Hypervolemia for Hypertension Pathophysiology: A Population-Based Study

1. Introduction

Hypertension is one of the most important health problems in the world. The mortality and morbidity are related to the severity and duration; therefore, early diagnosis and treatment provide favorable clinical outcome. It has been proposed that abnormal ion transport by the kidney and subsequent disruption of body fluid volumes are responsible for the development of hypertension [1].

Etiology has been defined in only 5% of hypertensive subjects, while the essential hypertension is the classification in the majority of subjects meaning that no clear underlying cause can be found. Up to now obtaining data about body fluids have many difficulties; therefore, studies performed in this field were in small size and related to a specific disease group [2]. It is not known if these body fluid changes represent the cause or rather the effect of hypertension. Clarifying the relation of body fluid and blood pressure rise in a population study could offer new therapeutic approaches with remarkable impact on this affection.

In bioimpedance spectroscopy (BIS), complex impedance is measured over a wide range of frequencies and the data are fit to a well-known model of biological tissue. The extracellular and intracellular resistance are obtained by Cole model [3, 4]. Total body water (TBW) is the sum of ECW + ICW. The basis of BIS is that at low frequency (zero) current, there is no conduction through biological cells and only the ECW is measured and at high frequency (infinite), both the ECW and ICW are fully measured.

A new tool (FMC BCM) (based on plausible scientific principles) has been introduced to the medical field that allows the ECW and ICW (and overhydration) to be measured routinely with reasonable accuracy for the first time in history. This tool has revealed that significant overhydration (OH) is present in 30% of European HD and PD patients [5] and a strong predictor of LV hypertrophy and mortality [6, 7]. It has also been found that BP is a poor predictor of OH during heart failure [8]. The obvious next step is to study the general population with this new tool.

Aim of this epidemiologic study is to determine the assessment of body fluids and the estimation of the compartmental distribution in ICW and ECW via BIS in a healthy population and their relation to blood pressure. It is the first study to show the compartmental distribution defects with relation to blood pressure in such a large population.

2. Subjects and Methods

3. Results

Study cases were aged between 18 and 89, with the mean of 47 ± 12.6 years. Characteristics of the study population were given in Table 1.

Mean BCM measurement quality given by the device itself was 93.22 ± 4.6 percent; measurement duration was 69.2 ± 33.6 seconds.

Mean systolic blood pressure (SBP) was 134.7 ± 20.0, and diastolic blood pressure was (DBP) 77.9 ± 11.6 mmHg. BMI was 28.5 ± 4.5 (15.8–50.6) kg/m², body surface area (BSA) was 1.89 ± 0.17 (1.33–2.47) m², OH (L) was 0.05 ± 1.05 L, OH/ECW (%) was 0.01 ± 5.94, ECW/height was 10.46 ± 1.36 L/m, TBW (L) was 39.9 ± 6.6, ECW (L) was 17.6 ± 2.79, ICW (L) was 22.3 ± 4.15, LTI (kg/m²) was 16.3 ± 2.85, FTI (kg/m²) was 11.9 ± 5.3, adipose tissue mass (ATM) was 32.82 ± 13.32 kg, lean tissue mass (LTM) was 46.44 ± 10.63 kg, and body weight was 79.2 ± 13 kg.

There was a weak positive correlation between ECW/height and SBP (r = 0.21, P < 0.001). There were a strong correlation between BMI and FTI (r = 0.82, P < 0.001), correlation even more pronounced among females (r = 0.93, P < 0.001), whereas no correlation between BMI and LTI in both genders but a weak positive correlation in male group (r = 0.16, P < 0.001). There was a negative correlation between eGFR and ECW/Height (P = 0.024, r = −0.06).

Age, BMI, ECW/height, and BSA increased significantly in normotensives than hypotensives and even increased in hypertensives than in both normotensives and hypotensives. FTI was higher in hypertensives than in both normotensives and hypotensives. Smoking rate was lower in hypertensives compared to normotensives (Table 2).

There was a statistical significance in relation to diabetes, smoking habits, and measurement time (whether morning or afternoon) in relation to SBP. In 52.5% of diabetics and 25.8% of cigarette smokers, 32.1% of afternoon measurements had SBP ≥ 140 mmHg, and in 68.2% of nondiabetics and 64.9% of nonsmokers, 61.2% of morning measurements had SBP < 140 mmHg (P < 0.05).

Risk factors for high SBP were increase of ECW/height, age, and BMI and presence of diabetes (Table 3). ECW/height 7.71 ± 0.87, 10.1 ± 1.17, 11.32 ± 1.28 L/m, FTI 4 ± 2.2, 9.5 ± 3.5, 17.1 ± 4.7 kg/m² and SBP 120 ± 17, 132 ± 19, 140 ± 21 mmHg, increased in as BMI increased (low, normal and obese), respectively, (P < 0.001) (Figure 1). ECW/height of 10.81 ± 1.24 and 10.41 ± 1.37 L/m, FTI of 14.7 ± 5.8 and 11.5 ± 5.1 kg/m², and SBP of 144 ± 22 and 133 ± 19 mmHg increased in diabetics compared to nondiabetics (P < 0.001) (Figure 2). ECW/height of 10.44 ± 1.29 and 10.47 ± 1.37 L/m (P > 0.05), FTI of 9.5 ± 4.1 12.3 ± 5.4 kg/m² (P < 0.001), SBP of 131 ± 18, 9.5 ± 4.1, and 135 ± 20 mmHg (P < 0.001) were lower in smokers than in nonsmokers (Figure 3).

Mean overhydration (OH) measured in the afternoon was more than morning measurements (0.0957 ± 1.03963 versus −0.0584 ± 1.06571, (P = 0.001)).

Receiver operating characteristic (ROC) curve with the performance of 0.60 (P < 0.001) that showed cut-off value of ECW/height was 10.06 L/m, with the 69% sensitivity and 45% specificity for SBP: 140 mmHg values.

According to hydration reference plot, there were 12 (0.6%) cases in Region 1, 405 (19.9%) in Region 2, 164 (8.1%) in Region 3, 22 (1.1%) in Region 4, 1431 (51.4%) in DX, and 1047 (51.4%) in Normal region (Figure 4). All cases were compared with DX region; in Region 1 there were no cases with CAD and increased incidence of DM 33.3% (P < 0.05). In Region 2 increased incidence of DM and CAD and increased BMI (20.3%, 7.7% and 29.7 kg/m², resp.) (P < 0.001). In Region 3 increased BMI 28.6 kg/m²(P < 0.001). In Region 4 increased incidence of CAD 13.6%, (P < 0.001). (Table 4).

4. Discussion

Hypertension is a public health problem. According to recent meta-analysis including 966 patients with masked hypertension (MH) and 2640 healthy controls with sustained normotension, controlled hypertension, and white coat hypertension, the prevalence of LVH ranged from 7 to 66% in MH and from 0.4 to 42% in non-MH counterparts (average 29 versus 9%, P < 0.01) [12]. Definition of the special risk factors (genetics, dietary habits, and life style), leading causes, and early diagnosis are important for prevention of the disease for individual population.

Bioelectrical impedance spectroscopy is a noninvasive, inexpensive, and portable method that has been used mainly for body-composition analysis over the past decade. It is a suitable tool for large epidemiological studies. Our data showed that high blood pressure subjects had increased volume parameters which are consistent with previous observations [13–16].

In present study E/I was higher in hypertensive group. But we know that E/I could be affected by changes in ICW as well as by ECW varying with hydration. This may lead to the spurious impression of overhydration in subjects with smaller ICW volumes so it does not reflect hydration alone [17]. Therefore we used other methods of expressing ECW as a measure of hydration.

Bomback et al. has reported that kidney handicap results in an altered aldosterone-ECV relationship that extends into the general population having negative volume, load, and inflammatory effects [18–20]. The expanded ECW in haemodialysis patients and the markedly elevated aldosterone levels seen in ESRD [18]. Low-dose mineralocorticoid receptor blocker therapy which provided better control of subclinical ECW expansion may attenuate the adverse effects of this receptor activation [19]. In obesity aldosterone level elevation and extracellular volume expansion are crucial for renal disease via aldosterone's nonepithelial, profibrotic, and proinflammatory effects [20]. In present study BMI highly correlated with FTI measured by bioimpedance. Obese people have significantly higher SBP than normal and low BMI group. Tagliabue et al. investigated the differences in the relationship between multifrequency impedance and body-water compartments (total body water (TBW) and extracellular water (ECW)) measured by dilution techniques in Italian and Dutch healthy subjects aged 19–41 years. In body build between the two groups, the main differences were height, trunk length, and the two ratios TBW/height and ECW/height. Population-specific prediction formulas for ECW (at 1 kHz) and TBW (at 100 kHz) were developed. The prediction errors for ECW and TBW were about 0.6 and 1.5 kg, respectively, in both groups. They concluded that the water distribution between the extra- and intracellular compartments was the major cause of error in the prediction of body water, and in particular of ECW from impedance measurements with a population-specific equation [21]. Deurenberg et al. found that different, more slender body build leads to overestimation of ECW of the Ethiopian population. That study indicated that the validity of predicted body water from impedance depends on the body build of the subjects, which should be taken into account to avoid systematic errors when applying prediction formulas from a reference population to another population under study [22]. Bartz et al. showed ECW measured by multifrequency impedance was underestimated in males and slightly overestimated in females of Indonesian people and concluded that the validation in a larger group of related population subjects was needed [23]. But at the same time it has been discovered that OH, as determined by ECW/Ht, is highly predictive of clinical disease. ECW/height correlated well with volume overload as assessed by echocardiography in PD patients [24, 25]. In the present study there was no obvious relationship between OH and blood pressure, but this might be expected because OH is calculated from the difference between the measured ECW and that expected which refer to patients of same age and sex, but they do not take into account racial differences.

A decrease in vascular compliance, such as what occurs with aging in larger blood vessels such as the aorta, may also contribute to the development of isolated systolic hypertension. Even in these conditions, however, it is quite possible that the kidney plays a role in maintaining the hypertensive response [26]. In present study we also found that advanced age and diabetic groups have high blood pressure.

Visser et al. showed that young healthy men with a higher BMI are associated with a larger increase in ECW during high salt intake, suggesting that altered sodium and fluid handling may be an early phenomenon in the pathophysiological consequences of weight excess and that dietary sodium restriction may have preventive potential in overweight subjects [27]. In present study BMI and FTI were higher in high blood pressure group. For every 10% increase in body fat, OH decreased by 1.2 L; obesity seems to afford some protection aginst OH in dialysis patients [28].

In present study high blood pressure group have lower GFR measured by MDRD. In the literature the relation of ECW increase and a positive correlation between ECW and HT in patients with poor renal function were demonstrated [29, 30]. Recently we also demonstrated the strong positive correlation of increased ECW and hypertension in peritoneal dialysis patients [25].

Diabetes history, smoking, and bioimpedance measurement time afternoon were all associated with increased ECW. Brizzolara et al. demonstrated that poor glucose control positively correlated with ECW and E/I ratio [31]. These observations proving that good or moderate long-term control IDDM patients have proportionately normal distributions of ECW and ICW excess. Osmotic effect of glucose may be responsible.

A causal relationship between exposure to smoking and increase in blood pressure (BP) is not yet clearly demonstrated [32]. In a study 2742 Turkish adults prospectively evaluated for more than 7 years revealed that current cigarette smoking played a protective role at borderline significance but former smokers uniformly exhibited significantly higher risk for the development of hypertension. In our study we took only currently smoking cases as smokers, so lower blood pressure in smokers may be due to current smokers [33]. Fat tissue index was also found less than expected in smokers.

In present study for the cut-off value 10.06 L/m of ECW/height; high blood pressure predicted positively in 69% and excluded in 45%.

To our knowledge, there were no such studies comparing the effect of time period on bioimpedance measurement. We revealed that afternoon measurements were associated with increased OH, may be due to circadian rhythm of cortisone or likely whole body measurements [34]. The other reason may be the fluid shifts to the extremities during standing. It is better to tailor bioimpedance measurement protocols accordingly after such epidemiological studies.

Although the cross-sectional design of our study and other limitations preclude to infer a causal relationship between high blood pressure and ECW/height ratio, increased age and BMI and diabetes history. All these risk factors must be taken into consideration for hypertension.

There is no satisfying explanation of changes of body water composition in high blood pressure. BIS a noninvasive, cheap, and easily repeatable method has the potential to improve the etiology and management of the various hypertensive states in the majority of patients all over the world.

Bioelectrical impedance analysis is a practical method for the followup of antihypertensive therapy in pregnancy [35] and in the future it might be used for all hypertensive subjects evaluating the effectiveness of pharmacological treatments by means of volume restoration accordingly.

Expected OH in the BCM device used is setup from a largely Caucasian population, but it remains to be demonstrated whether there are differences in hydration status in different ethnic groups. In the Turkish population however the results indicate no bias in hydration when measured with BCM. So other population studies will be needed in this field.

More than half of the studied population is in “N” region of the hydration reference (Wabel plot). Most of the cases are in DX region (includes N region) 70.4%. Region 1 (overhydrated-hypertensives) composed of a minor group of population (0.6%) and having statistically significant DM and absence of CAD. Hypertension may be volume dependent just as seen in most dialysis patients. In this group hypertension treatment strategy must be adjusted towards diuretics. Region 2 (normo/hypovolemic-hypertensives) composed of nearly 1/5 of the population and having statistically significant DM, CAD and obese cases in this group. Etiology of hypertension seems as volume independent and may be speculated as rennin dependent. Accordingly angiotensin converting enzyme inhibitors and angiotensin receptor blockers might be the first choice for probable etiology and related comorbidities. Region 3 (underhydrated-normal/low SBP) composed of 8.1% of whole population with increased number of obesity. Region 4 (overhydrated-normo/hypotensives) composed of 1.1% of population with increase incidence of CAD. This group needs further cardiac evaluation for congestive heart diseases and salt restriction must be logical approach for such cases.

Normal BP in the presence of overhydration would also be an important group to identify/screen to target treatment strategies appropriately. Hypertensive subjects who are not overhydrated would likely need different treatment strategies also. Measuring BP and hydration and taking into account the traditional risk factors may allow for better diagnosis and treatment. Blood pressure and hydration can be measured readily and this could help to direct the line of investigation in the search for primary causes.

High blood pressure may be accompanied by increased extracellular volume indices. In the future volume status assessment could be of use in evaluating the effectiveness of pharmacological intervention in the treatment of hypertension.