Effect of Lifetime Alcohol Consumption on Severity of NAFLD
Effect of Lifetime Alcohol Consumption on Severity of NAFLD
The study population consisted of 77 patients with histologically confirmed NAFLD who also had had completed the lifetime alcohol consumption survey at the time of initial enrolment into the study. The characteristics of the 77 patients are in Table 1. Forty-three patients (56%) were women and 52 (68%) were obese (Body mass index ≥30 kg/m). The presence of metabolic syndrome as defined by ATP III (Adult Treatment Panel III, Third report of the expert panel on detection, evaluation and treatment of high blood cholesterol in adults) criteria was seen in 39% of patients (n = 30). Sixty-seven patients (87%) had evidence of insulin resistance defined as HOMA (homeostasis model assessment) index ≥2.0. Overall, histological disease was mild in these patients with a mean NASH activity score of 3.9 (range 1–7) and mean global NASH score of 5.4 (range 1–9). Severe liver disease (bridging fibrosis or cirrhosis on liver biopsy) was seen in 17 patients (22%).
The overall median alcohol consumption among the 77 patients with NAFLD was 24 gram-years (range 0–2251). Although the patients were recruited into the study based on the definition of <40 g per week, the patient with the alcohol consumption 2251 gram-year was the outlier – once recruited the patient gave this history on the lifetime alcohol questionnaire. Removing the patient from the analysis did not change the findings median values reported. Those who consumed equal to or more than 24 gram-years were more likely to be men (73% vs. 40%, P = 0.01), had significantly lower per cent total body fat (BIA% fat 33 ± 12 vs. 43 ± 12, P = 0.001) and lower serum total cholesterol levels (194 ± 42 vs. 215 ± 58 mg/dl, P = 0.05), when compared to patients who consumed <24 gram-years (Table 1). In addition, patients who consumed ≥24 gram-years had significantly lower fibrosis scores on liver histology (1.2 ± 1.0 vs. 1.8 ± 1.2, P = 0.03). Although there was a trend towards lower insulin resistance in the higher consumption group (HOMA index values 5.2 ± 3.6 vs. 6.1 ± 3.9), this difference was not statistically significant (P = 0.31).
We also compared patients with lifetime alcohol consumption at cut-off values of 10 gram-years and 40 gram-years. In both settings, the per cent total body fat (BIA%) and the fibrosis score remained significantly lower in the group that consumed more alcohol (data not shown).
Table 2 compares patients who consumed alcohol vs. those who had had abstained from alcohol in their lifetime (non-drinkers). Twenty-five patients (32%) were abstinent from alcohol and were more likely to be women (83% vs. 44%, P = 0.0001) and had a higher body fat percentage measured by bioimpedance assay (44.9% vs. 34.5%, P = 0.0003). There was, however, no statistically significant difference in the overall NAFLD histology assessment (steatosis, lobular inflammation and hepatocyte ballooning grades or fibrosis stage) between the two groups.
Severe liver disease (fibrosis stage 3–4: F3/F4) was associated with increasing age, per cent total body fat (BIA%), serum aspartate aminotransferase and alkaline phosphatase levels and NASH histological feature of ballooning (P < 0.05 – < 0.001). As seen in Table 1 and Table 2, the number of patients who were non-drinkers or had lifetime alcohol consumption below less than 24 gram-years was significantly higher in the F3/F4 group compared to the group with less severe (F0–2) liver disease with regards to fibrosis severity. The presence of severe liver disease was inversely related to the quantity of lifetime alcohol consumption (≥24 vs. < 24 gram-years, relative risk −0.32, P = 0.005). In multivariate analysis, as shown in Table 3, after controlling for other characteristics including age, gender, BMI, BIA% fat, HOMA-IR index, liver enzymes and selected histological features, the independent risk factors for severe liver disease were an increase in age (years) and smaller quantities (<24 gram-years) of lifetime alcohol consumption. Patients who consumed equal to or above the median level of alcohol during their lifetime had a significantly lower risk of bridging fibrosis or cirrhosis (OR: 0.26, 95% CI: 0.07–0.97, P = 0.046). In addition, the protective effect of alcohol consumption was even seen in patients consuming above the lower cut-off level of 10 gram-years (odds ratio 0.24, 95% CI: 0.07–0.89, P = 0.03) when controlling for other factors, however, it was no longer significant above the higher cut-off value of ≥40 gram-years of lifetime alcohol consumption (data not shown).
Forty-three patients had a history being of abstinent from alcohol for ≤1 year of enrolment into the study, whereas 34 patients had a history of being abstinent from alcohol for over a year prior to enrolment. Those who had been abstinent for more than a year had a significantly higher per cent total body fat (BIA% 41.2 ± 12.4 vs. 33.6 ± 12.4, P = 0.01) and also a higher fibrosis score on liver biopsy (1.7 ± 1.2 vs. 1.2 ± 1.0, P = 0.04) compared to those who were still drinking or were abstinent for ≤1 year. There was a tendency among the patients with a longer duration of abstinence from alcohol consumption to have a higher BMI, HOMA and serum triglyceride levels although the differences between the two groups were not statistically different (P > 0.05). We also compared patients with a longer duration of abstinence from alcohol consumption (≤5 years vs. >5 years). The per cent total body fat was significantly higher (BIA% 41.7 ± 12.4 vs. 34.4 ± 12.5, P = 0.01) and there was a trend towards a higher fibrosis score (1.8 ± 1.2 vs. 1.3 ± 1.0) although this was not statistically significant (P = 0.07) in those patients who were abstinent for ≤5 years compared to those who were abstinent for >5 years.
Results
The study population consisted of 77 patients with histologically confirmed NAFLD who also had had completed the lifetime alcohol consumption survey at the time of initial enrolment into the study. The characteristics of the 77 patients are in Table 1. Forty-three patients (56%) were women and 52 (68%) were obese (Body mass index ≥30 kg/m). The presence of metabolic syndrome as defined by ATP III (Adult Treatment Panel III, Third report of the expert panel on detection, evaluation and treatment of high blood cholesterol in adults) criteria was seen in 39% of patients (n = 30). Sixty-seven patients (87%) had evidence of insulin resistance defined as HOMA (homeostasis model assessment) index ≥2.0. Overall, histological disease was mild in these patients with a mean NASH activity score of 3.9 (range 1–7) and mean global NASH score of 5.4 (range 1–9). Severe liver disease (bridging fibrosis or cirrhosis on liver biopsy) was seen in 17 patients (22%).
Lifetime Alcohol Consumption
The overall median alcohol consumption among the 77 patients with NAFLD was 24 gram-years (range 0–2251). Although the patients were recruited into the study based on the definition of <40 g per week, the patient with the alcohol consumption 2251 gram-year was the outlier – once recruited the patient gave this history on the lifetime alcohol questionnaire. Removing the patient from the analysis did not change the findings median values reported. Those who consumed equal to or more than 24 gram-years were more likely to be men (73% vs. 40%, P = 0.01), had significantly lower per cent total body fat (BIA% fat 33 ± 12 vs. 43 ± 12, P = 0.001) and lower serum total cholesterol levels (194 ± 42 vs. 215 ± 58 mg/dl, P = 0.05), when compared to patients who consumed <24 gram-years (Table 1). In addition, patients who consumed ≥24 gram-years had significantly lower fibrosis scores on liver histology (1.2 ± 1.0 vs. 1.8 ± 1.2, P = 0.03). Although there was a trend towards lower insulin resistance in the higher consumption group (HOMA index values 5.2 ± 3.6 vs. 6.1 ± 3.9), this difference was not statistically significant (P = 0.31).
We also compared patients with lifetime alcohol consumption at cut-off values of 10 gram-years and 40 gram-years. In both settings, the per cent total body fat (BIA%) and the fibrosis score remained significantly lower in the group that consumed more alcohol (data not shown).
Table 2 compares patients who consumed alcohol vs. those who had had abstained from alcohol in their lifetime (non-drinkers). Twenty-five patients (32%) were abstinent from alcohol and were more likely to be women (83% vs. 44%, P = 0.0001) and had a higher body fat percentage measured by bioimpedance assay (44.9% vs. 34.5%, P = 0.0003). There was, however, no statistically significant difference in the overall NAFLD histology assessment (steatosis, lobular inflammation and hepatocyte ballooning grades or fibrosis stage) between the two groups.
Factors Associated With Severe Liver Disease (Fibrosis Stage 3-4)
Severe liver disease (fibrosis stage 3–4: F3/F4) was associated with increasing age, per cent total body fat (BIA%), serum aspartate aminotransferase and alkaline phosphatase levels and NASH histological feature of ballooning (P < 0.05 – < 0.001). As seen in Table 1 and Table 2, the number of patients who were non-drinkers or had lifetime alcohol consumption below less than 24 gram-years was significantly higher in the F3/F4 group compared to the group with less severe (F0–2) liver disease with regards to fibrosis severity. The presence of severe liver disease was inversely related to the quantity of lifetime alcohol consumption (≥24 vs. < 24 gram-years, relative risk −0.32, P = 0.005). In multivariate analysis, as shown in Table 3, after controlling for other characteristics including age, gender, BMI, BIA% fat, HOMA-IR index, liver enzymes and selected histological features, the independent risk factors for severe liver disease were an increase in age (years) and smaller quantities (<24 gram-years) of lifetime alcohol consumption. Patients who consumed equal to or above the median level of alcohol during their lifetime had a significantly lower risk of bridging fibrosis or cirrhosis (OR: 0.26, 95% CI: 0.07–0.97, P = 0.046). In addition, the protective effect of alcohol consumption was even seen in patients consuming above the lower cut-off level of 10 gram-years (odds ratio 0.24, 95% CI: 0.07–0.89, P = 0.03) when controlling for other factors, however, it was no longer significant above the higher cut-off value of ≥40 gram-years of lifetime alcohol consumption (data not shown).
Duration of Abstinence From Alcohol
Forty-three patients had a history being of abstinent from alcohol for ≤1 year of enrolment into the study, whereas 34 patients had a history of being abstinent from alcohol for over a year prior to enrolment. Those who had been abstinent for more than a year had a significantly higher per cent total body fat (BIA% 41.2 ± 12.4 vs. 33.6 ± 12.4, P = 0.01) and also a higher fibrosis score on liver biopsy (1.7 ± 1.2 vs. 1.2 ± 1.0, P = 0.04) compared to those who were still drinking or were abstinent for ≤1 year. There was a tendency among the patients with a longer duration of abstinence from alcohol consumption to have a higher BMI, HOMA and serum triglyceride levels although the differences between the two groups were not statistically different (P > 0.05). We also compared patients with a longer duration of abstinence from alcohol consumption (≤5 years vs. >5 years). The per cent total body fat was significantly higher (BIA% 41.7 ± 12.4 vs. 34.4 ± 12.5, P = 0.01) and there was a trend towards a higher fibrosis score (1.8 ± 1.2 vs. 1.3 ± 1.0) although this was not statistically significant (P = 0.07) in those patients who were abstinent for ≤5 years compared to those who were abstinent for >5 years.