Calcium-phosphorus product is associated with non-alcoholic fatty liver disease in type 2 diabetes mellitus
Background: In the past decades, there is a markedly raise in the prevalence of type 2 diabetes mellitus (T2DM) and non-alcoholic fatty liver disease (NAFLD), causing increased global health burden. A large proportion of T2DM was complicated with NAFLD (40%-70%), and this strong association between these two diseases could possibly be owing to the pathophysiological mechanism they shared: insulin resistance. Recent studies demonstrate that insulin resistance was associated with serum calcium and phosphorous levels, so we hypothesize that prevalence of NAFLD in T2DM could be related to serum calcium and phosphorous. 

Aim: To determine whether calcium and phosphorus levels are associated with the prevalence of non-alcoholic fatty liver disease in patients of T2DM in China.

Method: 2715 patients with T2DM who were admitted to the ward of the department of Peking University People’s Hospital between January 2012 and February 2017 were enrolled in this study. Patients aged ≥18 years without reported history of heart failure, hypoparathyroidism, hyperparathyroidism, excessive alcoholic intake (man ≥30 (g/day), woman≥ 20 (g/day)) were included and patients with positive hepatitis B surface antigens or hepatitis C antibodies and estimated glomerular filtration rate ﹤60 ml/min/1.73m2 based on laboratory test were excluded. 1805 patients were included in the final analysis. Clinical characteristics were compared between NAFLD and non-NAFLD participants. Logistic regression models were used to detect the association between calcium-phosphorus product and the prevalence of NAFLD.

Result:
The prevalence of NAFLD is 52.2% in 1805 T2DM patients (mean age 56.2±13.3 years). Compared with the group without NAFLD, individuals combined with NAFLD were younger(54.4±13.6 vs. 58.9±12.5 (years), P<0.001), had shorter duration of diabetes(8.3 (3.0-14.0) vs. 11.0 (6.0-18.0) (years), P<0.001) and similar gender composition (male gender, 53.6% vs. 51%, P=0.309) between the two groups. Individuals with NAFLD have higher levels of BMI (27.23±3.81 vs. 23.85±3.16 (kg/m2), P<0.001), HbA1c (9.2% (7.8%-10.7%) vs. 8.9% (7.4%-10.7%), P=0.007), fasting plasma glucose (7.37 (6.01-9.40) vs. 6.91 (5.58-8.83) (mmol/L), P=0.001), triglyceride (1.86 (1.36-2.6) vs. 1.22 (0.91-1.64) (mmol/L), P<0.001) and uric acid (351±93 vs. 305±79 (umol/L), P<0.001). Additionally, high density lipoprotein cholesterol was lower (0.92 (0.80-1.06) vs. 1.04 (0.88-1.26) (mmol/L), P<0.001). Calcium-phosphorus product was higher in NAFLD group compared with the group without NAFLD (35.25±5.70 vs. 33.81±5.88 (mg2/dl2), P<0.001). There was no significant difference in vitamin D (37.77 (30.27-48.32) vs. 39.70 (30.72-51.42) (nmol/L), P=0.058) and parathyroid hormone (PTH) (32.30 (25.10-41.27) vs. 31.82 (24.13-39.57) (pmol/L), P=0.056) level between NAFLD group and non-NAFLD group.
In order to demonstrate the association between calcium-phosphorus product and NAFLD, individuals were divided into quartiles according to calcium-phosphorus product levels. The prevalence of NAFLD was 55.2%, 56.5%, 63.3%, 70.3% in Q1 (the first quartile, <30.73 (mg2/dl2)), Q2 (the second quartile, 30.73-34.44 (mg2/dl2)), Q3 (the third quartile, 34.44-38.65 (mg2/dl2)), Q4 (the fourth quartile, ≥38.65 (mg2/dl2)). Calcium-phosphorus product was significantly associated with the presence of NAFLD (odds ratio (OR): 1.029 (1.008-1.051)), after adjusting sex, age, SBP, BMI, HbA1c and TG. In multiple logistic regression model, the ORs for the prevalence of NAFLD were 1.097 (0.810-1.487) in Q2 group, 1.453 (1.062-1.988) in Q3 group, 1.766 (1.266-2.465) in Q4 group (P for trend＜0.001) in reference to Q1 group. Similar results were found after vitamin D and PTH was adjusted.
Discussion: Increasing level of calcium-phosphorus product is associated with the presence of NAFLD in this Chinese hospitalized T2DM population. This could be possibly related to the role of calcium and phosphorus metabolism in insulin resistance.