*3.1. Characterization of the Study Population Including Medical Details, Comorbidities, Medications, and Body Composition*

Table 1 shows the patients' demographic characteristics and body composition data upon hospital admission compared to the healthy control group. The two study groups did not significantly differ in age (*p* = 0.057). Body weight, BMI, body fat, and lean mass were significantly lower in patients with AN compared to controls (all: *p* < 0.001; Table 1). Regarding phase angle, i.e., the ratio of body cell mass to fat-free mass as an indicator of cellular health and integrity, AN patients had significantly lower values than controls (*p* < 0.001; Table 1).

Medical details, comorbidities, and current medications of the study populations are summarized in a supplemental table (Supplementary Table S1). Forty-eight percent of the patients were diagnosed with restrictive AN, 26% with purging AN, and 26% with atypical AN. In terms of comorbidities, AN patients had significantly more pericardial effusion (*p* < 0.001), episodes of depression (*p* < 0.001), and at least one comorbidity (AN = 96% vs. C = 33%, *p* < 0.001). No statistically significant differences existed for other medical disorders. In terms of medication, a significant difference between both groups existed only for psychopharmacological treatment, with none of the controls (C) but 16% of the AN patients receiving medication on admission (*p* = 0.021). No significant difference existed for oral contraceptives (*p* = 0.052), L-thyroxine (*p* = 0.632), or taking no medication (AN, 35% vs. C, 37%; *p* = 1.000).


**Table 1.** Demographic characteristics and bioimpedance data in patients with anorexia nervosa on admission and in the healthy control group.

Data are expressed as mean ± SD (range) or as median (25th/75th percentile). BMI, Body Mass Index; N/A, not applicable; AN, anorexia nervosa.

#### *3.2. Comparison of Physical Activity and MET Intensities*

PA data and time spent in different levels of physical activity of 50 hospitalized AN patients compared to 30 ambulatory healthy controls are outlined in Table 2. Both groups engaged in similar levels of activity in terms of average steps and total distance per day. However, patients with AN had a greater range regarding the step count; 2479–31876 vs. 6507–22948 steps (Table 2). Significant differences were observed in daily average METs with patients presenting lower median values than controls. Patients with AN spent significantly more time in very low (*p* = 0.004) and low (*p* < 0.001) levels of PA than controls. Conversely, AN patients spent significantly less time in PA below the very low PA level (*p* < 0.001), in moderate (*p* = 0.022) as well as in 6–9 MET vigorous activity level (*p* < 0.001; Table 2). However, no significant differences were found for markers of physical inactivity: Both groups spent nearly the same duration of time on recumbency and sleep.

**Table 2.** Physical activity and the division into different MET cut-offs in patients with anorexia nervosa on admission and in the healthy control group.


Data are expressed as mean ± SD (range) or as median (25th/75th percentile). LPA, light-intensity physical activity; MET, metabolic equivalent; MPA, moderate-intensity physical activity; PA, physical activity; VLPA, very light-intensity physical activity; VPA, vigorous-intensity physical activity.
