Results
The groups were well matched for age and sex ( Table 1 ). Weight and BMI were significantly higher in the NAFLD group when compared with controls ( Table 1 ). Liver fat content was 13±7%, ALT levels 55±33 UL, HbA1c 6.0±0.8% and fasting glucose 5.4±1.6 mmol/L in the NAFLD group. The control group did not self-report any disease.
As summarised in Table 2 , the average number of steps taken each day was significantly fewer in NAFLD compared with controls (8483±2926 vs 10377±3529 steps/day; p<0.01; figure 1A) as was total daily energy expenditure (2690±440 vs 2901±511 kcal/day; p<0.01; figure 1B). Average daily MET levels were significantly lower in the NAFLD group when compared with controls (1.2±0.2 vs 1.4±0.2 METs; p<0.01; figure 1D), as was AEE (classed as activity of >3.0 METs: 432±258 vs 732±345 kcal; p<0.01). People with NAFLD spent less time performing physical activity of any intensity (73±44 vs 124±49 min/day; p<0.01; figure 1C) than the controls, and a significant difference was also observed between the groups when the physical activity was divided up into intensity levels ( Table 1 ). Sedentary time, classed as activities up to 3.0 METs, was not statistically significantly different between the groups, but was higher in the NAFLD group (1318±68 vs 1289±60 min/day; p=0.047; figure 2A).
(Enlarge Image)
Figure 1.
Objectively measured physical activity levels were lower in non-alcoholic fatty liver disease (NAFLD) compared with healthy controls (data reported as daily means (SD)). (A) Steps. (B) Total energy expenditure. (C) Physical activity duration. (D) Average MET levels.
(Enlarge Image)
Figure 2.
Sedentary time was higher in non-alcoholic fatty liver disease (NAFLD) than healthy controls with fewer sedentary to active transitions (data reported as daily means (SD)). (A) Sedentary time. (B) Sedentary to active transitions.
Sedentary Activity
Distribution analyses of the lengths of sedentary bouts demonstrate that patients with NAFLD have the same duration of sedentary bouts (Lorentz area under curve (AUC) 0.19±0.03 vs 0.18±0.02; p=0.106) as their healthy counterparts. The number of transitions from being sedentary to active were lower in patients with NAFLD compared with controls, but just failed to achieve statistical significance (13±0.03 vs 15±0.03%; p=0.021; figure 2B).
Using the self-reported IPAQ, people with NAFLD reported lower levels of physical activity and more time spent sitting than their healthy counterparts (see Table 2). There was little correlation between the daily TEE recorded by the multisensor array and self-reported physical activity levels in the IPAQ across the whole group (r=−0.192; p=0.216). Sedentary time measured by the multisensor array was not associated with sitting time reported in the IPAQ (r=0.278; p=0.071).
Higher BMI was associated with lower average METs (r=−0.496; p<0.01), shorter physical activity duration (r=−0.494; p<0.01) and less moderate (r=−0.457; p<0.01) and vigorous activity undertaken (r=−0.445; p<0.01) in NAFLD. A trend towards a positive correlation between BMI and sedentary time was observed, however, this did not reach statistical significance (r=0.306; p=0.065). There was no correlation between liver fat, fasting glucose, HbA1c and ALT with any of the physical activity parameters measured by the multisensor array within the NAFLD group. Multivariate analyses were undertaken to control for BMI and age with respect to NAFLD. This showed that these factors contribute to lower activity levels and higher sedentary behaviour.