Weak evidence from recent review suggests there is no harm when using exercise to treat anorexia nervosa

Exercise is usually promoted as a healthy behaviour, with government guidelines on how much we should all do per week. At the right levels many people find exercise an enjoyable way to improve their health, but is this always the case?

For a particular group of people who have problems with eating and weight, this isn’t necessarily true. Individuals with anorexia nervosa sometimes find that excessive exercise can become a way to maintain a dangerously low weight, or to provide feelings of control. However, when not nutritionally supported extreme exercise can cause many health complications. Exercise is often viewed as a negative symptom of anorexia, so it’s therefore logical that treatment should aim to reduce or completely stop this behaviour, right?

Exercise is often viewed as a negative symptom of anorexia, so could physical activity ever be involved in the treatment of the condition?

Indeed this is what many therapists would advise. However not all would agree, as some may argue that restricting exercise during treatment may make matters worse. This is similar to when we thought that exercise after heart surgery would do more damage than good, when actually exercise aids recovery.

Restricting exercise could contribute towards a loss of control often felt during therapy; promote secretive and unsupervised exercise; and cause resistance and resentment towards the therapist and/or treatment. It also seems unreasonable to expect a patient to successfully engage in healthy and safe physical activity after treatment if the tools and methods for doing so are not fully explained during therapy.

So, could exercise ever be involved in the treatment of anorexia? Researchers from Canada have attempted to answer this question by conducting a systematic review on the impact of exercise training interventions for patients with anorexia, extending and updating a previous review by Zunker et al., (2011).

Methods

Moola, Gairdner, and Amara conducted a search using multiple databases to identify any type of research evidence that looked at the impact of exercise/physical activity on physical or psychosocial health outcomes in patients with anorexia.

The authors conducted a systematic search of Medline, EMBASE, CINAHL, PsychInfo, Scopus, Cochrane, and the Physical Education Index. Two independent authors screened the articles for inclusion.

No restrictions were placed on the age, gender, illness severity, or type of exercise. However, studies that included individuals with bulimia nervosa or eating disorder not otherwise specified were excluded. Effect sizes were not calculated for any of the interventions.

Results

Out of the initial 1,664 articles found by the systematic review search, the authors included 10 papers. Out of these: 5 were Randomised Controlled Trials (RCTs); 2 quasi experimental studies; 2 case studies; and 1 qualitative study.

The authors report the results for the following outcome measures:

Change in body composition or Body Mass Index (BMI)
- No studies reported a negative impact of exercise on weight, with 3 studies reporting an improvement in weight, BMI, or body composition.
Fitness
- One study reported enhanced fitness as a result of aerobic (sustained, moderate activity) training, whilst another study found no difference on fitness as a result of anaerobic (short, high intensity) training.
Strength
- Three studies reported an increase in strength in response to resistance (weight) training.
Eating disorder symptoms
- Only 2 out of the 10 studies reported results on eating disorder symptomology. In both cases there was no increase in eating disorder pathology. Both studies noted a decrease in food preoccupation, and food related emotional distress.
Psychological wellbeing and quality of life
- Six studies reported positive effects of the exercise interventions on measures of quality of life and psychological wellbeing.
Tolerability and acceptability of exercise
- Five of the 10 studies suggested that patients with anorexia tolerated the exercise and enjoyed the interventions.
Compliance
- Although most studies did not assess program compliance, 2 studies reported improved patient compliance as a result of integrating exercise and treatment.

Conclusions

This review provides some interesting findings, but ultimately we need better quality evidence before we can conclusively say that exercise is safe for people with anorexia

Eating disorders researchers need to aim higher before we can reliably answer this question

The authors concluded:

While studies are few and small sample sizes likely alter the significance of the findings, the research suggests that properly designed and safe adapted physical activity can “do no harm” whilst also offering physical and psychosocial benefits.

Although this study provides some preliminary evidence on the impact of incorporating exercise into the treatment of anorexia, a higher quality of research evidence is needed before we are able to draw firm conclusions.

Limitations

The authors noted that the existing research on this topic is poor and insufficient. The research studies included were varied (mainly quasi experimental or RCTs), insufficiently powered, and lacked information on blinding and randomization procedures.

Additionally, the review included a broad range of unspecific outcome measures that were not always comparable across studies. Only 6 out of the 10 studies included a comparison group and it is not clear whether the studies that did include a control group compared exercise in combination with treatment vs. treatment alone, or exercise vs. nothing. Other variables such as type of simultaneous psychological treatment, patient characteristics, and medication were not described or controlled for, and may have influenced results.

Furthermore, there was no assessment of publication bias, which could be quite high in this area. It is questionable how many researchers would be comfortable publishing research showing that an exercise intervention they proposed negatively impacted a vulnerable patient group.

Despite the limitations the review highlights an important field of research, and provides a useful starting point for future studies. Future research should aim to conduct high quality RCTs, with clear, standardised definitions on the type, duration, intensity, timing, and frequency of exercise. A matched comparison group and main outcome measure should be selected, and follow ups completed to measure relapse.