The study was based on a case control design and the cases were recruited from consecutive admissions to the coronary care unit of the General Hospital Kandy with the diagnosis of myocardial infarction. Kandy district is situated in the hilly central region of Sri Lanka where climatic conditions and vegetation are different from those in the low-lying plains in the island. The Cardio Coronary Unit, at the General Hospital Kandy is a tertiary care unit and it provides coronary care services to the entire Kandy district.

The cases presented with a history of chest pain, subsequently confirmed as myocardial infarction based on ECG changes and positive cardiac biomarker “Troponin T”. The Troponin-T test was performed using capillary blood as a bed side test using a standard kit (Roche CARDIAC Troponin T Sensitive^®) at the time of admission to the hospital and repeated 6 hours afterwards if the initial test was negative. The test was conducted according to the guide lines provided by the manufacturer. The positive Troponin T was considered essential along with ECG changes to confirm the diagnosis and recruitment. The patients who died immediately after the admission were excluded from the study. Controls were selected during the same period of time, among the patients who presented to the outpatient department of the same hospital with other ailments, unrelated to cardiovascular diseases. They were matched for age ± 2 years and gender. Finally the study comprised 205 cases and 197 controls for the analysis.

The study team trained two medical graduates for data collection. After obtaining consent from the selected patients, 5 ml of venous blood were taken within 12 hours of admission to measure their lipid profile, liver enzymes, serum albumin and total bilirubin levels. The reliability of laboratory information was ensured by retesting ten percent of duplicated blood samples in another reputed laboratory.

Once the patient’s general conditions were stable, details such as personal variables, information regarding known history of diabetes mellitus and hypertension, details on smoking habits, family history of ischaemic heart disease were collected using an interviewer-administered questionnaire. We defined a positive family history of IHD if a 1st degree male relative had IHD below 60 years of age or a female below 55 years of age. Anthropometric measurements were taken on third day, when patients were allowed to be in the standing position. Weight, height, waist circumference and hip circumference of each patient were measured by the trained interviewers using the metric units. Waist circumference was measured at the narrowest circumference between the costal margin and iliac crest. Hip circumference was taken at the maximum circumference over the buttocks. Blood pressure readings were taken from the records as they were measured regularly using the mercury sphygmomanometers.

The data were initially entered into a Microsoft Excel work sheet and then analyzed using the statistical software package, SPSS version 17. Univariate analysis was performed with the data of a categorical nature to assess the associations between the occurrence of MI and independent variables. Logistic regression was used to find out the independent risk factors of MI.

Associations between the cases and controls with regard to their liver enzymes [serum aspartate transaminase (AST), alanine transaminase (ALT)], serum albumin, serum bilirubin levels were analyzed separately. For the purpose of categorization of these values, cutoff values given by the manufactures were used. Odds ratio and trends were calculated to assess the associations. Association between serum bilirubin levels among cases and controls were analyzed using bilirubin fifth.