Prevalence of non-alcoholic fatty liver disease (NAFLD) among adults in urban Goa

Once considered innocuous, non-alcoholic fatty liver disease (NAFLD) has emerged as the leading cause of chronic liver disease.¹ It is not attributable to excessive alcohol consumption, drugs, toxins, infectious diseases or other identifiable causes.² NAFLD is now the most common liver disease in the western world.³ It has a pooled global prevalence of 25.4%, with the highest prevalence in the Middle East and South America (around 30%) and the lowest in Africa (~13%).^4–6 It is closely associated with metabolic syndrome and its components—hypertension, dyslipidaemia, hyperglycaemia, obesity and insulin resistance.⁷ Hepatocellular carcinoma and cirrhosis are the end stages of progression of NAFLD.^5,8 Individuals with NAFLD are at increased risk of liver-related, cardiovascular and all-cause mortality.⁹ NAFLD has also been associated with type 2 diabetes mellitus, atherosclerosis, cardiovascular disease, chronic kidney disease, polycystic ovarian syndrome, obstructive sleep apnoea, etc.¹⁰

In India, the prevalence of NAFLD has been reported to be 9% to 53%.^11–13 In the state of Goa, while the focus has been on the health effects of high level of alcohol consumption, no attention has been given to NAFLD. Hence, we decided to estimate the prevalence of NAFLD as well as associated factors in adults aged >30 years in an urban community in Goa.

We did a community-based cross-sectional study in an urban area in Goa, India. Individuals >30 years of age, with no alcohol use and living in the area for more than 6 months were recruited in the study. Pregnant and lactating women and individuals with any amount of alcohol intake, individuals on hepatic steatogenic medication in past one year were excluded from the study. Those testing positive for hepatitis B or C among those diagnosed with fatty liver on ultrasound were also excluded.

Considering the NAFLD prevalence to be 16.6¹⁴ in a study in an urban area in western India and absolute precision as 5%, we calculated the sample size to be 206. Taking into consideration non-responses and refusals to undergo ultrasound, we decided to recruit 236 participants in the study. The systematic random sampling technique was used to recruit the study subjects.

Data were collected using a structured questionnaire. For clinical examination and anthropometry, digital sphygmomanometer, nonstretchable measuring tape and portable weighing scale were used. For investigations, ultrasonography (Siemens Acuson S2000 with 6C1 HD and 4V1 transducer) was used by a trained radiologist for the diagnosis of NAFLD. The Hepatic Steatosis Ultrasound Images Assessment manual¹⁵ was used to assess the steatosis in the participants. NAFLD was graded into mild, moderate and severe NAFLD.¹⁵ Participants found to have fatty liver by ultrasound were tested for hepatitis B surface antigen and hepatitis C core antigen. Hepatitis C was tested with a rapid test kit (SD Bioline HCV, Abbot Laboratories) with manufacturer reported sensitivity and specificity of 99.3% and 100%, respectively. Calorie intake was measured by 24-hour dietary recall. Physical activity was measured by the Global Physical Activity questionnaire. For waist:hip ratio (WHR) we used cut-off of >0.9 for men and 0.85 for women. For biochemical parameters, we used the following cut-offs: 126 mg/dl for fasting blood sugar level, 200 mg/dl for serum cholesterol, 150 mg/dl for serum triglyceride, 1.2 mg/dl for serum bilirubin, 40 i.u./L for aspartate aminotransferase (AST) and 56 i.u./L for alanine aminotransferase (ALT). We categorized patients based on body mass index (BMI) into three categories as per Indian cut-offs: lean (18.5–22.9 kg/m²), overweight (23.0–24.9 kg/m²) and obese (>25 kg/m²).

Statistical analysis was done using IBM SPSS Statistics for Windows (Version 24.0. Armonk, NY: IBM Corp). NAFLD prevalence with 95% CI was calculated. Prevalence odds ratio with 95% CI was calculated for various associated factors. Chi-square was the test of significance used and a p value <0.05 was considered statistically significant. Ethics approval for the study was obtained from the Institutional Ethics Committee and written informed consent was obtained from all the participants recruited in the study.

Of the 278 adults approached for the study, 42 participants reported either alcohol intake or intake of hepatic steatogenic drugs and were excluded. Thus, a total of 236 participants were recruited in the study and interviewed. Fifteen participants did not report for laboratory investigations and 11 did not report for ultrasonography. Eventually complete data were available for 210 study participants and were included in the final analysis. Women comprised 62.9% of the study participants. The majority of participants were Hindus (56.7%) followed by Christians (31.4%) and Muslims (11.9%). Most participants belonged to socioeconomic class II (43.8%) by BG Prasad classification, closely followed by socioeconomic class III (37.6%). Around 11.9% belonged to class IV, 2.4% to socioeconomic class V and 4.3% to socioeconomic class I.

The overall prevalence of NAFLD was 34.8% (95% CI 28.5%– 41.4%). The prevalence among men (46.2%; 95% CI 35.3%–57.3%) was significantly higher (p=0.008) than among women (28%; 95% CI 20.8%–36.1%). As far as grading of NAFLD was concerned, 57.5% were mild, 38.4% moderate and 4.1% were severe.

Individuals >35 years of age had higher prevalence of NAFLD (39.4%) compared to those <35 years (6.7%). The observed difference in prevalence between the different age groups was statistically significant (p=0.006). The highest prevalence was seen among Hindus (38.7%) followed by Christians (36.4%) and the lowest among Muslims (12%). This difference was found to be statistically significant (p=0.037). Higher prevalence of NAFLD was observed in individuals belonging to higher socioeconomic classes I (44.4%) and II (48.9%). Prevalence was low in the lower socioeconomic classes IV (4.0%) and V (20%).

Individuals consuming a non-vegetarian diet were twice as likely to have NAFLD compared to vegetarians (OR 2.83; 95% CI 1.28– 6.24) and this difference in NAFLD prevalence between vegetarians and non-vegetarians was statistically significant (p=0.008). The prevalence of NAFLD was significantly higher (p=0.0001) among people with diabetes (52.1%) compared to those without diabetes (25%). We found a similar increased risk for NAFLD among hypertensive individuals (Table I).

Using high fasting blood sugar (>126 mg/dl) as a marker for poor glycaemic control, we found a higher prevalence of NAFLD (59.4%) among those with high fasting blood sugar compared to those with normal blood sugar (40.6%). We also found significant association of high cholesterol level (>200 mg/dl) and high triglyceride levels (>150 mg/dl) with NAFLD with individuals with high cholesterol 3.34 times (OR 3.34; 95% CI 1.84–6.09) as likely and individuals with high triglyceride 2.22 times (OR 2.22; 95% CI 1.17–4.24) as likely to have NAFLD compared to those with normal levels of these two biochemical markers (Table I).

We found no association between physical activity as measured by the Global Physical Activity Questionnaire and NAFLD (p=0.35) but BMI was significantly associated with NAFLD (p=0.001). WHR was also significantly associated with NAFLD (p=0.003). The mean values of BMI, fasting blood sugar level, total cholesterol, serum triglyceride, serum bilirubin, AST and ALT were significantly higher in subjects with NAFLD compared to those without NAFLD.

Binary logistic regression analysis by the forward stepwise Wald method identified only WHR, fasting blood sugar level and total cholesterol as significantly associated with NAFLD. The model correctly classified 75.2% of the cases. The logistic regression model was statistically significant (p<0.001) and Nagelkerke R² was 0.353 indicating the model explained 35.3% of the variability (Table II).

The prevalence of NAFLD was found to be 34.8% (95% CI 28.5%– 41.4%). A meta-analysis of 237 studies from Asia¹⁶ for the period 1999 to 2019 reported overall prevalence regardless of diagnostic method as 29.62% (95% CI 28.13–31.15). The pooled prevalence for India¹⁶ was reported to be 32.74% (95% CI 13.89–59.49). Mohan et al.¹⁷ reported an NAFLD prevalence of 32% in their study in Chennai, India. While Amarapurkar et al.¹⁴ reported a prevalence of 16.6% in an urban population in Mumbai, India.

The prevalence of NAFLD was significantly higher among men than women (46.2% v. 28.0%). Mohan et al.¹⁷ reported a higher prevalence among men (35.1%) compared to women (29.1%). Amarapurkar et al.¹⁴ also reported a higher prevalence among men compared to women (24.6% v. 13.6%). Li et al.¹⁶ in their meta-analysis of Asian studies also reported a higher prevalence among men compared to women (37.11% [95% CI 35.04–39.24] v. 22.67 [95% CI 20.61–24.88]). However, the relationship between gender and NAFLD is believed to be diametrically divided between either genders in several studies, some studies showing a preponderance of women while others show a preponderance of men in prevalence.¹⁸

Individuals >35 years of age had a significantly higher prevalence of NAFLD (39.4%) compared to those <35 years old (6.7%) and the prevalence increased with increasing age. NAFLD and ageing are believed to be strongly correlated and increasing age is considered robust epidemiological factor for NAFLD, non-alcoholic steato-hepatitis and fibrosis.¹⁹ The common age of presentation of NAFLD in India has been reported to be between 30 and 50 years.^14,20,21 A higher prevalence of NAFLD was observed in individuals belonging to higher and middle socioeconomic classes. Singh et al.²² in a hospital-based study reported that a majority of patients with NAFLD belonged to the middle-income group.

The prevalence of NAFLD was 52.1% among people with diabetes compared to 25% among those who did not have diabetes. Insulin resistance of varying degree is suggested to be the cellular level abnormality underlying NAFLD.¹¹ A strong association between type 2 diabetes mellitus and NAFLD has been documented.^17,23–26 A meta-analysis of 24 studies has reported the pooled prevalence of NAFLD to be 59.67% (95% CI 54.31–64.92) among people with diabetes.⁷

We found a significant increased risk for NAFLD among people who had hypertension and a similar association between hypertension and NAFLD was reported by Majumdar et al.²⁷ in their population-based study in Haryana (OR 2.7; 95% CI 1.4–5.3; p=0.003). We found a higher prevalence of NAFLD in individuals with fasting blood sugar >126 mg/dl (prevalence 59.4% v. 22.7%), total cholesterol >200 mg/dl (prevalence 52% v. 25%) and serum triglyceride >150 mg/dl (prevalence 49% v. 30.2%). Speliotes et al.²⁸ in their study observed that high fasting blood sugar was associated with NAFLD (OR 2.95; 95% CI 2.32–3.75; p=0.001). Significant association between high triglyceride level and NAFLD (OR 3.7; 95% CI 1.2–13.3) has been reported by Leite et al.²⁹ A similar association between high fasting blood sugar (OR 4.0; 95% CI 0.9–17.6, p=0.03), high cholesterol (OR 2.5; 95% CI 1.1–5.6; p=0.03) and high triglyceride level (OR 2.3; 95% CI 0.99–5.2; p=0.05) with NAFLD has been reported by Majumdar et al.²⁷

BMI and WHR were found to be significantly associated with NAFLD. Several studies^17,25,26 have also reported significant association between higher BMI and NAFLD. There is believed to be a close pathogenic connection between obesity and NAFLD as patients with NAFLD are often obese and obese people have a higher prevalence of NAFLD.¹¹

Limitations: Our study participants were aged >30 years, so there is a possibility of overestimation of the prevalence. We did not exclude other aetiologies such as Wilson disease or autoimmune hepatitis. Due to issues with high-density lipoprotein (HDL) testing at the institutional laboratory, we were unable to study the association between decreased HDL level and NAFLD as well as metabolic syndrome and NAFLD.

Considering the high prevalence of NAFLD in urban Goa, there is a need to focus on this important but seemingly invisible chronic disease before it reaches epidemic proportions given the changing patterns in lifestyle, diet and increasing burden of other noncommunicable diseases such as diabetes, obesity, hypertension and dyslipidaemia.