Association of weight changes with changes in histological features and blood markers in non-alcoholic steatohepatitis
Koutoukidis D., Jebb S., Tomlinson J., Cobbold J., Aveyard P.
<p><strong>Background and aims:</strong> Weight loss is recommended for patients with non-alcoholic steatohepatitis (NASH) but the impact of weight change on disease activity remains unclear. We examined the association between weight change (gain/loss) and changes in biochemical and histological features of NASH.</p> <p><strong>Methods:</strong> This was an analysis of the PIVENS and FLINT trials in adults with NASH who had liver biopsies at baseline and at either 1.5 years or 2 years. Multivariable regression models examined how weight change was associated with changes in (a) blood liver markers, (b) NASH resolution with no fibrosis worsening, (c) fibrosis improving with no NASH worsening, and (d) individual histological features.</p> <p><strong>Results:</strong> The BMI of the 421 participants was 34.4kg/m2 (SD:6.5) and their mean weight change was +0.4kg (SD:6.5). Weight change was independently and positively associated with changes in liver enzymes and the Fibrosis-4 score (all p<0.001). Each kg of weight loss was associated with 7% (95%CI: 3-10%, p<0.001) increase in odds of achieving NASH resolution with no fibrosis worsening and with 5% (95%CI: 1%-8%, p=0.01) increase in odds of achieving fibrosis improvement with no NASH worsening. Weight gain was associated with worsening of disease activity. For every kg of weight lost, the odds of fibrosis improving were 5% (95%CI: 2-8%, p=0.001). There was no evidence that the association between weight change and outcome depended upon pharmacological treatment, trial, body mass index, and baseline fibrosis.</p> <p><strong>Conclusions:</strong> Weight change was independently and monotonically associated with changes in biochemical and histological features of NASH. Guidelines for NASH management should incorporate recommendations for both avoidance of weight gain and support to lose weight.</p>