Charting the Pattern of Exercise in Stages of Anorexia Nervosa

Patients were significantly more active at weight restoration than at low weight.

Reprinted from Eating Disorders Review
March/April Volume 27, Number 2
©2016 IAEDP

One paradox in the course of anorexia nervosa (AN) is the increased levels of physical activity often reported among these very ill patients. A collaborative team of researchers from the New York State Psychiatric Institute, New York University, and Drexel University have measured physical exercise in AN patients at three time points from admission to follow-up. The researchers measured activity at low-weight, when weight was restored, and post-treatment, and compared these results with those from measured activity among age-matched heathy controls.

Loren M. Gianini, PhD, and her colleagues used a novel accelerometer to measure physical activity at these three time points in 24 women 15 to 49 years of age with DSM-5 AN and 24healthy controls (Int J Eat Disord. 2015. Dec 29. doi: 10.1002).The Intelligent Device for Energy Expenditure and Activity (IDEEA®; Minisun, Fresno, CA) is a microcomputer-based instrument that can identify at least 32 types of physical exercise, as well as the duration and intensity of walking or running. The device continually records for 24 hours, using 5 small sensors taped to the chest, thighs, and feet; the data are then transmitted to a small microcomputer on a waistband or belt.

The patients and controls wore the IDEEA at up to 3 time points: inpatient low-weight (Time 1), inpatient weight-restored (Time 2), and within 2 weeks of weight restoration (90% ideal body weight) on the impatient unit, and at follow-up (Time 3). Total activity time, including time standing and “fidgeting,” was recorded.

Sixty-one patients with AN were in the study; 45 patients wore the IDEEA for 3 consecutive days at low-weight, 35 at weight restoration, and 19 at one month after impatient discharge. Twelve wore the IDEEA at all three time points. Many healthy controls and patients with AN wore the IDEEA for 2 days but then removed it, resulting in a large amount of missing data for the third day of monitoring; thus, only data from the first two days of monitoring was used.

Physical activity across time

Contrary to the authors’ first hypothesis, patients were significantly more active at weight restoration than when they were at low weight. AN patients were also significantly more active at the one-month post-discharge point than at admission. After discharge (Time 3), patients were significantly more active than were healthy controls. The increase in physical activity, according to the authors, was primarily due to an increase in the amount of time spent on their feet, which was primarily composed of time spent standing (the device determined standing and fidgeting).

The authors also determined that elevated physical activity in AN is not directly related to the severity of the disorder or to general pathology. Another finding was that patients were more active during the day but less active at night than were controls. Some theories for this were that patients were more active during the day and then were fatigued in the evening and that patients were less socially active at all times, particularly in the evening. As for body mass index changes, the more time patients with AN spent on their feet (standing and walking), the more quickly they lost weight during the 12 months after discharge. 

This study is the first to identify a relationship between a particular type of objectively measured physical exercise and posttreatment weight. It also is the first to identify a relationship between a particular type of objectively measured physical exercise and posttreatment weight in AN patients. One suggestion from the authors is that inpatient units might develop interventions to limit standing and walking or to help patients stand and walk in a way that doesn’t lead to weight loss. Fidgeting did not differ between patients and controls and did not change with weight restoration; thus, it did not predict weight change after treatment.