Association of Reduced Renal Function with Hepatitis B Virus Infection and Elevated Alanine Aminotransferase

Summary

Background and objectives

Clinically, hepatitis B virus (HBV) infection is observed to be associated with nephropathy. However, previous population-based studies failed to show an association between HBV infection and CKD. Therefore, this cross-sectional study was designed to further explore this association.

Design, setting, participants, & measurements

A representative sample of 6854 Chinese adults aged 30–75 years was tested for levels of serum hepatitis B surface antigen, alanine aminotransferase (ALT), creatinine, urinary albumin/creatinine ratio, and potential CKD risk factors.

Results

Neither HBV infection nor elevated ALT (ALT+; ≥ sex-specific 90th percentile of ALT levels of noninfected persons) was significantly associated with reduced estimated GFR (eGFR < 60 ml/min per 1.73 m²). Compared with noninfected persons, HBV-infected persons with ALT+, but not those with ALT− (P=0.26), were more likely to have reduced eGFR (odds ratio, 4.07; 95% confidence interval, 1.18–14.0; P=0.03). Further analysis with a general linear model revealed a significant difference in eGFR (mean ± SEM) between HBV-infected and noninfected persons (87.8±0.8 versus 90.2±0.4 ml/min per 1.73 m²; P=0.002). This difference was mainly derived from that between HBV-infected persons with ALT+ and noninfected persons, with an average difference in eGFR of −4.5 (95% confidence interval, −0.9 to −8.1; P=0.01). HBV infection and ALT+, alone or in combination, were not significantly associated with albuminuria or CKD.

Conclusions

HBV infection with elevated ALT, rather than HBV infection alone, was associated with reduced renal function.

Introduction

CKD has a prevalence of 10.3%–13.7% in different countries (1), and its incidence is rising throughout the world (2,3). CKD increases the risk for cardiovascular diseases and death (4). Thus, CKD imposes a heavy burden on global public health, and identifying its risk factors is essential.

Hepatitis B virus (HBV) infection is also a public health problem. Worldwide, >350 million people are chronic HBV carriers (5). Apart from liver involvement, HBV infection can lead to HBV-associated nephropathy (6,7), yet population-based or case-control studies have not shown the expected association between HBV infection and reduced renal function, CKD, or proteinuria (8–11). However, a case-control study revealed that urine albumin excretion in children with hepatitis B was significantly higher than in inactive HBV carriers and non-HBV carriers (12). ALT is an indicator for hepatitis B in HBV-infected persons (12,13); we thus hypothesized that HBV infection with elevated ALT rather than HBV infection alone may be associated with reduced renal function or albuminuria.

To test this hypothesis, we explored the association of HBV infection with renal function and albuminuria in a large cross-sectional study in a suburb of Beijing.

Materials and Methods

Study Population and Sampling Method

This study was performed between April 2008 and March 2009 in Pinggu District, a suburb of Beijing. The target population was residents aged 30–75 years. Multistage clustered sampling was used to recruit study participants, with the probability proportional to population size in each stage. In stage 1, 4 of 18 towns were selected. In stage 2, 5 villages were selected from each town, which resulted in the selection of 20 of 76 villages and a total of 8189 eligible households. One person was selected from each household with a randomly allocated Kish grid. Ultimately, 6925 (84.6%) people agreed and participated.

The following participants were excluded from analyses: a renal transplant recipient with positive serum hepatitis B surface antigen (HBsAg), 23 participants with missing serum creatinine measurements, and 47 with missing data on any of the confounding variables. Furthermore, 23 participants with missing urinary albumin measurements were excluded from the CKD and albuminuria analyses. Eventually, 6854 (99%) participants were included in the renal function analyses and 6831 (99%) in the CKD and albuminuria analyses.

This study was approved by the local ethics committee. Written informed consent was obtained from each participant before data collection.

Data Collection and Measurements

Trained medical staff used a questionnaire to collect data on age, sex, education, smoking, and history of hypertension, diabetes, kidney disease, nephrolithiasis (14), and nephrotoxic medications, including nonsteroidal anti-inflammatory drugs and herbs containing aristolochic acid.

Waist circumference and BP were measured using standard protocols and techniques. Venous blood samples were collected after an overnight fast of 8–12 hours. All blood and urinary samples were tested at the Peking Union Medical College Hospital laboratory.

HBsAg and hepatitis C virus (HCV) antibody were detected by chemiluminescence (Abbott Architect i2000SR). Serum HBsAg positivity was considered to be HBV infection (15), and is herein designated as HBV+ (or as HBV− otherwise). Fasting blood glucose, serum lipids, and ALT levels were measured using an Olympus AU5400 autoanalyzer (Olympus, Japan). Considering the sex disparity in the upper limit of normal (ULN) for ALT and the controversies over the ALT value that defines hepatitis (16,17), the following three sets of sex-specific ALT cutoffs were used: (1) the sex-specific 90th percentile values of noninfected persons in our study (28 U/L for women and 40 U/L for men); (2) 1.3 times the ULN proposed by the American Association for the Study of Liver Disease (AASLD) (25 U/L for women and 39 U/L for men) (15); and (3) the ALT cutoff used in the Third US National Health and Nutrition Examination Survey III (NHANES III) (31 U/L for women and 40 U/L for men) (18). Elevated ALT was defined as an ALT value above the cutoffs, and designated as ALT+ (or as ALT− otherwise).

Diabetes was defined as a fasting serum glucose concentration >7.0 mmol/L or a previous diagnosis of diabetes (19). Hypertension was defined as systolic BP ≥140 mmHg or diastolic BP ≥90 mmHg or by the use of antihypertensive medications in the previous 2 weeks, irrespective of BP. Central obesity was defined as waist circumference ≥90 cm in men or ≥80 cm in women.

Outcome Measures

First-void, mid-stream morning urine samples were collected to measure the urinary albumin/creatinine ratio (ACR). Urine albumin was measured by using immunoturbidimetric methods and serum or urinary creatinine by means of the Jaffè kinetic method on an Olympus AU5400 autoanalyzer. Albuminuria was defined as a urinary ACR ≥3.39 mg/mmol (30 mg/g), and macroalbuminuria as an ACR ≥33.9 mg/mmol (300 mg/g) (20). Considering the imperfection in equation-based GFR estimation, the following three equations were used to calculate the estimated GFR (eGFR): (1) the Chinese eGFR Investigation Collaboration equation: eGFR = 175 × Scr^−1.234 × age^−0.179 × 0.79 (if female) (21); (2) the Modification of Diet in Renal Disease (MDRD) equation: eGFR = 186 × Scr^−1.154 × age^−0.203 × 0.742 (if female) (22); and (3) the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation: eGFR = 141 × min(Scr/κ,1)^α × max(Scr/κ,1)^−1.209 × 0.993^Age × 1.018 (if female). For men, κ=0.9 and α =−0.411, whereas for women, κ=0.7 and α=−0.329. In addition, min(Scr/κ,1) = the minimum of Scr/κ or 1, whereas max(Scr/κ,1) = the maximum of Scr/κ or 1 (23). As part of this, 97 serum samples with creatinine ranging from 0.50 to 10.2 mg/dl were retested by an enzymatic method and the following calibrating equation was developed (R²=1.0): calibrated Scr = 0.0128 × Scr − 0.288. In the CKD-EPI equation, calibrated Scr was used. Reduced eGFR was defined as eGFR <60 ml/min per 1.73 m² and CKD was defined as the presence of albuminuria and/or reduced eGFR (24).

Statistical Analyses

Population characteristics were expressed as mean (±SD) or median (interquartile range) for continuous variables and as frequencies or percentages for category variables. Characteristics of the three groups were compared with variance analyses for continuous variables and chi-squared tests for categorical variables.

Primarily, the sex-specific 90th percentile values for ALT were adopted as the ALT cutoff and eGFR was estimated by the Chinese eGFR Investigation Collaboration equation.

Logistic regression was used to calculate the odds ratios of having reduced eGFR, albuminuria, macroalbuminuria, and CKD, adjusted for potential confounders of age (as continuous variable), sex, diabetes, hypertension, waist circumference (central obesity or not), history of nephrolithiasis, and levels of total cholesterol (≥ or <200 mg/dl) and HDL cholesterol (≥ or < 40 mg/dl). First, the odds ratios of having reduced eGFR, albuminuria, and CKD for HBV infection and ALT+ were calculated. The study population was then divided into the following three groups according to HBV infection status and ALT profile: HBV-infected persons with ALT+, HBV-infected persons with ALT−, and noninfected persons. The odds of having reduced eGFR, albuminuria, and CKD relative to noninfected persons were calculated.

To confirm our findings, we also used general linear models to calculate and compare mean eGFR among HBV-infected persons with ALT+, HBV-infected persons with ALT−, and noninfected persons, adjusted for the aforementioned confounders. Multiple comparisons were adjusted using the Bonferroni correction.

Sensitivity analyses were performed using the MDRD or CKD-EPI equation to estimate eGFR, or 1.3 times the AASLD ULN or the NHANES III cutoffs as the ALT cutoffs, or albuminuria defined by sex-specific Chinese ACR cutoffs (2.79 mg/mmol for men and 3.9 mg/mmol for women) (25).

All data were analyzed using the SPSS statistical software package (version 14.0; SPSS Inc, Chicago, IL). All tests were two tailed. For multiple comparisons in general linear models, P<0.02 was considered statistically significant; for all other analyses including those in logistic regression, P<0.05 was considered statistically significant.

Results

Characteristics of the study population by HBV status and ALT levels are summarized in Table 1. The prevalence of HBV infection was 4.8% and no HBV-infected person was found to be coinfected by HCV. Only nine HBV-infected persons, whose ALT levels were normal during the study, were ever treated with IFN or nucleoside or nucleotide analogs. The groups differed significantly by mean eGFR (P=0.03) and frequency of diabetes and hypercholesterolemia (P=0.02 and P<0.001, respectively). Reduced eGFR was more frequent in HBV-infected persons with ALT+ than in noninfected persons (3.7% versus 1.4%; P=0.05).

Table 1.

Characteristics of a representative sample of 6854 people from a suburb of Beijing, China, by hepatitis B status and alanine aminotransferase levels

Characteristic	Total (n=6854)	Noninfected Persons (n=6526)	HBV-Infected Persons		Pa
			With ALT− (n=247)	With ALT+ (n=81)
Age (yr)	50.7±10.5	50.7±10.5	51.2±9.8	48.6±9.5	0.15
Male sex	3425 (50)	3245 (49.7)	136 (55.1)	44 (54.3)	0.19
Hypertension	3440 (50.2)	3289 (50.4)	110(44.3)	41 (50.6)	0.20
Diabetes	613 (8.9)	572 (8.8)	27 (10.9)	14 (17.3)	0.02
Waist circumference (cm)	86.1±10.2	86.1±10.2	85.5±10.2	87.1±11.7	0.43
ALT, U/L	17 (13–24)	17 (13–23)	19 (15–25)	46 (36–57)	<0.001
eGFR, ml/min per 1.73 m2	91.1±14.6	91.2±14.6	89.6±12.4	87.8±16.9	0.03
eGFR <60 ml/min per 1.73 m2	95 (1.4)	91 (1.4)	1 (0.4)	3 (3.7)	0.11
HDL cholesterol <40 mg/dl	1085 (15.8)	1021 (15.6)	46 (18.6)	18 (22.2)	0.13
Total cholesterol ≥200 mg/dl	2708 (39.5)	2617 (40.1)	64 (25.9)	27 (33.3)	<0.001
Triglycerides ≥200 mg/dl	1092 (15.9)	1049 (16.1)	29 (11.7)	14 (17.3)	0.18
History of nephrolithiasis	509 (7.4)	484 (7.4)	19 (7.7)	6 (7.4)	0.99
Illiteracy	773 (11.3)	743 (11.4)	27 (10.9)	3 (3.7)	0.12
Smoking	2924 (42.7)	2772 (42.5)	118 (47.8)	34 (42.0)	0.26
Nephrotoxic medication	1287 (18.8)	1229 (18.8)	41 (16.6)	17 (21.0)	0.60
Albuminuria	731 (10.7)	703 (10.8)	19 (7.7)	9 (11.1)	0.31
Macroalbuminuria	63 (0.9)	60 (0.9)	1 (0.4)	2 (2.5)	0.27
CKD	777 (11.4)	748 (11.5)	20 (8.1)	9 (11.1)	0.27

Data are mean ± SD, n (%), or median (interquartile range), unless otherwise indicated. ALT+, ALT levels ≥ ALT cutoff (the sex-specific 90th percentile ALT values of noninfected persons in our study [28 U/L for women and 40 U/L for men]); ALT−, ALT levels < ALT cutoff. HBV, hepatitis B virus; ALT, alanine aminotransferase; eGFR, estimated GFR.

^aComparisons among the three subgroups were conducted with univariate variance analysis for continuous variables and chi-squared tests for categorical variables.

After adjusting for potential confounders, neither HBV infection nor ALT+ was significantly associated with reduced eGFR, albuminuria, and CKD, no matter what ALT cutoff or eGFR equation was adopted (only data based on the 90th percentile ALT cutoff and the eGFR derived from the Chinese eGFR Investigation Collaboration equation shown in Table 2).

Table 2.

Association of HBV infection and elevated ALT with reduced eGFR, albuminuria, and CKD

	HBV Infection	ALT+
Reduced eGFR
Model 1a	1.04 (0.37–2.93)	1.22 (0.64–2.35)
Model 2b	1.15 (0.41–3.30)	1.26 (0.65–2.44)
Albuminuria
Model 1a	0.79 (0.53–1.19)	1.15 (0.58–2.28)
CKD
Model 1a	0.77 (0.52–1.19)	1.12 (0.89–1.42)

Data are odds ratios (95% confidence intervals). Multivariate logistic regressions were used to estimate odds ratios. All were P<0.05. ALT+, ALT levels ≥ the sex-specific 90th percentile values of noninfected persons in our study (28 U/L for women and 40 U/L for men); reduced eGFR, eGFR estimated with the Chinese eGFR Investigation Collaboration equation <60 ml/min per 1.73 m². HBV, hepatitis B virus; ALT, alanine aminotransferase; eGFR, estimated GFR.

^aModel 1: Adjusted for age, sex, hypertension, diabetes, waist circumference, HDL cholesterol levels, total cholesterol levels, and nephrolithiasis;

^bModel 2: Adjusted as in for Model 1, plus albuminuria.

The association of HBV infection with and without elevated ALT with reduced eGFR is shown in Table 3. HBV-infected persons with ALT+ were 4.07 times as likely to have reduced eGFR as noninfected persons (95% confidence interval, 1.18–14.0; P=0.03), but there was no statistically significant difference in the odds between HBV-infected persons with ALT− and noninfected persons (P=0.26). Neither HBV infection with ALT+ nor HBV infection with ALT− was significantly associated with albuminuria or CKD compared with non-HBV-infected persons. Similar outcomes were found if eGFR was estimated by the MDRD or the CKD-EPI equation, or the ALT cutoffs were defined as 1.3 times the AASLD ULN or the NHANES III cutoffs, or albuminuria was defined by the sex-specific Chinese eGFR Investigation Collaboration ACR cutoffs. If the nine participants ever treated with antiviral medications were excluded, the results remained unchanged (data not shown).

Table 3.

Association of HBV infection with and without elevated ALT with reduced eGFR

	HBV Infection with ALT+ (n=81)		HBV Infection with ALT− (n=247)
	OR (95% CI)	P	OR (95% CI)	P
Using the Chinese eGFR Investigation Collaboration equation to estimate eGFR
And the 90th percentile as the ALT cutoff (28 U/L for women and 40 U/L for men)
Model 1	4.07 (1.18–14.0)	0.03	0.32 (0.04–2.33)	0.26
Model 2	4.59 (1.33–15.9)	0.02	0.34 (0.05–2.54)	0.29
And 1.3 times the AASLD ULN as the ALT cutoff (25 U/L for women and 39 U/L for men)
Model 1	3.41 (1.01–11.4)	0.05	0.33 (0.05–2.44)	0.28
Model 2	3.97 (1.17–13.5)	0.03	0.35 (0.05–2.62)	0.31
And the ALT cutoff adopted by NHANES III (31 U/L for women and 40 U/L for men)
Model 1	5.36 (1.56–18.4)	0.01	0.30 (0.04–2.22)	0.23
Model 2	5.55 (1.58–19.5)	0.01	0.33 (0.04–2.44)	0.28
Using MDRD equation to estimate eGFR and the 90th percentile as the ALT cutoff
Model 1	3.22 (1.08–9.58)	0.04	0.19 (0.03–1.35)	0.10
Model 2	3.40 (1.15–10.1)	0.03	0.21 (0.03–1.50)	0.12
Using CKD-EPI equation to estimate eGFR and the 90th percentile as the ALT cutoff
Model 1	5.66 (1.27–25.2)	0.02	0.40 (0.05–2.95)	0.37
Model 2	6.40 (1.32–31.1)	0.02	0.45 (0.06–3.43)	0.44

Logistic regressions were used to estimate odds ratios and the comparison group is non-HBV-infected persons. Model 1: Adjusted for age, sex, hypertension, diabetes, waist circumference, HDL cholesterol levels, total cholesterol levels, and nephrolithiasis. Model 2: Adjusted as in for model 1, plus albuminuria. ALT+, ALT levels ≥ cutoffs; ALT−, ALT levels < cutoffs; reduced eGFR, eGFR < 60 ml/min per 1.73 m². HBV, hepatitis B virus; ALT, alanine aminotransferase; eGFR, estimated GFR; OR, odds ratio; 95% CI, 95% confidence interval; AASLD, American Association for the Study of Liver Disease; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; NHANES III, Third US National Health and Nutrition Examination Survey; ULN, the upper limit of normal; MDRD, Modification of Diet in Renal Disease Study.

In general linear models, after adjusting for potential confounders, eGFR (mean ± SEM, ml/min per 1.73 m²) was significantly lower in HBV-infected persons than in noninfected persons (87.8±0.8 versus 90.2±0.4; P=0.002), but did not differ significantly between participants with ALT+ and those with ALT− (P=0.42). Further analysis revealed that the eGFR difference between HBV-infected persons and noninfected persons was mainly derived from the difference between HBV-infected persons with ALT+ and noninfected persons, with an average difference of −4.5 ml/min per 1.73 m² (95% confidence interval, −0.9 to −8.1; P=0.01). Secondary analyses, using 1.3 times the AASLD ALT ULN and the NHANES III ALT cutoff, respectively, as the ALT cutoff, or modeling the logarithm of serum creatinine or eGFR estimated by the MDRD and the CKD-EPI equations, respectively, all resulted in similar trends (data not shown).

Discussion

In Chinese adults aged 30–75 years, we found that HBV infection with elevated ALT, rather than HBV infection or elevated ALT alone, was significantly associated with reduced renal function. HBV infection and elevated ALT, alone or in combination, were not significantly associated with albuminuria or CKD.

HBV infection is a public health challenge, especially in China, where its overall prevalence is as high as 7% (26). Clinically, HBV infection is observed to be associated with nephropathy (6). In patients with diabetes, it also poses additional risk of developing ESRD (27). However, as in previous population-based or case-control studies (8–11), we failed to test the statistically significant association of HBV infection with albuminuria, macroalbuminuria, reduced eGFR, or CKD. In further analysis, we found a significant association between reduced eGFR and HBV infection with ALT+, and a tendency of higher frequency of macroalbuminuria in HBV-infected persons with ALT+ (Table 1), albeit insignificant. Kilic et al. (28) also associated active hepatitis B with higher urine albumin excretion in children, which was alleviated by effective anti-HBV therapy.

In HBV-infected persons, ALT is usually an indicator for hepatitis B (12,13). Our finding and that of Kilic et al. (28) thus indicate that hepatitis B rather than HBV infection alone may be associated with renal involvement. For this, there are some potential explanations. First, in hepatitis B patients, the host’s immune response to HBV is active (12,13), and plays a major role in the pathogenesis of both hepatitis B and HBV-associated nephropathy (6,7,12,13). Deposition of immune complex of HBV antigen-antibody is universal in HBV-associated nephropathy. It leads to both glomerular and interstitial tubular damage, clinically presenting as macroalbuminuria and decline in renal function (6,7). Second, HBV replication in hepatitis B patients (12,13) may promote apoptosis of renal tubular cells and subsequently contributes to renal function deterioration (29). Third, chronic liver injury may elicit insulin resistance and potentiate the deleterious effect of diabetes on the kidney (27,30).

Our finding that HBV infection was associated with reduced renal function independently of albuminuria may broaden our concept of HBV-associated renal involvement. In turn, HBV-associated renal involvement may be more common than expected and should be taken more seriously. This finding may also partially account for the high percentage of chronic renal failure in adults with HBV-associated nephropathy (about 30%) (31). However, because this was a cross-sectional study with a relatively small number of persons with reduced renal function (eGFR <60 ml/min per 1.73 m²), a large-scaled cohort study is required to confirm our conclusion.

Our study has some potential limitations. First, its cross-sectional design makes it difficult to infer causality and our finding may be challenged by the clinic-based studies in which ALT levels were lower in patients with chronic renal failure (32). However, our study, as well as another population-based study (10), showed no difference in ALT levels between persons with and without reduced eGFR. This disparity may be ascribed to the fact that the persons with reduced eGFR among the general population (median creatinine in our study, 1.4 mg/dl) usually have renal function much better than the patients with chronic renal failure in the clinic-based studies (median creatinine ≥7.9 mg/dl) (32). Furthermore, in our study, ALT levels in HBV-infected persons with ALT+ were much higher than the others (Table 1). Thus, our finding should not be just a false positive. Second, the absence of an ALT cutoff definitively indicating hepatitis or active immune response to HBV, and the imperfection in equation-based estimation of renal function, may bring uncertainty into the interpretation of our findings. However, in our study, no matter which eGFR equation or ALT cutoff was used, consistent outcomes were achieved. Hence, our findings should be reliable. Third, because HBV DNA could not be measured, we could not rule out that in some HBV-infected persons, high ALT levels were caused by factors other than HBV infection, such as sex, age, obesity, dyslipidemia, HCV infection, and history of medication (including antiviral medication), which may concomitantly affect the kidney and liver. Our findings were adjusted for all of these factors. Fourth, because only a single measurement of HBsAg was obtained from each participant, we could not differentiate acute from chronic HBV infection. Nevertheless, acute HBV infection usually has a negligible percentage in persons with HBsAg positivity (33) and accordingly its contribution to our findings may also be negligible.

In conclusion, HBV infection with elevated ALT, rather than HBV infection or elevated ALT alone, was significantly associated with reduced renal function.

Disclosures

None.