Introduction Fatty liver disease (FLD) is common, with an estimated prevalence between 10 and 24% of the general population. Most patients are asymptomatic, and identified via incidental abnormal liver function tests, particularly alanine transaminase (ALT) and γ-glutamyltransferase (GGT). There are few population data on the prevalence or determinants of abnormal liver function.
Aim To determine the characteristics and associations of people with elevated values 1.5 times normal of both ALT and GGT in a population sample (upper limit of normal (ULN)=ALT ♂37 ♀31 IU/l; GGT ♂55 ♀38 IU/l) as an indicator of FLD.
Method The London Life Sciences Prospective Population Study (LOLIPOP study) is a population based cohort study of cardiovascular risk and outcomes in west London, an area with a high ethnic minority population. Adults aged 35–74 from 58 general practices were invited to cardiovascular screening which included questionnaire, clinical measurement and blood sampling. Response rate was 62%. Cross sectional data were obtained on 31 507, European white (n=9222), Indian Asian (n=19 769) and Black (n=2516). ALT and GGT were measured from a single serum sample. To identify key socio-demographic variables and potentially modifiable lifestyle factors, including BMI and alcohol intake, we used logistic regression models examining associations with elevated ALT and GGT.
Results The number with GGT and ALT measures was 31 465 (99.9% of total). The prevalence of elevated GGT above 1.5 times the ULN (GGT 1.5) was 10.4% (3265/31 465), for ALT 8% (2517/31 465) and for both 3.2% (996/31 465). 40% of those with ALT1.5 had GGT1.5, and 32% of those with GGT 1.5 had ALT 1.5. The independent odds of having both raised were increased in: younger people (OR 3.00 (95% CI 2.28 to 3.94) for 35–44 compared to 65–74 age group), south Asians (OR 1.29 (95% CI 1.10 to 1.51) compared to White), greater alcohol intake (OR 6.52 (95% CI 5.36 to 7.95) for >28 units per week compared to no alcohol), raised BMI (OR 2.34 (95% CI 1.94 to 2.83) obese compared to normal), and diabetes (OR 1.32 (95% CI 1.07 to 1.63), with a dose related gradient for alcohol consumption. Gender, deprivation, and taking statin were not significantly associated, and no interaction observed between alcohol and gender or BMI.
Conclusion The prevalence of abnormal liver function tests, suggesting FLD, was common (directly standardised prevalence of 3.3% in this age group in the England and Wales population). BMI, and alcohol independently increased risk across the range of their values, highlighting the importance of considering both in preventing severe liver disease and counselling patients.