This was a hospital-based unmatched case control study, conducted in the largest tertiary care hospital in Gampaha district during the period of 1st of June to 31st of August 2019. The hospital has four medical wards, five surgical wards, and three obstetrics and gynecology wards. The case group was the patients admitted to Medical, Surgical, and Gynecology wards and who were newly diagnosed to have fatty liver disease or diagnosed as a fatty liver disease within the last 6 months or less by ultrasonography.
NAFLD was diagnosed based on the presence of fatty liver according to established ultrasound criteria which were increase echogenicity of the liver compared to kidney and spleen, obliteration of vascular architecture, and deep attenuation of ultrasonic signals. Presences of two out of above three criteria were taken to diagnose as NAFLD.15 It was diagnosed by four well-trained medicals officers at the radiology department in the hospital. The control group was the patients admitted to the same units and who were not diagnosed as a fatty liver disease by ultrasonography. The exclusion criteria for both cases and controls were the patients who consumed alcohol above the safe limit among current drinkers and formal drinkers, pregnant women, pediatric age groups (≤ 14 years of age), patients who were diagnosed as hepatitis, cirrhosis, liver metastasis and liver cancers, cholestasis, alcoholic liver disease, and dengue hemorrhagic fever.
Sample size calculation was done based on odds ratio as 1.96 for elevated triglyceride for NAFLD,11 the anticipated proportion of elevated triglyceride among controls as 46%, power as 80%, and with 5% significant level. The control per case ratio was 2:1. According to the calculation, a minimum of 101 cases and 202 controls were needed. After adding 5% for the non-responders, it was decided to recruit 106 cases and 212 controls. Those who full fill inclusion criteria were recruited consecutively until the required sample size was achieved.
The study instruments were a pre-tested interviewer-administered questionnaire (IAQ) and a record sheet. The IAQ has three parts; part 1- assessed socio-demographic risk factors, part 2- assessed medical risk factors, and part 3- assessed other risk factors (family history of liver disease, Frequency of Tea, Coffee, and Fast food consumption, Physical activity level, Alcohol consumption, Methotrexate usage. The face and content validity of the questionnaire was assured. The validated global physical activity questionnaire (GPAQ) was used to access the physical activity levels.16 The validity of the GPAQ ranged from low to moderately high (r = 0.25–0.63).17 Pretesting was done at another hospital of a radiology unit with the patients who were attended to abdominal ultrasound scan. The record sheet was used to record the hemoglobin level, height, weight. The height was measured by using a standard measuring tape. The weight was measure by using a calibrated scale.
Data collection was done at the Radiology department of the hospital by the first author. Because all the admitted patients also send for scans to the radiology department and it was done by four trained medical officers. Therefore, with the help of the radiology medical officers, it could get relatively more accurate information whether the patients had or had not fatty Liver Disease. Before the data collection, obtained permission from the consultant radiologists in the Radiology department. An ultrasound scan was done using 5 MHz 50 mm convex probe which was made in china.
The safe limit of alcohol consumption was assessed based on Asian standard values which were 14 units per week per men and 7 units per week for women.18 One unit of alcohol was included in the 25–30 ml of arrack or whiskey, 50 ml of illicit alcohol, a half-pint of beer or toddy, 175 ml glass of wine. Those were taken into account for calculating units of alcohol per week.18 Hypertension, diabetic Mellitus, and dyslipidemia status were cross-checked with the BHTs.
Body Mass Index (BMI) for Asian cut off value was taken as a BMI of ≥ 23 Kg/m2 as overweight in this study.19 The physical activity level was assessed by using mainly three types of activities as the activity at work, travel to and from places, recreational activities. It was calculated by using Metabolic Equivalent (MET). In typical week, adequacy of exercises asses as by using Moderate intensity physical activity of 150 minutes or Vigorous-intensity activity of 75 minutes or combination of moderate and vigorous-intensity physical activity achieving at least 600 MET minutes. MET minutes were calculated in a case of combined activity as follows. Moderate intensity activity MET min/week equals 4.0*Moderate intensity activity minutes*Moderate activity days, Vigorous-intensity activity MET min/week equals 8.0*Vigorous intensity activity minutes* Vigorous-intensity activity days and walking/cycling MET min/week equals 3.3*walking/cycling minutes*Walking/Cycling days”.16 If the participant not fulfilled the above criteria was taken as inadequately physical activity level.
The cutoff of tea cups consumption was taken as > 4 cups per day and ≤ 4 cups per day and for the coffee, consumption was taken as ≥ 1 cups per day or none use per day.11,20 The fast foods were considered as cakes, snacks, bread, instant noodles, processed fish and meat, buns, pastries, hot dogs, etc. The cut off for fast food consumption was taken as > 1 time and the ≤ 1 time per week.21 The average hemoglobin value among the Asian women was 11 to 13 g/dl and among men, it was 13 to 15 g/dl. Therefore, the cut off value was taken as > 13 g/dl for females and > 15 for males. The family history of liver disease was considered as if any family member from a maternity or paternity side had suffered from any kind of a liver disease condition such as fatty liver, cirrhosis, NASH, etc.
The statistical analysis was done by using SPSS 16 version. The results were interpreted by using the odds ratios (OR), 95% confident intervals (95% CI), and the p values. The variables with p-value < 0.2 in the bivariate analysis were selected for the multiple logistic regression. The purposeful selection was used to perform. Hosmer and Lemeshow test was used as the test for goodness of fit and it was observed as satisfactory (chi-square 4.48, p value = 0.61)
The permission was taken from the Director of the hospital and the consultants. The interview was conducted after obtaining the informed written consent of the patients. Measures were taken to minimize the disturbances to the routine ward works. The data collection was done by avoiding the routine daily ward round times. Ethical clearance was taken from the Ethics Review Committee at the Postgraduate Institute of Medicine, Colombo.