Malnutrition–Inflammation–Atherosclerosis (MIA) Syndrome Components in Hemodialysis and Peritoneal Dialysis Patients

Abstract

Background. Malnutrition, inflammation, and atherosclerosis (MIA syndrome) are common in end-stage renal disease (ESRD) patients. Each component of MIA syndrome is the predictor of outcomes in ESRD patients. In this cross-sectional study, we aimed to compare both dialysis modalities for MIA syndrome components. Material and Methods. Thirty hemodialysis (HD) (mean age 44 ± 11 years, 14 male and 16 female, mean time on dialysis: 31.0 ± 19.0 months) and 30 continuous ambulatory peritoneal dialysis (CAPD) patients (41 ± 9 years, 12 male and 18 female, mean time on dialysis: 25.5 ± 21.5 months) were included. In order to determine malnutrition in ESRD patients, serum albumin level and anthropometric measurements were used. For inflammation, serum C-reactive protein level, erythrocyte sedimentation rate, and fibrinogen levels were measured. Mean-carotid artery intima media thickness (m-CIMT), presence of carotid plaque and serum homocysteine level were used to determine atherosclerosis. Results. Five CAPD patients (16%) and one HD patient (3%) was hypoalbuminemic. HD and CAPD groups were similar for inflammation. Mean-CIMT and serum homocysteine level were higher in HD patients than CAPD patients. There was a positive correlation between homocysteine and m-CIMT. Conclusion. Before choosing renal replacement therapy, malnutrition, inflammation, and atherosclerosis parameters must be investigated in ESRD patients. Hemodialysis seems to be more advantageous for malnutrition components than CAPD. Both dialysis modalities seem to be similar for inflammation, and CAPD modality has superiority for atherosclerosis. Before choosing the type of renal replacement therapy, assessment of MIA syndrome components could be useful for individualization of the decision on which dialytic modality is appropriate in ESRD patients.

Keywords:

• MIA syndrome
• atherosclerosis
• malnutrition
• hs-CRP
• albumin
• homocysteine

INTRODUCTION

Malnutrition, inflammation, and atherosclerosis are common and the main causes of morbidity and mortality in end-stage renal disease (ESRD) patients. Also, all of these parameters are the predictors of outcomes in this population.[1–5] Since the first report about association between elevated C-reactive protein (CRP)[6] and increased mortality, several groups have reported similar findings in both hemodialysis (HD)[7],[8] and peritoneal dialysis (PD) patients alike.[9],[10] Inflammation has been associated with malnutrition and atherosclerosis, and this link has led to describe MIA syndrome (Malnutrition–Inflammation–Atherosclerosis). MIA syndrome was defined as the interaction between increased pro-inflammatory cytokine levels and malnutrition and atherosclerosis in ESRD patients.[11] One, two, or all components of MIA syndrome could be present in ESRD patients.

It was reported that as the components of MIA syndrome increases, the morbidity and mortality increases in PD patients.[12] However, according to our knowledge, there was no study about comparing both dialysis modalities according to MIA syndrome components. In the present study, we aimed to compare both dialysis modalities for MIA syndrome components.

MATERIAL AND METHODS

Study Cohort

In this cross-sectional study, 30 hemodialysis (HD) and 30 continuous ambulatory peritoneal dialysis (CAPD) patients were included. All HD patients have been dialyzed using standard bicarbonate containing dialysate and semisynthetic dialyzer for 4 h three times per week. CAPD patients have been dialyzed with two liters of dialysate four times in a day. Patients more than 60 years old, dialyzed less than 6 months, hospitalized within the last 3 months, and with atherosclerotic cardiovascular disease according to ECG and clinical signs (cerebrovascular accident, peripheral artery disease, myocardial infarctions, coronary artery catheterization history) were excluded. None of the patients had a history of acute or chronic infection, such as tuberculosis, within the last 3 months. The study was approved by the local Ethical Committee.

Study Protocol

Demographic and clinical data; primary renal disease; time on duration; smoking history; use of erythropoietin, anti-hypertensive, and anti-hyperlipidemic treatment were recorded. A patient with uncontrolled hypertension was defined as a patient with systolic blood pressure higher than 140 mm Hg and/or diastolic blood pressure higher than 90 mm Hg, despite anti-hypertensive treatment use. Due to the effect on homocysteine metabolism, medications containing vitamin B6, B12, and folic acid were stopped 4 weeks before the study.

Biochemical Measurements

After an overnight fasting, all blood samples were drawn from the antecubital vein in the mid-week dialysis session for HD patients. Likewise, in HD patients, blood samples were drawn after an overnight fast in CAPD patients. Blood samples were taken into appropriate Vacutainer, centrifuged within 5 min for 4000 cycle/minute, and serum was separated. Blood samples for serum homocysteine level were stored at −20°C until analysis. Serum creatinine, albumin, transferrin, calcium, phosphorus, total cholesterol, triglyceride, HDL cholesterol, and complete blood count were performed by routine procedures in the department of clinical chemistry. Serum albumin level was measured by bromcresol purple method, and serum C-reactive protein (CRP) was measured with nephelometric method (Behring BN 200, Germany). LDL cholesterol level was calculated using the Friedewald formula.[13] Serum homocysteine level was measured by HPLC (High Performance Liquid Chromatography) method described before elsewhere[14] with a Hewlett Packard 1100 series detector.

Anthropometric and Blood Pressure Measurements

The average of three body weight measurements, before the dialysis session of the same week, was used for HD patients. In the CAPD patients, the measurement of body weight after peritoneal dialysate discharged was used at the time blood samples were drawn. Mid-arm circumferences, handgrip, biceps and triceps thicknesses were measured. An average of three blood pressure measurements, before hemodialysis session in the same week, was considered as blood pressure value for HD patients. For CAPD patients, an average of three blood pressure measurements were also used. Hypertension was defined as systolic blood pressure higher than 140 mm Hg and/or diastolic blood pressure higher than 90 mm Hg or any anti-hypertensive treatment use.

B-Mode High-Resolution Carotid Doppler Examination

Carotid Doppler examination was performed using an echocardiographic system (model 5000; Advanced Technology Laboratories, Bothell, WA, USA) equipped with a variable (2–4 MHz) phased array, cross-sectional transducer, and harmonic imaging. A trained sonographer who was blinded to clinical and biochemical data of the patients scanned the common carotid arteries bilaterally in all patients. The far wall of the common carotid artery, 0.5–1 cm proximal to the beginning of the carotid bulb, was used for carotid artery intima media thickness (CIMT) measurement. CIMT was measured as the distance between the leading edge of lumen-intima echo and media-adventitia echo as defined previously. CIMT were measured during end-diastole. The average of both side CIMT measurements was considered as mean-CIMT. The number of atherosclerotic plaques (soft plaque = grey echo, calcified plaques = white echo) were recorded. All carotid artery measurements were performed in plaque-free arterial segments. M-CIMT >1 mm was defined as atherosclerosis.[2] The intra-observer variability for IMT measurements was 3.8%, and the absolute difference between duplicate IMT measurements was 0.02 mm.

MIA Components

Serum albumin, transferrin levels, and body mass index (BMI) were used for assessment of nutritional status. Serum CRP, ferritin, plasma fibrinogen, erythrocyte sedimentation rate (ESR), and white blood cell count were used for assessment of inflammatory status. Lipid parameters, serum homocysteine level, and m-CIMT value were used for assessment of atherosclerosis. Serum albumin level <3.5 g/dL was defined as hypoalbuminemia and a patient with serum albumin <3.5 g/dL was defined as a malnourished patient. Serum CRP level >10 ng/dL (normal range 0–5 ng/dL) was defined as presence of inflammation. In addition, serum homocysteine level >15 μmol/L was defined as presence of hyperhomocysteinemia. M-CIMT >1 mm was defined as presence of atherosclerosis.

Statistical Analysis

Data are presented as mean ± SD. P value less than 0.05 was considered as statistically significant. A comparison between two groups was performed by Student t test for normally distributed variables, if necessary. Mann-Whitney U test was used to compare non-normal distributed variables. Chi-square test was used to compare nominal variables. Correlation was performed by Spearman correlation test. The independent effect of serum homocysteine level on m-CIMT was studied in a reduced multiple-regression model based on unvaried predictors of m-CIMT.

RESULTS

Patient Characteristics

There was no difference in age, gender distribution, BMI, time on dialysis, systolic and diastolic blood pressures, number of diabetics, smoking history, and use of anti-hyperlipidemic drugs between HD and CAPD groups. However, rate of erythropoietin use was higher in HD patients than CAPD patients (67% vs. 33%, P = 0.01) (Table 1).

Table 1 Demographic and clinical data of both groups

	HD patients (n = 30)	CAPD patients (n = 30)	P
Gender (M/F)	14/16	12/18	NS
Age (year)	43.9 ± 11.0	41.4 ± 9.3	NS
BMI (kg/m^2)	25.6 ± 5.8	27.3 ± 4.8	NS
Time on dialysis (months)	31.0 ± 19.0	25.5 ± 21.5	NS
Systolic BP (mm Hg)	130.6 ± 19.9	129.3 ± 25.8	NS
Diastolic BP (mm Hg)	79.6 ± 12.4	76.0 ± 16.1	NS
Diabetic patient	5 (20 %)	5 (20 %)	NS
Hypertensive patient	8 (26 %)	13 (43 %)	NS
Antihyperlipidemic drug usage	7 (23 %)	6 (20 %)	NS
Number of smokers (%)	6 (20 %)	5 (16 %)	NS
EPO usage	20 (67 %)	10 (33 %)	0.01

NS, nonsignificant.

HD and CAPD groups were compared with Student t-test for normally distributed variables and Mann-Whitney-U Test was used to compare non-normal distributed variables. Chi-square test was used to compare nominal variables.

Outcome Measurements

Nutritional Parameters

Serum BMI, albumin, and transferrin levels were used for assessment of nutritional status. Both HD and CAPD groups were similar for BMI (25.6 ± 5.8 vs. 27.3 ± 4.8 kg/m², P = 0.22), serum transferrin (148.4 ± 53.9 mg/dL vs. 167.9 ± 61.6 mg/dL, P = 0.19), and albumin levels (4.01 ± 0.44 g/dL vs. 3.92 ± 0.49 g/dL, P = 0.47). Hypoalbuminemic patient ratio was higher in CAPD patients than HD patients (16% vs. 3%, p = 0.19), despite not reaching a statistically significant level. There was also no difference in anthropometric parameters between HD and CAPD patients (Table 2).

Table 2 Differences between HD and CAPD patients

Parameters	HD patients	CAPD patients	P
Nutritional parameters
BMI (kg/m^2)	25.6 ± 5.8	27.3 ± 4.8	NS
Serum albumin (g/dL)	4.01 ± 0.44	3.92 ± 0.49	NS
Transferrin (mg/dL)	148.4 ± 53.9	167.9 ± 61.6	0.19
Inflammation parameters
Serum CRP (mg/L)	17.3 ± 23.6	21.6 ± 36.6	NS
Erythrocyte sedimentation rate (mm/h)	51.2 ± 35.5	80.7 ± 34.8	0.002
White blood count (10^3/mm^3)	7.0 ± 2.4	8.0 ± 2.0	NS
Ferritin (ng/mL)	500.2 ± 328.7	448.2 ± 388.0	NS
Fibrinogen (mg/dL)	388.4 ± 126.9	499.2 ± 126.4	0.001
Atherosclerosis parameters
Total cholesterol (mg/dL)	162.1 ± 53.8	210.9 ± 38.7	0.0001
Triglyceride (mg/dL)	180.7 ± 123.1	245.2 ± 133.8	0.05
LDL cholesterol (mg/dL)	92.7 ± 34.1	122.4 ± 32.8	0.002
HDL cholesterol (mg/dL)	32.3 ± 9.3	43.0 ± 16.5	0.003
Serum homocysteine (μmol/L)	30.3 ± 32.7	18.1 ± 7.0	0.05
m-CIMT (mm)	1.18 ± 0.16	1.06 ± 0.13	0.003
Atherosclerotic plaque presence	5 (16 %)	6 (20%)	NS

CRP, C-reactive protein.

m-CIMT, mean carotid artery intima media thickness.

Inflammatory Parameters

Serum CRP, plasma fibrinogen, ESR, and complete blood count were used for assessment of inflammatory status. ESR and plasma fibrinogen levels were higher in CAPD patients than HD patients, however, CRP levels were similar in both groups. In addition, rate of patients with serum CRP level >10 mg/L was similar in both groups (13 patients in each group). Serum ferritin levels were also similar in the study groups.

Atherosclerosis Parameters

Lipid parameters, serum homocysteine level, and m‐CIMT were used to evaluate atherosclerosis. The presence of carotid artery plaques was also assessed. In five HD patients (16%), atherosclerotic plaques were found (in three patients at both carotid arteries, in two patients at only the left side) and in six CAPD patients (20%), atherosclerotic plaques were found (in four patients at both carotid arteries, in one patient at the left side, and in one patient at the right side). Rate of patients with atherosclerotic plaque was similar in both groups (P = 0.73). Patients who had carotid plaque had higher serum CRP levels as compared to patients without carotid plaque (35.1 ± 15.6 mg/l vs. 16.0 ± 3.3 mg/l, p = 0.04 (mean ± SE).

Despite similarity in serum vitamin B₁₂ and folic acid between HD and CAPD groups (for vitamin B₁₂ 507.0 ± 279.4 pg/mL vs. 482.5 ± 196.7 pg/mL, P = 0.69; and folic acid, 11.3 ± 6.4 ng/mL vs. 12.0 ± 6.3 ng/mL, P = 0.64, respectively), serum homocysteine level was higher in HD patients (30.3 ± 32.7 μmol/L vs. 18.1 ± 7.0 μmol/L, P = 0.05, respectively). Twenty-six patients (86%) in the HD group and 20 patients (66%) in the CAPD group were hyperhomocysteinemia (p = 0.06). Comparison results of HD and CAPD patients for nutrition, inflammation, and atherosclerosis parameters are given in Table 2.

Increased m-CIMT is accepted as a sign of early atherosclerosis. In this study, we found that m-CIMT was higher in HD patients than CAPD patients, despite CAPD patients having a more atherogenic lipid profile (1.18 ± 0.16 mm vs. 1.06 ± 0.13 mm, P = 0.003, respectively). Twenty-six HD patients and 19 CAPD patients had m-CIMT >1 mm (P = 0.03). However, among the atherosclerosis parameters, serum homocysteine level was found higher in HD patients than the CAPD group.

When the association between serum homocysteine and albumin levels in the whole study group was analyzed, none of the hypoalbuminemic patient was hyperhomocysteinemia, but 46/54 (85%) of normoalbuminemic patients were hyperhomocysteinemia (P = 0.0005). Despite this association, there was no statistically significant correlation between serum albumin and serum homocysteine levels (P = 0.74).

Correlations

We found positive correlations between serum albumin level and serum total cholesterol level (r = 0.33, P = 0.009), and hemoglobin level (r = 0.33, P = 0.009). There was a negative correlation between CRP level and serum albumin level (r = −0.59, P = 0.0004), and serum total cholesterol level (r = −0.26, p = 0.03). There was a positive correlation between serum homocysteine level and m-CIMT (r = 0.47, P = 0.0001). In addition, we found that serum homocysteine level was independent and a strong predictor of m-CIMT by multivariate analysis.

Only one HD patient and seven CAPD patients had no components of MIA syndrome. According to this point of view, CAPD modality was found to be superior to HD modality (P = 0.05). Comparison of both groups for MIA syndrome components is shown in Figures 1 and 2.

Figure 1 MIA syndrome components in HD patients.

Figure 2 MIA syndrome components in CAPD patients.

DISCUSSION

Although it has been over 40 years since chronic dialysis treatment was started, the mortality rate in ESRD patients remains several-fold higher as compared to the general population. A major cause of mortality in ESRD patients is predominantly due to cardiovascular reasons. Approximately 50% of the patients had lost their life as a consequence of cardiovascular diseases, due to accelerated atherosclerosis.[15],[16] It is well-documented that HD patients have had approximately 10-fold higher serum pro-inflammatory cytokine levels than the healthy population.[17] Several studies have linked high levels of pro-inflammatory cytokines with poor outcomes in ESRD patients. Therefore, MIA syndrome was defined as the interaction between increased pro-inflammatory cytokine, and malnutrition and atherosclerosis in ESRD, as suggested by Stevinkel et al.[11] In this study, we aimed to compare both renal replacement therapies according to MIA syndrome components and found that PD modality has superiority against HD modality in terms of MIA syndrome components.

Hypoalbuminemia is the most powerful predictor of death in these ESRD patients.[18] It was reported that the prevalence of malnutrition varies 18–76% in dialysis patients according to type of dialysis modality.[19–22] In the present study, only one patient in the HD group and five patients in the CAPD group were hypoalbuminemic. Malnutrition prevalence was higher in the CAPD group than the HD group, but not statistically significant. For the malnutrition component of MIA syndrome, HD modality seems to be more advantageous than CAPD modality.

End-stage renal disease is a chronic inflammatory state. Since the first report by Bergström et al.,[6] several groups have reported an association between elevated CRP level and mortality for both HD and CAPD patients.[7–10] In the present study, according to the inflammation component of MIA syndrome, CAPD and HD patients had similar serum CRP levels. In each group, 13 patients had serum CRP levels higher than 10 mg/dL, but other inflammatory markers (ESR and plasma fibrinogen levels) were higher in CAPD patients than HD patients. Recent published reports have given conflicting results about which dialysis modality had higher inflammatory response. Several studies have shown that many CAPD patients have evidence of inflammatory response.[9],[10],[23],[24] However, data presented by Haubitz et al.[25] and Takahashi et al.[26] has shown that CAPD patients have lower inflammatory status than HD patients. In the present study, both groups were similar for inflammatory status. Except for serum CRP level, other acute phase parameters, such as ESR and fibrinogen, were higher in CAPD patients than HD patients. Due to contaminated dialysate and/or bio-incompatible membranes, it is likely that these inflammatory mediators (ESR and plasma fibrinogen levels) could increase after the hemodialysis session. We measured these parameters before the hemodialysis session and this may explain lower ESR and fibrinogen levels in HD patients.

When considering serum CRP levels, HD seems to be similar to CAPD for inflammation. Mean-carotid artery intima media thickness measured with carotid ultrasonography is used as a non-invasive method to evaluate atherosclerosis. In autopsy study, a close correlation between coronary artery atherosclerosis and extent of atherosclerotic lesions in the carotid arteries has been revealed.[27] In a study by Hodis et al.[28] it was concluded that the absolute intima-media thickness and the progression of intima-media thickness predicted risk for coronary events beyond that predicted by coronary arterial measures of atherosclerosis and lipid parameters. In the present study, in spite of having more atherogenic lipid profile and a higher fibrinogen level, m-CIMT was lower in CAPD patients than HD patients, but serum homocysteine level was higher in HD patients. Likewise, in a study by Zoccali et al.[29] it was reported that HD patients have a higher fibrinogen level and increased m-CIMT than CAPD patients and serum homocysteine level was similar between both groups. In the same study, they reported increased prevalence of carotid artery plaque as the serum CRP level increased. We also found that patients with atherosclerotic carotid artery plaque had higher CRP levels than patients without plaque. Prevalence of carotid artery plaque has been reported as high as 72% in predialytic ESRD patients. Carotid plaque prevalence in our study was lower, probably due to younger patients and lower time on dialysis in the study group. In terms of atherosclerosis, CAPD treatment is superior, as compared to HD treatment.

Because hyperhomocysteinemia is an independent risk factor for development of atherosclerotic cardiovascular disease in the non-uremic population, it has received considerable attention in the last two decades. An elevated homocysteine level was detected in most of ESRD patients and correlated with the presence of vascular disease. The relationship between hyperhomocysteinemia and premature atherosclerosis is not well understood. Most studies suggest that homocysteine may enhance lipoprotein oxidation,[30] increase smooth muscle cell proliferation,[31] induce endothelial activation of factor V, and reduce protein C activation[32] by arterial and venous endothelial cells. It is now well-established that uremic patients have a high prevalence (84–92%) of hyperhomocysteinemia.[33–36] However, there are a few reports on comparing homocysteine levels between HD and CAPD patients. Moustapha et al.[37] investigated serum homocysteine levels and the prevalence of hyperhomocysteinemia in both HD and CAPD patients. They reported HD patients had a higher homocysteine level than CAPD patients (29.8 μmol/L vs. 19.9 μmol/L, P = 0.001). Also, prevalence of hyperhomocysteinemia in HD patients was higher than CAPD patients (90.8% and 67.4%). Consistent with this report, we also found higher serum homocysteine levels (30.3 ± 32.7 μmol/L vs. 18.1 ± 7.0 μmol/L, P = 0.05) and higher prevalence of hyperhomocysteinemia in HD patients in the present study (86% vs. 66%, P = 0.06).

In many studies, hyperhomocysteinemia is shown as a risk factor for atherosclerosis and association between hyperhomocysteinemia and m-CIMT was defined. Mallamaci et al.[38] reported that hyperhomocysteinemia predicts cardiovascular outcomes in hemodialysis patients. Zoccali et al.[29] showed positive correlation between CIMT, serum homocysteine level, and time on dialysis. Additionally, it was reported that serum homocysteine level was an independent predictor of both m-CIMT and presence of atherosclerotic plaque. Consistent with these results, we also found serum homocysteine level as an independent predictor of m-CIMT. Libetta et al.[39] showed the strong association correlation between plasma homocysteine and m-CIMT, and this correlation emphasizes the role of hyperhomocysteinemia as a major risk factor for atherosclerosis. However, it was reported that dialysis patients with cardiovascular disease had lower plasma homocysteine levels, as well as a greater prevalence of malnutrition and hypoalbuminemia, than those without cardiovascular diseases.[40–42] Plasma homocysteine levels reflected correlation with serum albumin levels and these findings are trying to be explained by a phenomenon of reverse epidemiology.[41] This suggests that serum homocysteine levels are also a marker for nutritional status of the patients.[42] In the present study, we also observed high serum homocysteine levels only in dialysis patients who had normal serum albumin levels, which supports the hypothesis.

Only one HD patient and seven CAPD patients had no components of MIA syndrome. Both groups were different for the distribution of MIA components. Stenvinkel et al.[12] reported that there was a relationship between the number of complications (malnutrition, inflammation, and atherosclerosis) and mortality rate in peritoneal dialysis patients. As the components of MIA syndrome increase, the morbidity and mortality increase and the survival rate decreases in peritoneal dialysis patients.[12] In the present study, we found that peritoneal dialysis modality has superiority to hemodialysis modality regarding prevalence of patients who had no components of MIA syndrome. However, due to the cross-sectional nature of the study and relatively small number of the patients, further prospective larger studies are needed to answer this important question.

In conclusion, hemodialysis modality seems to be more advantageous for malnutrition components than CAPD modality. Hemodialysis and CAPD seem to be similar for inflammation, and CAPD modality has superiority for atherosclerosis as compared to HD. Before choosing the type of renal replacement therapy, assessment of MIA syndrome components could be useful for individualization of the decision on which dialytic modality is appropriate in ESRD patients.