O12 Liver frailty index predicts poor physical activity volume and intensity (GENEActiv accelerometer) in advanced chronic liver disease: UK prospective case-control study

<h3>Background</h3> Physical frailty leads to poor clinical outcomes in advanced chronic liver disease (AdvCLD). However, the volume and intensity of physical activity in AdvCLD remains unknown limiting the ability to tailor interventions for frailty. This study investigated (a) physical activity profiles in AdvCLD vs healthy controls (HC), (b) predictors of physical activity volume/intensity and their association with frailty, and (c) the physical activity volume and intensity achievable in AdvCLD. <h3>Method</h3> Outpatients with AdvCLD awaiting liver transplant were recruited, alongside HC (Birmingham, UK). Participants wore a GENEActiv® wrist-accelerometer for 14-days to measure physical activity volume and intensity, including most active continuous periods of the day. Measures of physical frailty included liver frailty index (LFI), quadriceps strength, and thigh muscle ultrasound. <h3>Results</h3> 42 AdvCLD (65% male, mean age 56, UKELD 52, 53% ArLD, 12% MASLD, 31% Autoimmune) and 17 HC (59% male, age 47) were recruited. Participants with ArLD had significantly lower volume of physical activity (mean[SD] acceleration 13.4[4.6] vs 29.2[8.9]mg, p&lt;0.01) and spent proportionately more time at low physical activity intensities than HC (Intensity gradient -2.9 vs -2.4, p&lt;0.01). Older age and refractory ascites independently predicted both lower volume (B=-0.24, p=0.02; B=-7.6, p&lt;0.01, respectively) and intensity (B=-0.001, p=0.02; B=-0.15, p=0.04, respectively). Lower LFI (less frail) and higher quadriceps strength were associated with higher physical activity volume (B=-4.68, p=0.03; B=0.11, p&lt;0.01, respectively) and intensity (B= -0.21, p&lt;0.01; B=0.004, p&lt;001, respectively); but not chair stands. Physical activity volume and intensity across quartiles of LFI revealed a non-significant downward trend in volume as frailty worsened (fig. 1A). Physical activity intensity differed significantly by LFI quartile (p=0.002, fig. 1B), with higher intensity in the 1st (most robust) and 2nd quartile than the 4th quartile (most frail) (-2.7[0.2] vs 3.1[0.1], p=0.004; -2.8[0.2] vs -3.1[0.1], p=0.02, respectively). The most active continuous one-minute of the day was at a vigorous intensity in the most robust LFI quartile (red line, fig. 1C), but only a slow walk for the frailest LFI quartile (purple line, fig. 1C). <h3>Conclusion</h3> Physical activity (especially intensity) is significantly reduced in patients with AdvCLD, being more pronounced in the frailest. Older age, refractory ascites and ArLD are associated with lower physical activity. The LFI is a valuable predictor of physical activity and could be used to prioritise and tailor exercise interventions. This study highlights realistic exercise targets in patients with AdvCLD (i.e. short bursts (&lt;5mins) of brisk walking) rather than those for the general population.