Homocysteine and IgA nephropathy: observational and Mendelian randomization analyses.

BACKGROUND

High levels of plasma homocysteine occur almost uniformly in patients with end-stage renal disease (ESRD). IgA nephropathy (IgAN) is the most common form of primary glomerulonephritis and a common cause of ESRD in young adults. Here, we aimed to detect whether homocysteine was elevated and associated with clinical-pathologic manifestations of IgAN patients and tested its causal effects using a two-sample Mendelian randomization (MR) approach.

METHODS

For observational analysis, 108 IgAN patients, 30 lupus nephritis (LN) patients, 50 minimal change disease (MCD) patients, and 206 healthy controls were recruited from April 2014 to April 2015. Their plasma homocysteine was measured and clinical-pathologic manifestations were collected from medical records. For MR analysis, we further included 1686 IgAN patients. The missense variant methylenetetrahydrofolate reductase C677T (rs1801133) was selected as an instrument, which was genotyped by TaqMan allele discrimination assays.

RESULTS

Majority of IgAN patients (93.52%, 101/108) showed elevated levels of plasma homocysteine (>10 μmol/L). Plasma homocysteine in IgAN patients was significantly higher than that in MCD patients (median: 18.32 vs. 11.15 μmol/L, Z = -5.29, P < 0.01) and in healthy controls (median: 18.32 vs. 10.00 μmol/L, Z = -8.76, P < 0.01), but comparable with those in LN patients (median: 18.32 L vs. 14.50 μmol/L, Z = -1.32, P = 0.19). Significant differences were observed in sub-groups of IgAN patients according to quartiles of plasma homocysteine for male ratio (22.22% vs. 51.85% vs. 70.37% vs. 70.37%, χ = 14.29, P < 0.01), serum creatinine (median: 77.00 vs. 100.00 vs. 129.00 vs. 150.00 μmol/L, χ = 34.06, P < 0.01), estimated glomerular filtration rate (median: 100.52 vs. 74.23 vs. 52.68 vs. 42.67 mL·min·1.73 m, χ = 21.75, P < 0.01), systolic blood pressure (median: 120.00 vs. 120.00 vs. 125.00 vs. 130.00 mmHg, χ = 2.97, P = 0.05), diastolic blood pressure (median 80.00 vs. 75.00 vs. 80.00 vs. 81.00 mmHg, χ = 11.47, P < 0.01), and pathologic tubular atrophy and interstitial fibrosis (T) (T0/T1/T2: 62.96%/33.33%/3.70% vs. 29.63%/40.74%/29.63% vs. 24.00%/48.00%/28.00% vs. 14.81%/37.04%/48.15%, χ = 17.66, P < 0.01). The coefficient of each rs1801133-T allele on homocysteine levels after controlling age and sex was 7.12 (P < 0.01). MR estimates showed causal positive effects of homocysteine on serum creatine (β = 0.76, P = 0.02), systolic blood pressure (β = 0.26, P = 0.02), diastolic blood pressure (β = 0.20, P = 0.01), and pathologic T lesion (β = 0.01, P = 0.01) in IgAN.

CONCLUSIONS

By observational and MR analyses, consistent results were observed for associations of plasma homocysteine with serum creatinine, blood pressures, and pathologic T lesion in IgAN patients.