Epidemiological aspects of metabolic dysfunction-associated steatotic liver disease

Metabolic dysfunction-associated steatotic liver disease (MASLD) is common in the global population and is quickly emerging as a global health challenge. If left untreated, persons with MASLD may progress to develop liver-related complications such as liver cirrhosis, which in turn may lead to major adverse liver outcomes (MALO), and hepatocellular carcinoma (HCC). The objective of this thesis was to investigate how we can improve the identification and management of persons with MASLD in register-based studies and clinical care, respectively. In Study I, we performed an external validation of the Toronto HCC risk index (THRI), which can be used to predict the development of HCC in persons with liver cirrhosis. The objective was to assess if the THRI can be used to guide recommendations on HCC surveillance. Here, we used a cohort of persons with liver cirrhosis (n=2,491) that had been in contact with Karolinska University Hospital between 2004-2017. Persons were stratified into three risk groups (low, intermediate, and high) based on their calculated THRI score. We found that the THRI had a modest discriminative ability (Harrell's C-index=0.69), and a good model calibration as indicated by the Calibration-in-the-large (0.11), calibration slope (1.24, not different from 1, p=0.66) and goodness-of-fit. Persons in the high-risk group had a 7.1-fold higher rate of HCC compared to persons in the low-risk group. However, the score was only able to identify a small subset of persons at low risk (5.3%). Together, these results indicate a limited utility of the THRI in guiding decisions on HCC surveillance in persons with liver cirrhosis in our population. In Study II, we assessed how well a register-based definition of MASLD captures a population with true MASLD. Here, 150 persons with an International Classification of Diseases 10th revision (ICD-10) code for MASLD at Karolinska University Hospital were randomly selected for inclusion. The MASLD diagnosis was validated upon medical chart review and persons were classified as true or false positive for MASLD. The validity of the ICD-10 code for MASLD was confirmed by calculation of the positive predictive value (PPV), which we found to be 0.82 using the ICD-10 code for MASLD alone. When excluding persons with ICD-10 codes for other liver diseases or alcohol abuse disorders, the PPV improved to 0.91. This supports that the ICD-10 code for MASLD is accurate, but that exclusion of other liver- or alcohol-related ICD-10 codes is preferred when conducting register-based research. In Study III, we identified all persons with newly diagnosed MASLD between 2006 and 2020 in the Swedish National Patient Register and matched these on age, sex, inclusion year, and municipality with up to 10 comparators from the general population. Our primary outcome of interest was severe depression or anxiety, which was defined using ICD-10 codes for depression or anxiety (ICD-10 F32-33 and F40-43), together with a medical operation (MVO) code, which indicates in which clinical setting or specialty healthcare visits have taken place. A secondary outcome of interest was progression to MALO, which was defined as a composite diagnosis including decompensated cirrhosis, HCC, liver transplantation and liver-related death. In analyses where MALO was the outcome, we restricted the study population to persons with MASLD and included severe depression or anxiety as a time-varying exposure. We found that persons with MASLD were at a higher risk of incident severe depression (adjusted hazard ratio [HR]=1.8, 95% confidence interval [CI]=1.5-2.1) compared with the matched comparators. We did not find an association between severe depression and MALO (aHR=1.0, 95%CI 0.6-1.5). However, this analysis was limited by few outcomes of MALO, and interpretation should therefore be made with caution. In Study IV, we included persons with or at high risk of MASLD (n=63,587) in the Stockholm CREAtinine Measurements (SCREAM) cohort. The SCREAM cohort is a healthcare utilization cohort that includes data on all persons seeking healthcare Stockholm region between 2006 and 2021. We defined persons with or at high risk of MASLD as either having an ICD-10 code or anatomical therapeutic chemical code corresponding to MASLD, obesity or type 2 diabetes. All persons with or at high risk of MASLD were then stratified into four groups based on estimated glomerular filtration rate (eGFR) (>90, 60-89, 30-59 and