (2001) criteria.
The interdependence between the risk factors and the interaction with age were studied using multiple linear regression models, with each risk factor as a continuous variable ( Table 3 ). We observed a strong, significant association between the ABPM systolic BP and waist circumference and weaker associations with HDL cholesterol, triglycerides, fasting plasma glucose, and Homeostasis Model of Assessment–Insulin Resistance in the younger age groups. A gradual decrease was seen with advancing age in the association between the ABPM systolic BP and waist circumference (p <0.001 for interaction) that was not seen with the other MS components ( Figure 1 and Table 3 ). The relatively strong association between the ABPM systolic BP and hs-CRP apparently increased with age, although the age effect failed to reach statistical significance. The waist circumference, used as the dependent variable, was quite strongly associated with the ABPM systolic BP, HDL cholesterol, triglycerides, and fasting plasma glucose in the younger age groups ( Table 3 ). These associations were stronger using the waist circumference than using the body mass index (data not shown). The association between the waist circumference and Homeostasis Model of Assessment–Insulin Resistance and fasting plasma glucose remained stable or increased with age. In contrast, the association with ABPM systolic BP, HDL cholesterol, and triglycerides disappeared or markedly decreased with age. Thus, the BP and lipid levels (but not fasting plasma glucose) were dissociated from obesity with advancing age.
| Age Group (years) | p Value | Model R 2 | ||||||
|---|---|---|---|---|---|---|---|---|
| 40–49 (Referent) | 50–59 | 60–69 | 70–79 | ≥80 | Overall Effect | Interaction | ||
| Ambulatory blood pressure monitoring-systolic blood pressure | ||||||||
| Waist circumference | 0.375 | 0.356 | 0.352 | 0.188 ⁎ | 0.011 ⁎ | <0.001 | <0.001 | 7.9 |
| 9.9 | 9.6 | 7.6 | 2.4 | 0.01 | ||||
| High-density lipoprotein cholesterol | −0.23 | −0.23 | −0.29 | −0.28 | −0.18 | <0.001 | 0.48 | 5.9 |
| 5.3 | 4 | 4.8 | 4 | 2.1 | ||||
| Log_triglycerides | 0.004 | 0.003 | 0.003 | 0.003 | 0.004 | <0.001 | 0.31 | 2.9 |
| 4.2 | 2.1 | 1.7 | 2.3 | 4.2 | ||||
| Fasting plasma glucose | 0.145 | 0.159 | 0.220 | 0.213 | 0.042 | <0.001 | 0.11 | 4.2 |
| 2 | 2.1 | 2.9 | 2.3 | 0.1 | ||||
| Homeostasis Model of Assessment–Insulin | 0.022 | 0.02 | 0.028 | 0.024 | 0.017 | <0.001 | 0.14 | 5.5 |
| Resistance | 5.5 | 4.2 | 6.6 | 6.2 | 3.5 | |||
| Log_high-sensitivity C-reactive protein | 0.009 | 0.008 | 0.011 | 0.008 | 0.012 | <0.001 | 0.06 | 8.7 |
| 8.1 | 6.5 | 9.8 | 5.5 | 12.8 | ||||
| Waist circumference | ||||||||
| Ambulatory blood pressure monitoring- | 0.264 | 0.269 | 0.216 | 0.130 ⁎ | 0.009 ⁎ | <0.001 | <0.001 | 9.7 |
| systolic blood pressure | 9.9 | 9.6 | 7.6 | 2.4 | 0.01 | |||
| High-density lipoprotein cholesterol | −0.386 | −0.473 † | −0.441 | −0.392 | −0.251 ‡ | <0.001 | 0.002 | 19.4 |
| 20.5 | 22.6 | 18.6 | 11.1 | 4.9 | ||||
| Log_triglycerides | 0.006 | 0.005 ‡ | 0.004 ⁎ | 0.004 ‡ | 0.002 ⁎ | <0.001 | <0.001 | 10.9 |
| 17.5 | 10.4 | 7.6 | 8.1 | 1.9 | ||||
| Fasting plasma glucose | 0.226 | 0.298 ‡ | 0.297 ‡ | 0.312 ‡ | 0.425 † | <0.001 | 0.023 | 10.5 |
| 6.8 | 9.6 | 8.6 | 7.2 | 13.4 | ||||
| Homeostasis Model of Assessment–Insulin | 0.038 | 0.038 | 0.036 | 0.032 | 0.043 | <0.001 | 0.205 | 20.7 |
| Resistance | 23.7 | 21.4 | 18.2 | 16.8 | 28.9 | |||
| Log_high-sensitivity C-reactive protein | 0.009 | 0.01 | 0.009 | 0.006 † | 0.005 ‡ | <0.001 | 0.001 | 10.4 |
| 11.1 | 12.1 | 9.6 | 3.8 | 2.5 | ||||
| Log_high-sensitivity C-reactive protein | ||||||||
| Waist circumference | 12.2 | 12 | 11.3 | 6.5 ⁎ | 5.2 † | <0.001 | <0.001 | 9.9 |
| 11.1 | 12.1 | 9.6 | 3.8 | 2.5 | ||||
| Ambulatory blood pressure monitoring- | 8.73 | 7.64 | 8.56 | 6.5 | 10.29 | <0.001 | 0.182 | 9.8 |
| systolic blood pressure | 8.1 | 6.5 | 9.8 | 5.5 | 12.8 | |||
| High-density lipoprotein cholesterol | −7.43 | −9.65 | −8.24 | −8.51 | −6.91 | <0.001 | 0.403 | 7.5 |
| 5.8 | 7.9 | 5.4 | 4.7 | 3.4 | ||||
| Log_triglycerides | 0.11 | 0.12 | 0.1 | 0.11 | 0.13 | <0.001 | 0.918 | 5.0 |
| 4.3 | 4.9 | 4.6 | 4.4 | 6.8 | ||||
| Fasting plasma glucose | 6.19 | 5.35 | 7.93 | 7.44 | 8.56 | <0.001 | 0.164 | 5.9 |
| 3.8 | 5.6 | 5.1 | 3.7 | 5.0 | ||||
| Homeostasis Model of Assessment–Insulin | 0.96 | 0.88 | 0.97 | 0.74 | 1.04 | <0.001 | 0.284 | 10.3 |
| Resistance | 11.1 | 9.5 | 10.9 | 7.7 | 14.8 | |||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
