The study was done on patients with type 2 diabetes with or without macrovascular complications such as coronary artery disease. Patients were recruited from three multispecialty hospital centers in Southern India. All the protocols were submitted to and approved by the Human Ethics Committee of Rajiv Gandhi Centre for Biotechnology and participating hospitals (Indian Institute of Diabetes, PRS Hospital and Madras Medical Mission). A written informed consent was obtained from all study subjects. All patients subsequently underwent a standardized clinical and laboratory evaluation.
The study subjects (n = 556) were divided into five groups: (i) Patients with type 2 diabetes mellitus (DM; n = 101), (ii) diabetes patients with coronary artery disease (DM + CAD) diagnosed within 5 years of detection of type 2 diabetes (n = 103), (iii) patients with DM who had CAD diagnosed five years after detection of diabetes (n = 109), (iv) patients with CAD (n = 122) and without diabetes and (v) normal healthy volunteers (n = 121).
Patients with hypertension, cardiac diseases other than coronary artery disease, peripheral vascular disease, thrombotic stroke, nephropathy, retinopathy, inflammatory disease of any cause, subjects with other systemic and metabolic diseases, asthma, malignancy, liver diseases, kidney diseases and pregnant women were excluded from the study. Diabetes was assessed by recording HbA1c and/or fasting blood sugar (FBS) levels while CAD was diagnosed by a positive treadmill test and/or coronary angiography.
Enzyme linked assay for measurement of cyclophilin A in plasma
Blood samples were collected after overnight fasting. Blood was collected into EDTA containing tubes and centrifuged for 7 minutes at 2500 g within 30 minutes of collection. The separated plasma was transferred to a fresh tube and used directly for the immunoassay as per the supplier’s protocol. Cyclophilin A levels in plasma were determined with a sandwich immunoassay kit (Uscn Life Science Inc, Product No. SEA979Hu). The linearity of the kit was assayed by testing samples spiked with a known concentration of cyclophilin A and their serial dilutions. Spiking with known concentrations of the protein guaranteed a recovery range of 83–102%. No significant cross-reactivity or interference with any other proteins was observed. All samples were analyzed in duplicate. To maintain assay precision, samples with a CV > 12% were excluded.
C-Reactive Protein (CRP) was measured with a quantikine sandwich Human CRP Immunoassay (R & D Systems, Product No: DCRP00). A 93-100% recovery range was observed when samples were spiked with 1:2 to 1:16 concentrations of CRP.
Medication history was recorded for 270 subjects of the 556 subjects. Medications were grouped into (i) antiplatelet aggregating agents (Ecospirin, Clopidogrel, Cilostazol), (ii) Antihypertensive agents (Calcium channel blockers, ACE inhibitors, vasodilators), (iii) Statins (atorvastatin, Lovastatin) and (iv) Metformin derivatives (glucophage, glycimet).
The distribution of all variables in the five groups was studied by describing the mean, median, range and standard deviation. Predictor variables such as age, HbA1C, fasting blood sugar (FBS) were grouped into two based on median values; cyclophilin A levels in the two groups compared using student t test. The same procedure was used for comparing cyclophilin levels in groups with different levels of CRP, and in patients using metformin or not. Cyclophilin values in the five groups were compared with ANOVA (F = 54.75, p < 0.001), followed by multiple comparisons using pairwise t tests with pooled variance by the Holms’ method. We did a multinomial logistic regression analysis for estimating the prevalence odds ratios for presence of disease, with the normal subjects as the reference. The odds for prevalence of the other four conditions, i.e., diabetes and no heart disease (DM), heart disease and no diabetes (CAD), diabetes and heart disease of up to 5 years duration (DM + CAD 5Y), and diabetes and heart disease beyond 5 years but below 10 years (DM + CAD 10Y), among those subjects with high cyclophilin levels were compared with prevalence odds in those with low cyclophilin values, adjusted for age and sex. All statistical tests were done with an estimated power =0.80. All statistical tests were done in R (R Core Team , 2013, R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.); graphs were also created in R.