Abstract
Background
This study aimed to investigate the association of both body mass index (BMI) and waist circumference (WC) with left ventricular mass (LVM) in hypertensive predialysis chronic kidney disease (CKD) patients.
Methods
From 2004 to 2005, 206 consecutive incident adult patients from the outpatient CKD clinics of two hospitals in Greece were included. Inclusion criteria were the presence of CKD and hypertension. BMI (kg/m2), WC (cm) and LVM (g) were assessed annually for 3 years.
Results
The mean age was 68.1 years, mean BMI 29.1 kg/m2 and mean WC was 103.7 cm. The median LVM was 245.7 g (n = 179). In the cross-sectional data, linear regression models showed that WC {β = 1.2 [95% confidence interval (CI) 0.15; 2.3]}, and not BMI [β = 2.1 (95% CI: −0.70; 4.8)], was significantly associated with LVM. After adjustment for age, sex, primary renal disease, smoking and history of cardiovascular disease, both BMI [β = 4.7 (95% CI: 2.0; 7.4] and WC [β = 1.2 (95% CI: 0.14; 2.3)] were significantly associated with LVM. These associations were pronounced in CKD stage 1–3, but not in CKD stage 4–5. In the longitudinal analysis, linear mixed models adjusting for confounders showed that both an increase in BMI [β = 2.9 (95% CI: 0.74; 5.1)] and an increase in WC [β = 1.1 (95% CI: 0.28; 1.8)] were significantly associated with an increase in LVM.
Conclusions
In hypertensive predialysis CKD patients, both BMI and WC were associated with LVM in CKD stage 1–3, but not in CKD stage 4–5. In the longitudinal analysis, both an increase in BMI and WC were associated with an increase in LVM. Future studies should focus on mechanisms responsible for the associations between anthropometric variables and LVM.
Keywords: body mass index, chronic kidney disease, hypertension, left ventricular mass, waist circumference
INTRODUCTION
Cardiovascular events, mortality and the progression to end-stage renal disease (ESRD) are major complications of chronic kidney disease (CKD) [1–3]. Previous studies have shown that a patient with CKD is at higher risk of developing a cardiovascular event than reaching ESRD [4–6]. Left ventricular hypertrophy (LVH), a common comorbidity in CKD patients and hypertensive patients [7–10], is a strong risk factor for cardiovascular events and mortality [11–14]. Therefore, identification of modifiable risk factors for increased LV mass (LVM) to prevent LVH and its major consequences in CKD patients is needed, especially in those with hypertension as this is an important risk factor for increased LVM.
Obesity, a worldwide health problem, is a potential risk factor for an increase in LVM and for chronic diseases like CKD [15, 16]. Once patients have acquired CKD, obesity adds to the increased risk of cardiovascular events [17]. Studies in dialysis [18] and predialysis CKD patients [17, 19, 20] recently suggested that abdominal obesity is a better predictor for adverse events than body mass index (BMI). BMI does not distinguish between weight from muscle and fat, and this might obscure the relationship between BMI and outcome. In contrast, waist circumference (WC) detects abdominal obesity and visceral fat, which appear to be the most important fat deposits affecting cardiovascular risk [21]. So far, to our knowledge, studies investigating both BMI and WC as risk factors for LVM in (hypertensive) predialysis CKD patients are lacking. Furthermore, longitudinal studies with LVM as an outcome are very limited in this patient group [22].
Therefore, this longitudinal study aimed to investigate the association of both BMI and WC with LVM in hypertensive predialysis CKD patients.
MATERIALS AND METHODS
Patients
From 2004 to 2005, 206 consecutive incident adult patients from the outpatient CKD clinics of two hospitals in Greece were included in this study. Prior to the study, inclusion patients were followed up for 3 months to confirm the presence of CKD. The study population included patients with hypertension only (93% of study population), getting a more homogenous patient population in which all had this important risk factor for increased LVM. Hypertension was defined as systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg or if they were using antihypertensive medication. Exclusion criteria were a history of malignancy with a ‘free of disease’ period of <5 years, and the presence or history of inflammation or a major cardiovascular event, defined as stroke, myocardial infarction or acute ischaemic heart disease, during the last 3 months prior to study entry. The definition of CKD was based on the Clinical Practice Guidelines for Chronic Kidney Disease (Kidney Disease Outcomes Quality Initiative) [23]. The study was approved by local ethical committees of the two hospitals, and patients participated after providing informed consent.
Data collection at baseline and follow-up
At study entry, the patients underwent a detailed review of their medical history and careful clinical examination. BMI, WC and LVM were assessed annually for a period of 3 years. Additionally, baseline measurements, including demographic characteristics, primary renal disease, diabetic status, smoking habits, medication and blood pressure, were assessed. A full haematology and biochemical screen was performed and estimated glomerular filtration rate (eGFR) (mL/min/1.73 m2) was calculated by the CKD-EPI formula [24].
Independent variables
BMI was calculated by dividing body weight (kg) by height2 (m2). BMI was used as a continuous variable and was categorized into normal weight (18.5–24.9 kg/m2), overweight (25–29.9 kg/m2) and obese (≥30 kg/m2). WC was measured in centimetres with the patient erect and the tape horizontally placed just above the iliac crest.
Outcome variables
The primary outcome was LVM. The secondary outcomes were LVM corrected for height2.71 and LVM corrected for body surface area. We chose the uncorrected LVM as the primary outcome, because the anthropometric variables (which are used to calculate the corrected LVM outcomes) were also used as independent variables.
LVM was assessed by two-dimensional (2D)-mode echocardiographic screening and Doppler for valve flow, which were performed within 1 week and no longer than 1 month from study entry, by a single cardiologist from each hospital, who followed a predefined protocol for the recordings and measurements [25]. Left ventricular end-diastolic diameter (LVEDD), interventricular septum (IVS) and posterior wall (PW) thickness were measured according to the Penn convention [26]. LVM was estimated in grams with the Devereux formula: [LVM=1.04×[(LVEDD+PW+IVS)3−LVEDD3]−13.6 g] [27].
Statistical analysis
Data are presented as mean and standard deviation (SD) (for normally distributed data), median and inter-quartile range (for non-normally distributed data) or as frequency in percentage. To compare the patient characteristics across BMI categories, we used tests for trend of means (for normally distributed data), and medians (for non-normally distributed data) and an χ2 test for trend provided (for categorical variables). A two-tailed P-value of <0.05 was considered as statistically significant.
We used linear regression analysis to test the association between both BMI and WC with LVM. For each association (i.e. BMI with LVM and WC with LVM), three models were built. Model 1 was an unadjusted model and Model 2 was adjusted for variables fulfilling the criteria for confounding [28], i.e. age, sex, primary renal disease, smoking and a history of CVD. Model 3 was adjusted for the confounders of Model 2 plus eGFR, in order to examine to what extent the associations were mediated by eGFR.
Additionally, the following two analyses were performed to test whether an interaction existed between both BMI and WC with CKD stage, systolic blood pressure, pulse pressure and sex. First, the corresponding multiplicative terms were introduced into the linear regression models. Secondly, unadjusted and adjusted linear regression models were performed by CKD stage, by tertiles of systolic blood pressure groups, by tertiles of pulse pressure groups and by sex.
Linear mixed models were used to analyse the association between both BMI and WC with LVM using 3-year follow-up data, thereby adjusting for multiple measurements in the same patient and for possible confounders. For both associations (i.e. change in BMI with change in LVM and change in WC with change in LVM), the same three models were built as described above. The analyses were performed within IBM SPSS Statistics 19 and SAS version 9.2.
RESULTS
Baseline characteristics
Table 1 shows the characteristics of the whole patient population (n = 206) and for those with at least one follow-up measurement on LVM (n = 107), overall and by BMI category. Of the 206 individuals, the mean age was 68.1 years, mean BMI was 29.1 kg/m2 (SD = 5.0) and mean WC was 103.7 cm (SD = 12.6). The 179 patients having data available on LVM had a median LVM of 245.7 g (males 297.4 g; females 220.8 g), a median LVM corrected for height of 69 g/m2.71 (males 70.4 g/m2.71; females 64.6 g/m2.71) and a median LVM corrected for body surface area of 137.6 g/m2 (males 154.0 g/m2; females 122.5 g/m2). The proportion of females, the proportion of patients with a history of a cardiovascular disease and the proportion of patients using diuretics, aspirin and statins increased with increasing BMI (P < 0.05). In addition, LVM (height corrected) decreased with increasing BMI (P < 0.05). Remarkably, in this group of predialysis CKD patients with hypertension, the proportion of current smokers decreased, eGFR estimated by CKD-EPI and MDRD increased and systolic blood pressure and LDL decreased with increasing BMI category (P < 0.05). It should be noted that for the 107 persons with follow-up data on LVM, systolic blood pressure and LDL did not significantly change by BMI category.
Table 1.
Baseline characteristics of pre-RRT CKD patients with hypertension, by BMI categories for all persons at baseline (n = 206) and for persons with follow-up data on left ventricular mass (n = 107)
| All persons (n = 206) |
Persons with follow-up data on left ventricular mass (n = 107) |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| All | Baseline BMI (kg/m2) |
P-value, test for trend | All | Baseline BMI (kg/m2) |
P-value, test for trend | |||||
| Normal weight 18.5–24.9 kg/m2 | Overweight 25–29.9 kg/m2 | Obesity ≥30 kg/m2 | Normal weight 18.5–24.9 kg/m2 | Overweight 25–29.9 kg/m2 | Obesity ≥30 kg/m2 | |||||
| Number (%) | 206 (100) | 47 (22.8) | 79 (38.3) | 80 (38.8) | 107 (100) | 20 (18.7) | 34 (31.8) | 53 (49.5) | ||
| Age (years) | 68.1 (60.5–73.7) | 65.0 (51–71) | 68.5 (60.4–74.4) | 70.4 (62.0–74.0) | 0.03 | 69.8 (62.8–74.0) | 64.8 (52.9–69.9) | 70.1 (63.0–74.3) | 71.2 (64.0–74.1) | 0.04 |
| Male (%) | 52.9 | 72.3 | 60.8 | 33.8 | <0.001 | 45.8 | 65.0 | 55.9 | 32.1 | 0.02 |
| Diabetes mellitus (%) | 32.0 | 25.5 | 27.8 | 40.0 | 0.07 | 37.4 | 25.0 | 35.3 | 43.4 | 0.14 |
| Smoking (%) | ||||||||||
| Never | 58.3 | 36.2 | 62.0 | 67.5 | 0.002 | 65.4 | 45.0 | 61.8 | 75.5 | 0.02 |
| Ex-smoker | 26.2 | 40.4 | 21.5 | 22.5 | 20.6 | 35.0 | 17.6 | 17.0 | ||
| Current | 15.5 | 23.4 | 16.5 | 10.0 | 14.0 | 20.0 | 20.6 | 7.5 | ||
| History of cardiovascular disease (%) | 25.6 | 13.0 | 20.3 | 38.5 | 0.001 | 37.4 | 20.0 | 29.4 | 49.1 | 0.01 |
| Primary renal disease (%) | ||||||||||
| Glomerulonephritis | 8.7 | 8.5 | 6.3 | 11.3 | 0.10 | 6.5 | 10.0 | 2.9 | 7.5 | 0.41 |
| Diabetic nephropathy | 15.5 | 14.9 | 13.9 | 17.5 | 15.0 | 10.0 | 17.6 | 15.1 | ||
| Hypertensive nephrosclerosis | 17.0 | 12.8 | 13.9 | 22.5 | 26.2 | 15.0 | 23.5 | 32.1 | ||
| Other | 25.8 | 19.1 | 33.0 | 22.4 | 28.9 | 30.0 | 38.4 | 22.7 | ||
| Unknown | 33.0 | 44.7 | 32.9 | 26.3 | 23.4 | 35.0 | 17.6 | 22.6 | ||
| eGFR | ||||||||||
| CKD epi, mL/min/1.73 m2 | 42.3 (27.2–68.8) | 35.0 (22.6–24.3) | 41.7 (29.5–73.2) | 48.9 (29.4–68.0) | 0.06 | 44.9 (32.0–74.5) | 38.6 (29.0–59.2) | 48.7 (30.1–86.1) | 50.6 (35.3–71.8) | 0.04 |
| CKD stage (%) | ||||||||||
| CKD stage 1 | 13.6 | 4.3 | 10.1 | 7.5 | 0.04 | 15.0 | 10.0 | 23.5 | 11.3 | 0.54 |
| CKD stage 2 | 24.8 | 14.9 | 25.3 | 31.3 | 23.4 | 15.0 | 17.6 | 30.2 | ||
| CKD stage 3 | 38.3 | 40.4 | 39.2 | 36.3 | 39.3 | 50.0 | 35.3 | 37.7 | ||
| CKD stage 4–5 (not on RRT) | 23.3 | 40.4 | 25.3 | 25.0 | 22.4 | 25.0 | 23.5 | 20.8 | ||
| Antropometric measures | ||||||||||
| Waist circumference (cm) | 103.7 (12.6) | 90.1 (7.3) | 100.8 (7.1) | 114.5 (9.8) | <0.001 | 106.9 (13.2) | 91.5 (6.8) | 100.7 (5.8) | 116.6 (10.2) | <0.001 |
| Left ventricular massa | ||||||||||
| Left ventricular mass (g) | 245.7 (192.7–331.5) | 228.4 (190.0–297.4) | 253.6 (191.6–341.1) | 253.3 (209.1–339.2) | 0.09 | 244.5 (190.0–330.6) | 219.6 (163.4–320.2) | 242.4 (173.3–335.0) | 250.2 (207.6–323.1) | 0.17 |
| Left ventricular mass height2.71 indexed (g/m2.71) | 69.0 (52.5–86.0) | 60.1 (48.9–74.4) | 68.6 (50.4–85.5) | 75.7 (58.7–95.7) | 0.001 | 67.7 (50.8–86.0) | 57.0 (47.7–75.4) | 60.9 (45.2–81.9) | 74.0 (58.3–94.4) | 0.008 |
| Left ventricular mass body surface area indexed (g/m2) | 137.6 (105.5–172.2) | 138.7 (106.1–173.7) | 139.9 (105.4–181.7) | 134.8 (104.6–169.6) | 0.79 | 130.8 (102.6–167.7) | 127.1 (105.8–187.2) | 128.8 (94.5–181.7) | 132.4 (103.4–163.7) | 0.98 |
| Blood pressure | ||||||||||
| Systolic blood pressure (mmHg) | 142.5 (18.3) | 146.4 (21.6) | 143.6 (17.5) | 139.1 (16.5) | 0.03 | 139.3 (17.3) | 139.3 (17.3) | 138.7 (18.6) | 139.7 (16.8) | 0.92 |
| Diastolic blood pressure (mmHg) | 81.2 (10.9) | 82.5 (13.3) | 81.3 (9.6) | 80.5 (10.7) | 0.31 | 79.1 (11.1) | 80.6 (12.2) | 77.1 (10.3) | 79.8 (11.2) | 0.80 |
| Pulse pressure (mmHg) | 61.2 (16.6) | 63.9 (19.2) | 62.3 (16.9) | 58.7 (14.5) | 0.09 | 60.2 (15.5) | 58.7 (15.4) | 61.7 (16.9) | 59.9 (14.8) | 0.77 |
| Lipids | ||||||||||
| LDL (mg/dL) | 128.4 (36.3) | 134.9 (38.6) | 132.0 (36.1) | 120.9 (34.3) | 0.04 | 121.9 (32.3) | 128.0 (36.8) | 125.1 (30.4) | 117.6 (31.7) | 0.22 |
| HDL (mg/dL) | 51.9 (13.2) | 54.0 (12.5) | 54.0 (12.8) | 48.7 (13.6) | 0.03 | 52.3 (14.9) | 55.4 (16.3) | 55.7 (15.0) | 48.9 (13.8) | 0.10 |
| Inflammatory factors | ||||||||||
| CRP (mg/L) | 2.0 (0.9–5.0) | 2.0 (0.6–4.0) | 2.0 (1.0–6.0) | 2.2 (0.6–5.0) | 0.51 | 1.0 (0.31–3.5) | 0.95 (0.17–2.15) | 1.0 (0.39–3.5) | 1.13 (0.38–4.1) | 0.53 |
| Medications (%) | ||||||||||
| ACEi | 39.8 | 34.0 | 45.6 | 37.5 | 0.87 | 39.3 | 25.0 | 47.1 | 39.6 | 0.43 |
| ARBs | 26.2 | 21.3 | 29.1 | 26.3 | 0.63 | 27.1 | 20.0 | 26.5 | 30.2 | 0.39 |
| Diuretics | 53.4 | 36.2 | 39.2 | 57.5 | 0.03 | 59.8 | 40.0 | 55.9 | 69.8 | 0.02 |
| Aspirin | 17.0 | 6.4 | 17.7 | 22.5 | 0.02 | 18.7 | 5.0 | 17.6 | 24.5 | 0.06 |
| Statin | 34.0 | 19.1 | 39.2 | 37.5 | 0.06 | 35.5 | 25.0 | 44.1 | 34.0 | 0.74 |
NA, not applicable; RRT, renal replacement therapy. Results of continuous variables are presented as mean (SD) or median (25th and 75th percentile).
aData available on 179 persons.
Figure 1a–c shows the mean BMI, mean WC and median LVM by CKD stage. BMI and WC were lowest in CKD stage 4–5, whereas LVM was highest in CKD stage 4–5.
FIGURE 1:
Mean BMI in kg/m2 (a), mean WC in cm (b) and median LVM in g (c), per CKD stage (not on RRT) (n = 206).
Cross-sectional associations
Figure 2 and Table 2 show the association between BMI and LVM and that between WC and LVM. In the unadjusted analysis (Model 1), WC, and not BMI (as a continuous variable), was significantly associated with LVM. After adjustment for confounders (Model 2), both BMI and WC were significantly associated with LVM. Additional adjustment for eGFR increased all βs to some extent (Model 3). When analysing the association between BMI as a categorical variable with LVM, patients who were overweight or obese had a significantly higher LVM compared with those with normal weight, both in the unadjusted and adjusted analyses (Table 2).
FIGURE 2:
Correlation between BMI (left figure) and WC (right figure) with LVM (n = 206).
Table 2.
The association between both BMI and WC with LVM using linear regression analysis in cross-sectional data
| Left ventricular mass (g) (n = 179) |
Left ventricular mass height2.71 indexed (g/m2.71) (n = 179) |
Left ventricular mass, body surface area indexed (g/m2) (n = 179) |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Model 1: unadj | Model 2: adja | Model 3: adjb | Model 1: unadj | Model 2: adja | Model 3: adjb | Model 1: unadj | Model 2: adja | Model 3: adjb | |
| BMI (kg/m2) continuous | 2.1 (−0.70; 4.8) | 4.7 (2.0; 7.4) | 5.5 (2.9; 8.1) | 1.3 (0.59; 2.0) | 1.4 (0.65; 2.1) | 1.6 (0.90; 2.3) | −0.048 (−1.9; 0.92) | 0.27 (−1.1; 1.7) | 0.74 (−0.62; 2.1) |
| BMI categorical | |||||||||
| Normal weight: 18.5–24.9 kg/m2 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |
| Overweight: 25–29.9 kg/m2 | 36.4 (0.11; 72.7) | 37.3 (5.1; 69.6) | 50.6 (19.5; 81.7) | 9.0 (−0.12; 18.1) | 6.7 (−1.8; 15.3) | 10.1 (1.8; 18.4) | 7.3 (−11.4; 26.0) | 5.7 (−11.3; 22.7) | 13.4 (−3.0; 29.9) |
| Obesity: ≥30 kg/m2 | 39.0 (3.11; 74.9) | 58.9 (24.6; 93.2) | 71.2 (38.2; 104.2) | 16.0 (7.0; 25.0) | 14.5 (5.4; 23.6) | 17.6 (8.8; 26.4) | 0.29 (−18.2; 18.7) | 4.7 (−13.3; 22.8) | 11.9 (−5.4; 29.3) |
| Waist circumference (cm) | 1.2 (0.15; 2.3) | 1.2 (0.14; 2.3) | 1.7 (0.61; 2.7) | 0.46 (0.19; 0.74) | 0.28 (0.00; 0.57) | 0.39 (0.10; 0.67) | 0.02 (−0.54; 0.58) | −0.18 (−0.74; 0.39) | 0.06 (−0.49; 0.61) |
aModel 2: adjusted for age, sex, smoking, primary renal disease and history of cardiovascular disease.
bModel 3: adjusted for confounders of Model 2 + eGFR.
When adding interaction terms in the linear regression models, those of BMI with CKD stage, systolic blood pressure, pulse pressure and sex, and the interaction terms of WC with CKD stage, systolic blood pressure, pulse pressure and sex failed to reach statistical significance. Nevertheless, when performing linear regression models by CKD stage, the adjusted models showed that both BMI and WC were associated with LVM in CKD stage 1–3, but not in CKD stage 4–5 (Table 3), suggesting an interaction between BMI and CKD stage and between WC and CKD stage in association with LVM. As CKD stage 4–5 contained fewer patients than CKD stage 1–2 and CKD stage 3, we performed these analyses also by eGFR in tertiles, and obtained similar results. After adjustment for age, smoking and history of cardiovascular disease, the association between both BMI and LVM and WC and LVM was more pronounced in males [β = 7.4; 95% confidence interval (CI): 2.2; 12.6 and β = 3.8; 95% CI: 0.81; 6.7, respectively) than in females (β = 2.13; 95% CI: 0.11; 4.2 and β = 1.02; 95% CI: −0.33; 2.7, respectively).
Table 3.
The association between both BMI and WC with LVM, by CKD stage using linear regression analysis in cross-sectional data
| Left ventricular mass (g) (n = 179) |
||||||
|---|---|---|---|---|---|---|
| Model 1: unadj |
Model 2: adja |
|||||
| CKD stage 1–2 (n = 64) | CKD stage 3 (n = 66) | CKD stage 4 (n = 49) | CKD stage 1–2 (n = 64) | CKD stage 3 (n = 66) | CKD stage 4 (n = 49) | |
| BMI (kg/m2) continuous | 1.3 (−2.9; 5.5) | 4.2 (−0.18; 8.5) | 3.6 (−2.2; 9.4) | 4.7 (1.0; 8.7) | 6.7 (2.1; 11.3) | 4.3 (−2.0; 10.7) |
| Waist circumference (cm) | 1.4 (−0.3; 3.1) | 2.3 (0.32; 4.2) | 1.3 (−0.7; 3.3) | 1.8 (0.01; 3.5) | 2.3 (0.28; 4.2) | 0.59 (−1.7; 2.9) |
aModel 2: adjusted for age, sex, primary renal disease, smoking and history of cardiovascular disease.
Longitudinal associations
Table 4 shows that in persons with BMI, WC and LVM measurements both at baseline and at 3 years (n = 71), the mean BMI did not significantly change over 3 years, whereas the mean WC was significantly lower after 3 years and the median LVM significantly higher after 3 years. In 66.2% of this patient population with 3-year follow-up data, LVM had increased over 3 years with a median (25th–75th percentile) increase of 54.6 g (27.3–75.6 g).
Table 4.
Mean BMI, mean WC and median left ventricular mass at baseline and at 3 years follow-up (n = 71a)
| Baseline | At 3 years follow-up | Baseline versus 3 year follow-up P-valueb | |
|---|---|---|---|
| BMI (kg/m2) continuous | 30.6 (5.4) | 30.4 (5.6) | 0.99 |
| Waist circumference (cm) | 107.9 (12.5) | 105.2 (14.1) | 0.006 |
| Left ventricular mass (g) | 227.3 (165.6; 294.2) | 253.7 (198.0–333.7) | 0.002 |
aData based on 71 persons who had BMI, WC and LVM measurements both at baseline and at 3 years.
bParametric paired t-test (for BMI and WC) and the Wilcoxon matched-pair signed-rank test (for LVM) were used.
Table 5 shows the associations between the change in BMI and the change in LVM as well as between the change in WC and the change in LVM. In the unadjusted analysis, a change in BMI was not associated with a change in LVM. In contrast, an increase in WC was associated with an increase in LVM. After adjustment (Model 2), both an increase in BMI and an increase in WC were associated with an increase in LVM. Additional adjustment for eGFR increased the effect to some extent (Model 3).
Table 5.
The association between change in both BMI and WC and change in LVM, per year using linear mixed modelling
| Left ventricular mass (g) (n = 107) |
Left ventricular mass height2.71 indexed (g/m2·71) (n = 107) |
Left ventricular mass body surface area indexed (g/m2) (n = 107) |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Model 1: unadj | Model 2: adja | Model 3: adjb | Model 1: unadj | Model 2: adja | Model 3: adjb | Model 1: unadj | Model 2: adja | Model 3: adjb | |
| BMI (kg/m2) | 0.82 (−1.50; 3.16) | 2.9 (0.74; 5.1) | 3.6 (1.5; 5.7) | 1.39 (0.43; 2.35) | 0.61 (−0.41; 1.6) | 0.58 (−0.44; 1.6) | 1.19 (−2.4; 0,01) | −0.57 (−1.17; 0.57) | −0.17 (−1.26; 0.91) |
| Waist circumference (cm) | 1.19 (0.39; 1.99) | 1.1 (0.28; 1.8) | 1.3 (0.52; 2.0) | 0.68 (0.31; 1.05) | 0.28 (−0.11; 0.67) | 0.27 (−0.12; 0.67) | 0.07 (−0.34; 0.49) | −0.09 (−0.50; 0.31) | 0.01 (−0.38; 0.41) |
aModel 2: adjusted for age, sex, primary renal disease, smoking and history of cardiovascular disease.
bModel 3: adjusted for confounders of Model 2 + eGFR.
DISCUSSION
In this study of predialysis hypertensive CKD patients, we noted that WC, and not BMI, was significantly associated with LVM. However, after adjustment for confounders, both BMI and WC were associated with LVM. The results suggested that this association was pronounced in CKD stages 1–3 but not in CKD stages 4–5. In the longitudinal analysis, both an increase in BMI and an increase in WC were associated with an increase in LVM. The identification of these anthropometric variables as risk factors for LVM is essential for the prevention of LVH and its major consequences like cardiovascular mortality, especially in light of the current obesity epidemic.
Comparison of BMI and WC
Data from observational studies show that in predialysis CKD patients and dialysis patients measures of central obesity and visceral fat like WC and waist-to-hip ratio (WHR) may better predict adverse clinical outcomes including mortality than BMI [17–20, 29]. The unadjusted results of our observational study confirm this hypothesis and showed that WC, and not BMI, was significantly associated with LVM. A likely explanation for this is that WC, and not BMI, mainly measures abdominal fat which reflects visceral fat, the most important fat deposit affecting cardiovascular risk [21]. In line with this, other studies have shown that WC is highly correlated with both visceral fat [30, 31] and subcutaneous fat [31], whereas BMI is only highly correlated with subcutaneous fat [31]. In contrast, WHR is highly correlated with visceral fat, and not with subcutaneous fat [31]. Indeed, some studies found WHR to be a better predictor for adverse outcomes than both WC and BMI [17]. Unfortunately, WHR was not available in our study.
After adjustment for potential confounders, we found that both BMI and WC were significantly associated with LVM. These adjusted associations of both BMI and WC with LVM were, however, pronounced in CKD stages 1–3 but not in CKD stages 4–5. In patients with higher CKD stages in whom muscle wasting is common, a lower BMI and WC may not (only) reflect lower visceral fat (and lower CVD risk), but may also reflect lower muscle mass (and higher CVD risk). It should be noted that so far most studies in predialysis patients, which have shown an association between anthropometric measures and adverse clinical outcomes like mortality or CVD, were performed in CKD stage 3 [17, 19], whereas little is known about these associations in CKD stage 4. The study by Kramer et al. [20] including patients with CKD stages 1–4 found no significant interaction between anthropometric variables and CKD stage in association with all-cause mortality. In our study, we did not find any significant interaction terms either when adding interaction terms in the regression models. However, when analysing the data by CKD stage, the results suggested an interaction between BMI and CKD stage and between WC and CKD stage in association with LVM. With regard to the lower BMI in advanced CKD stages, it cannot be overemphasized that within CKD patients, obesity cannot be seen as a protective factor for adverse outcomes, as illustrated by a study of Honda et al. [32], in which low muscle mass at any level of fat mass was a strong predictor for mortality in CKD patients.
Longitudinal associations
Studies with longitudinal data on LVM in patients with CKD are extremely scarce [22]. A study by Eckardt et al. [33] showed that the LVMI did not significantly change over 3 years in patients with CKD stage 3 and 4. In contrast, our study showed a significant increase in LVM in 3 years. To our knowledge, studies examining the association between change in BMI and change in WC with change in LVM are lacking. The study by Okumura et al. [22] was the first one on factors associated with change in eGFR and LVM in CKD patients with hypertension. However, anthropometry was not assessed in this study. Interestingly, our study showed for the first time that an increase in BMI and an increase in WC were associated with an increase in LVM persisted after adjustment for potential confounders. This crucial finding could help the prevention of LVH, and eventually its major consequences like cardiovascular mortality.
However, several longitudinal studies examined the association between anthropometric measures and CVD or mortality in this patient group [17, 19, 20]. As mentioned before, these studies have shown that WC and WHR may be better predictors for adverse outcomes than BMI.
Potential mechanisms
Traditionally, the association between obesity and LVM has been explained through haemodynamic changes. However, in our study, the mean pulse pressure and mean systolic blood pressure were lower in the obese group in comparison with the normal weight group. This could indicate that, at least in predialysis CKD patients with hypertension, the association between obesity and LVM is more complex than simple haemodynamic effects.
Numerous studies have shown that increased visceral and subcutaneous adiposity causes dysfunction of adipohormones and elevates systemic inflammatory cytokines [34]. Adipohormones, such as adiponectin and leptin, have a proliferative effect on smooth muscle cells and lead to increased vascular stiffness and LVH. On the other hand, cytokines secreted by the adipocytes such as hs-CRP and IL-6 lead to endothelial dysfunction which may also cause LVH [35]. It should be noted however that in our study, the median C-reactive protein (CRP) was similar across the BMI groups, which suggests that the association between obesity and LVM may not be explained by CRP. However, it is of notice that there was an increasing use of statins and aspirin across the BMI groups (Table 1), and this might have influenced (decreased) the level of inflammation and therefore of CRP.
Identification of the mechanisms through which obesity influences LVM is essential for the prevention of LVH and its major consequences like cardiovascular events and mortality [11–14]. The association between obesity and LVM indicates that obesity can be used as a modifiable risk factor in the reduction of CVD mortality. Rider et al. [36] proved that weight loss leads to a reduction in LVM in obese non-CKD persons. Nonetheless, future studies are needed to confirm this effect in CKD patients.
Strengths and limitations
The main strength of this study was the availability of longitudinal data on BMI, WC and LVM, the inclusion of patients from all CKD stages and use of the chronicity criterion. However, our study has some limitations. First, 2D echocardiography was used despite the limitations of this technique for the quantification of LVM, mainly regarding geometric assumptions and the dependence on adequate endocardial and epicardial border definitions of the LV [37]. Cardiac magnetic resonance imaging (CMRI) is considered to be the ‘gold standard’ [37]. However, CMRI is not widely available and quite costly, especially if applied in large numbers of patients and for a longitudinal follow-up. In order to minimize subjectivity and to limit variations in echocardiography performance and measurements, the 2D echocardiographic screening was performed by a single cardiologist in each hospital, who followed a predefined protocol for the recordings, and the final reading and measurements for all echo studies were made by one cardiologist. Furthermore, the echo study was performed within 1 week and no longer than 1 month from study entry, with the patient in steady state, thus avoiding fluctuations in fluid volume. Secondly, cross-sectional data are prone to selection bias. In contrast to the general population, we found in this predialysis hypertensive CKD population that the proportion of current smokers decreased from the lowest to highest BMI category, that eGFR was progressively higher and systolic blood pressure and LDL were progressively lower from the lowest BMI category on. An explanation for the lower proportion of current smokers could be the medical consultation encouraging smoking cessation and the lower proportion of males in the highest BMI category. An explanation for the progressively lower systolic blood pressure could be the higher use of diuretics and the better kidney function and for the lower LDL could be the higher use of statins from the lowest to the highest BMI categories, respectively. Furthermore, the highest eGFR in the highest BMI category may be the effect of obesity on kidney function through multiple mechanisms (e.g. haemodynamic effects, adipokines, inflammation). Thirdly, although it is possible that BMI and WC cause LVM, in the cross-sectional analyses, it is not possible to distinguish causes and effects. Fourthly, the study may suffer from confounding, as it was not possible to correct for confounders that were not measured or were unknown. Therefore, among others, in this observational study, causality cannot be inferred.
CONCLUSION
Our results suggest that both BMI and WC were associated with LVM in hypertensive predialysis CKD patients, and this association was present in CKD stages 1–3, but not in CKD stages 4–5. Within the latter CKD stages, the relationship is more complex due to the varied and multiple mechanisms by which uraemia may influence LVM and peripheral muscle wasting. In the longitudinal analysis, both an increase in BMI and an increase in WC were associated with an increase in LVM. Although traditionally the association between obesity and LVM has been explained mainly through haemodynamic changes, our results suggest that the increase in LVM may not exclusively be due to blood pressure. These results indicate that patients in the early stages of CKD should be advised to lose weight, if they are overweight or obese, and by this protect themselves from adverse cardiovascular events. Future studies should confirm these results and should focus on mechanisms responsible for the associations between anthropometric variables and LVM.
CONFLICT OF INTEREST STATEMENT
None declared.
ACKNOWLEDGEMENTS
We would like to acknowledge our thanks to Dr N. Kotzadamis and Dr A. Kelesidis for kindly agreeing to use the data from the Department of Nephrology, General Hospital of Veria, and Prof. V. Vargemezis and Prof. Pl. Pasadakis, Department of Nephrology, Dimokrition University of Thrace, for their advice and support to Dr. K.I.'s Thesis. The research leading to these results has received funding from the European Community's Seventh Framework Programme under grant agreement number HEALTH-F2-2009-241544 (SysKID).
REFERENCES
- 1.Di AE, Chowdhury R, Sarwar N, et al. Chronic kidney disease and risk of major cardiovascular disease and non-vascular mortality: prospective population based cohort study. BMJ. 2010;341:c4986. doi: 10.1136/bmj.c4986. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Matsushita K, van d V, Astor BC, et al. Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet. 2010;375:2073–2081. doi: 10.1016/S0140-6736(10)60674-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sarnak MJ. Cardiovascular complications in chronic kidney disease. Am J Kidney Dis. 2003;41:11–17. doi: 10.1016/s0272-6386(03)00372-x. [DOI] [PubMed] [Google Scholar]
- 4.Dalrymple LS, Katz R, Kestenbaum B, et al. Chronic kidney disease and the risk of end-stage renal disease versus death. J Gen Intern Med. 2011;26:379–385. doi: 10.1007/s11606-010-1511-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Keith DS, Nichols GA, Gullion CM, et al. Longitudinal follow-up and outcomes among a population with chronic kidney disease in a large managed care organization. Arch Intern Med. 2004;164:659–663. doi: 10.1001/archinte.164.6.659. [DOI] [PubMed] [Google Scholar]
- 6.O'Hare AM, Choi AI, Bertenthal D, et al. Age affects outcomes in chronic kidney disease. J Am Soc Nephrol. 2007;18:2758–2765. doi: 10.1681/ASN.2007040422. [DOI] [PubMed] [Google Scholar]
- 7.Cerasola G, Nardi E, Mule G, et al. Left ventricular mass in hypertensive patients with mild-to-moderate reduction of renal function. Nephrology (Carlton) 2010;15:203–210. doi: 10.1111/j.1440-1797.2009.01178.x. [DOI] [PubMed] [Google Scholar]
- 8.Levin A, Singer J, Thompson CR, et al. Prevalent left ventricular hypertrophy in the predialysis population: identifying opportunities for intervention. Am J Kidney Dis. 1996;27:347–354. doi: 10.1016/s0272-6386(96)90357-1. [DOI] [PubMed] [Google Scholar]
- 9.Nardi E, Palermo A, Mule G, et al. Left ventricular hypertrophy and geometry in hypertensive patients with chronic kidney disease. J Hypertens. 2009;27:633–641. doi: 10.1097/HJH.0b013e3283220ecd. [DOI] [PubMed] [Google Scholar]
- 10.Paoletti E, Bellino D, Cassottana P, et al. Left ventricular hypertrophy in nondiabetic predialysis CKD. Am J Kidney Dis. 2005;46:320–327. doi: 10.1053/j.ajkd.2005.04.031. [DOI] [PubMed] [Google Scholar]
- 11.Levy D, Garrison RJ, Savage DD, et al. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med. 1990;322:1561–1566. doi: 10.1056/NEJM199005313222203. [DOI] [PubMed] [Google Scholar]
- 12.London GM, Pannier B, Guerin AP, et al. Alterations of left ventricular hypertrophy in and survival of patients receiving hemodialysis: follow-up of an interventional study. J Am Soc Nephrol. 2001;12:2759–2767. doi: 10.1681/ASN.V12122759. [DOI] [PubMed] [Google Scholar]
- 13.Paoletti E, Cassottana P, Bellino D, et al. Left ventricular geometry and adverse cardiovascular events in chronic hemodialysis patients on prolonged therapy with ACE inhibitors. Am J Kidney Dis. 2002;40:728–736. doi: 10.1053/ajkd.2002.35680. [DOI] [PubMed] [Google Scholar]
- 14.Pierdomenico SD, Cuccurullo F. Risk reduction after regression of echocardiographic left ventricular hypertrophy in hypertension: a meta-analysis. Am J Hypertens. 2010;23:876–881. doi: 10.1038/ajh.2010.80. [DOI] [PubMed] [Google Scholar]
- 15.Wang Y, Chen X, Song Y, et al. Association between obesity and kidney disease: a systematic review and meta-analysis. Kidney Int. 2008;73:19–33. doi: 10.1038/sj.ki.5002586. [DOI] [PubMed] [Google Scholar]
- 16.Yusuf S, Hawken S, Ounpuu S, et al. Obesity and the risk of myocardial infarction in 27,000 participants from 52 countries: a case-control study. Lancet. 2005;366:1640–1649. doi: 10.1016/S0140-6736(05)67663-5. [DOI] [PubMed] [Google Scholar]
- 17.Elsayed EF, Tighiouart H, Weiner DE, et al. Waist-to-hip ratio and body mass index as risk factors for cardiovascular events in CKD. Am J Kidney Dis. 2008;52:49–57. doi: 10.1053/j.ajkd.2008.04.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Postorino M, Marino C, Tripepi G, et al. Abdominal obesity and all-cause and cardiovascular mortality in end-stage renal disease. J Am Coll Cardiol. 2009;53:1265–1272. doi: 10.1016/j.jacc.2008.12.040. [DOI] [PubMed] [Google Scholar]
- 19.Evans PD, McIntyre NJ, Fluck RJ, et al. Anthropomorphic measurements that include central fat distribution are more closely related with key risk factors than BMI in CKD stage 3. PLoS One. 2012;7:e34699. doi: 10.1371/journal.pone.0034699. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kramer H, Shoham D, McClure LA, et al. Association of waist circumference and body mass index with all-cause mortality in CKD: the REGARDS (Reasons for Geographic and Racial Differences in Stroke) study. Am J Kidney Dis. 2011;58:177–185. doi: 10.1053/j.ajkd.2011.02.390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zoccali C, Seck SM, Mallamaci F. Obesity and the epidemiology and prevention of kidney disease: waist circumference versus body mass index. Am J Kidney Dis. 2011;58:157–159. doi: 10.1053/j.ajkd.2011.05.009. [DOI] [PubMed] [Google Scholar]
- 22.Okumura K, Io H, Matsumoto M, et al. Predictive factors associated with change rates of LV hypertrophy and renal dysfunction in CKD patients. Clin Nephrol. 2012;79:7–14. doi: 10.5414/CN107569. [DOI] [PubMed] [Google Scholar]
- 23.National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification and stratification. Am J Kidney Dis. 2002;39(Suppl 1):S1–S266. [PubMed] [Google Scholar]
- 24.Levey AS, de Jong PE, Coresh J, et al. The definition, classification, and prognosis of chronic kidney disease: a KDIGO Controversies Conference report. Kidney Int. 2011;80:17–28. doi: 10.1038/ki.2010.483. [DOI] [PubMed] [Google Scholar]
- 25.Sahn DJ, DeMaria A, Kisslo J, et al. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation. 1978;58:1072–1083. doi: 10.1161/01.cir.58.6.1072. [DOI] [PubMed] [Google Scholar]
- 26.Devereux RB, Alonso DR, Lutas EM, et al. Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings. Am J Cardiol. 1986;57:450–458. doi: 10.1016/0002-9149(86)90771-x. [DOI] [PubMed] [Google Scholar]
- 27.Devereux RB, Reichek N. Echocardiographic determination of left ventricular mass in man. Anatomic validation of the method. Circulation. 1977;55:613–618. doi: 10.1161/01.cir.55.4.613. [DOI] [PubMed] [Google Scholar]
- 28.Jager KJ, Zoccali C, Macleod A, et al. Confounding: what it is and how to deal with it. Kidney Int. 2008;73:256–260. doi: 10.1038/sj.ki.5002650. [DOI] [PubMed] [Google Scholar]
- 29.Zoccali C, Torino C, Tripepiand G, et al. Assessment of obesity in chronic kidney disease: what is the best measure? Curr Opin Nephrol Hypertens. 2012;21:641–646. doi: 10.1097/MNH.0b013e328358a02b. [DOI] [PubMed] [Google Scholar]
- 30.Sanches FM, Avesani CM, Kamimura MA, et al. Waist circumference and visceral fat in CKD: a cross-sectional study. Am J Kidney Dis. 2008;52:66–73. doi: 10.1053/j.ajkd.2008.02.004. [DOI] [PubMed] [Google Scholar]
- 31.Seidell JC, Perusse L, Despres JP, et al. Waist and hip circumferences have independent and opposite effects on cardiovascular disease risk factors: the Quebec Family Study. Am J Clin Nutr. 2001;74:315–321. doi: 10.1093/ajcn/74.3.315. [DOI] [PubMed] [Google Scholar]
- 32.Honda H, Qureshi AR, Axelsson J, et al. Obese sarcopenia in patients with end-stage renal disease is associated with inflammation and increased mortality. Am J Clin Nutr. 2007;86:633–638. doi: 10.1093/ajcn/86.3.633. [DOI] [PubMed] [Google Scholar]
- 33.Eckardt KU, Scherhag A, Macdougall IC, et al. Left ventricular geometry predicts cardiovascular outcomes associated with anemia correction in CKD. J Am Soc Nephrol. 2009;20:2651–2660. doi: 10.1681/ASN.2009060631. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Ebinc H, Ebinc FA, Ozkurt ZN, et al. Impact of adiponectin on left ventricular mass index in non-complicated obese subjects. Endocr J. 2008;55:523–528. doi: 10.1507/endocrj.k07e-098. [DOI] [PubMed] [Google Scholar]
- 35.McQuarrie EP, Patel RK, Mark PB, et al. Association between proteinuria and left ventricular mass index: a cardiac MRI study in patients with chronic kidney disease. Nephrol Dial Transplant. 2011;26:933–938. doi: 10.1093/ndt/gfq418. [DOI] [PubMed] [Google Scholar]
- 36.Rider OJ, Francis JM, Ali MK, et al. Beneficial cardiovascular effects of bariatric surgical and dietary weight loss in obesity. J Am Coll Cardiol. 2009;54:718–726. doi: 10.1016/j.jacc.2009.02.086. [DOI] [PubMed] [Google Scholar]
- 37.Glassock RJ, Pecoits-Filho R, Barberato SH. Left ventricular mass in chronic kidney disease and ESRD. Clin J Am Soc Nephrol. 2009;4(Suppl 1):S79–S91. doi: 10.2215/CJN.04860709. [DOI] [PubMed] [Google Scholar]