Research Article

Physical Activity Intervention for Patients with Anorexia Nervosa: A Systematic Review and Meta-Analysis

Qingling Zhang^1*, Rufu Xu², Lingling Chu³, Xi She³, Dan Feng³ and Lei Dong³

¹Department of Psychosomatic, the Second Affiliated Hospital of Army Medical University, China

²Medicine and Clinical Epidemiology Center, the Second Affiliated Hospital of Army Medical University, China

³Department of Nursing, the Second Affiliated Hospital of Army Medical University, China

Submission: December 09, 2019; Published: January 06, 2020

*Corresponding author: Qingling Zhang, Psychosomatic Department , the Second Affiliated Hospital of Army Medical University, Chongqing, 400037, China

How to cite this article: Qingling Zhang, Rufu Xu, Lingling Chu, Xi She, Dan Feng, Lei Dong. Physical Activity Intervention for Patients with Anorexia Nervosa: A Systematic Review and Meta-Analysis. Psychol Behav Sci Int J. 2020; 14(2): 555881. DOI: 10.19080/PBSIJ.2019.10.555881.

Abstract

Purpose: To explore the intervention effects of physical activity in patients with anorexia nervosa (AN).

Methods: Six databases were searched and randomized controlled trials that isolated the intervention effects of physical activity in AN patients were included. Pooled event rates were calculated using random effects models.

Results: Of 7301 studies identified, 7 studies of randomized controlled trials were eligible. Pooled results showed no significant differences in the influence of physical activity on some anthropometric variables, such as weight (mean difference (MD)= -1.58, 95% CI: -4.26, 1.10, Z=1.16, P =0.25, I2=0%) and weight increment (MD= 0.04, 95% CI: -1.50, 1.57, Z=0.05, P =0.96, I2=51%), body mass index (BMI) (MD= -0.4, 95% CI: -1.23, 0.42, Z=0.96, P =0.34, I2=0%)and BMI increment (MD= 0.12, 95% CI: -0.49, 0.73, Z=0.38, P =0.70, I2=0%), and percentage body fat (MD= -0.62, 95% CI: -2.6, 1.37, Z=0.61, P =0.54, I2=0%), whereas compared to the control group, the intervention group had significantly improved muscular strength and agility and time to vital-sign stabilization significantly shortened in some studies.

Conclusion: Meta-analysis provides low-moderate quality evidence that a certain amount of physical activity with adequate food supply in the treatment of patients with AN does not lead to a decrease in weight, BMI or body fat percentage. Furthermore, a certain amount of physical activity benefits the increase of bone mineral density, muscular strength and agility. Moreover, a certain amount of physical activity can improve compliance with the treatment regimen and shorten the time to vital-sign stabilization.

Keywords: Anorexia; Exercise; Physical activity; Weight; Body mass index

Introduction

Anorexia nervosa (AN) is the most dangerous eating disorder [1], which mainly affects adolescents and young adults [2]. Mortality rates of 5%-10% at 10 years have made AN the psychiatric disorder with the highest mortality [3]. Patients with AN may have abnormal vital signs, such as low heart rate, low blood pressure, abnormal lab values and loss of bone density [4,5]. AN is also associated with significant psychiatric comorbid conditions, including obsessive-compulsive disorders, anxiety, and depression, as well as considerable impairment in psychosocial functioning [6,7].

AN includes two clinical subtypes. One is the restricting type, with the primary loss of weight through excessively restricting caloric intake, and the other is the binge eating/purging type, with recurrent binge eating and purging through self-induced vomiting or laxatives, enemas or diuretics misuse. AN has three main elements. The patient with AN inordinately restricts food intake to the point of markedly reduced body weight yet remains excessively concerned about obesity or weight gain and has the distorted body image of being overweight [1]. Severe malnutrition and weight loss are considered core components of AN [8].

AN is often preceded by excessive physical activity [9]. Because some people think physical activity would lead to weight loss, exercise is not often prescribed in the clinical circumstance of individuals with AN [10]. In fact, some patients with AN are placed on very strict bed rest. However, it has been suggested that moderate physical activity during refeeding of patients with AN should be safe and beneficial for the restoration of body composition, the preservation of bone mineral density, and the management of mood [10,11]. Therefore, how physical activity should be managed during the clinical care of patients with AN is still controversial.

Methods

This review is reported according to PRISMA guidelines. PRISMA is short for Preferred Reporting Items for Systematic Reviews and Meta-Analyses [12]. The protocol for this systematic review was not registered.

Search strategy

Six databases, including EMBASE, PubMed, Medline, CINAHL, BIOSIS Previews and Cochrane Library, were searched from their inception to 19 November 2018. The free words ‘anorexia or anorexic’ and ‘physical activity or physical exertion or physical fitness or exercise or sports’ were used to identify relevant studies. Secondary searches consisted of manually searching the reference lists of relevant systematic searches to obtain potentially eligible additional primary studies.

Study selection

Reports of randomized controlled trials (RCTs) that met the following inclusion criteria were included. First, study subjects were adolescent or adult patients diagnosed with anorexia nervosa. Second, patients were treated with usual care in the control group while patients in the intervention group were treated with usual care and a certain amount of physical activity, which included aerobic exercises and strength training. Third, the primary outcomes involved body weight, body weight increment, BMI and BMI increment, whereas the secondary outcomes included percentage body fat, lean body mass, muscle strength or endurance, and quality of life.

Data extraction

Two authors independently extracted the data and assessed trial quality. When disagreements appeared, the two authors resolved the difference through discussion. If no consensus was achieved, a third person was consulted. Descriptive data extracted from each study included study location, characteristics of study participants (type, sample size, age, sex), interventions (amount and duration of exercise, type, mode of delivery, other interventional components), outcomes (body weight, BMI, fat mass and lean body mass, quality of life, bone mineral density, muscular strength and agility, time to vital-sign stabilization). To pool some variables, we also obtained data through calculation according to given information. For example, we obtained the standard deviation (SD) of the body weight increment, BMI increment, skeletal muscle mass (SMM) increment and percentage body fat increment in this way in the Del Valle 2016 study.

Risk of bias and evidence grade assessment

The possible bias of the included studies was assessed independently by two authors using the Cochrane Risk of Bias Tool [13]. This tool has 7 items, including sequence generation, allocation concealment, blinding of participants and personnel to the study protocol, blinding of outcome assessment, incomplete outcome data, selective outcome reporting and others. According to these criteria, each study was considered to be at high, unclear, or low risk of bias. We evaluated the quality of evidence for each outcome according to the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach [14]. The quality of evidence was downgraded from high quality by one level for each of the following criteria: risk of bias, inconsistency, indirectness, imprecision and other considerations. The quality of evidence for each outcome was classified as very low, low, moderate, or high quality of evidence. GRADE profiler (version 3.6, GRADEpro) was used to construct the summary tables for these outcomes.

Data synthesis

The descriptive characteristics and outcomes of the included studies were summarized and synthesized . We pooled study results based on explicit inclusion and weighting criteria. We assessed the intervention effects of physical activity on anthropometry in patients with AN. The mean changes from baseline in body weight, BMI, and percentage body fat were treated as continuous variables; thus, they were expressed as the mean difference (MD) with 95% confidence intervals (95% CIs). All meta-analyses were performed using Review Manager Version 5.3. Random- or fixed-effect approaches were used to determine effect size depending on the heterogeneity or homogeneity of the studies. We used the Cochrane chi-square test and I2 statistics to detect the heterogeneity among the included studies [15]. A P-value<0.10 or I2≥50% indicated significant heterogeneity, and a random-effects model was used to pool the effect size. Otherwise, a fixed-effects model was used. Because the number of included studies was < 10, publication bias was not assessed.

Results

Study selection

The initial electronic database search yielded a total of 7297 articles. Through additional manual searches of reference literature lists, 4 articles were found. After further identification according to the type of articles, screening of titles, abstracts, or the removal of duplicates and reading of full texts,7 randomized controlled trials (209 participants, including 5 male patients) were included for meta-analysis and systematic review. The search flowchart is shown in Figure 1.

Study characteristics

The descriptive characteristics of the seven included studies are as follows (see Table 1 and Table 2). Studies were published between 2000 and 2017. Three were conducted in Spain [16-18], one in Canada [19], one in South Africa [20], one in Denmark [21] and one in the USA [22]. Two studies recruited inpatients (n=55), two recruited daily intrahospital care patients (n=72), two recruited outpatients (n=34), and one recruited both inpatients and outpatients (n=24), which was a double-blind, randomized, controlled, crossover study. The intervention duration ranged from 2 sessions daily over 9 days to 2-3 sessions every week over 8-12 weeks. The anthropometric outcomes after the physical activity intervention included body weight, BMI, percentage body fat and lean body weight. Some studies reported the mean value of anthropometric indicators pre- and postintervention [16,17,20]. Some studies reported the increment mean value of anthropometric indicators after intervention [18,19,21,22]. Three studies reported muscular strength, and one study reported agility. One study reported time to vital-sign stabilization, and one study reported the Short-Form 36-item Quality of Life Questionnaire (SF36) score.

From Figure 2, we can see that risk of bias across the 7 included studies was scored on 49 items. The two reviewers were consistent on 46 items. The 93.88% consistency indicated very good agreement. After discussion with the third author, consensus was reached on the disagreements. One study had all domains judged as low risk of bias, three studies had one or two domains as unclear risk, whereas the others each had one item in unclear risk of bias and one item in high risk of bias. The main weakness of the included studies was that allocation concealment was not mentioned in 6 studies. The GRADE evidence profiles for the outcomes are shown in Table 3. The GRADE level of evidence was moderate for body weight, BMI and percentage body fat post intervention and low for body weight increment, BMI increment and percentage body fat increment after intervention. The main reasons for the outcomes with moderate and low evidence were that the sample size was too small, and some studies were at high risk of bias.

Effects of interventions

Only body weight, BMI, percentage body fat and increments in these measures were suitable for meta-analysis. There were no significant differences in the above-mentioned anthropometric variables between the intervention and control groups (Figures 3-7). In addition, the subjects in the intervention group significantly increased the peak torque of their knee extensors and flexors and elbow flexors compared to those in the control group [20]. Some training programs also showed beneficial effects on agility. The time for restoring vital-sign stabilization was significantly reduced in the intervention group compared with the control group [22].

Discussion

This systematic review suggested that the existing RCTs on physical activity interventions in patients with anorexia nervosa were very limited. Due to a limited number of studies, we could only conduct a small- sample- size meta-analysis. However, our review was able to indicate that a certain amount of physical activity was not detrimental to anorexic patients on important anthropometry indicators. Moreover, appropriate exercise could bring some benefits for anorexic patients. Compared to previous systematic reviews [23,24], which included not only RCTs but also quasi-randomized studies and single-group studies or which were aimed to patients with anorexia and bulimia nervosa, our review only included RCTs studying physical activity interventions with AN patients.

Intervention effects of physical activity on anthropometry of anorexic nervosa patients

Doctors often attempt to restrain physical activity in anorexic nervosa patients simply because they think most of these patients may engage in excessive exercise, which can play an important role in the pathogenesis and ongoing weight loss of anorexic nervosa [25]. However, according to Elzbieta Kostrzewa [26], during treatment, the high-level physical-activity group decreased their total activity, while the low-level physical-activity group on average slightly increased their activity. From this point of view, doctors ought to worry whether patients with AN have appropriate exercise instead of excessive exercise during treatment. Although some studies have indicated that those who exercise (aerobic exercise, strength training or yoga) increase in weight and body fat compared with non-exercisers [27,28], our review does not support this finding. However, our meta-analysis suggests that appropriate physical activity with adequate food supply does not lead to a decrease in weight, BMI or body fat percentage for AN patients. In fact, adequate food supply is very necessary for anorexic nervosa patients to gain weight whether they exercise or not. From the 7 included studies, calorie intake was mentioned in 5, among which, two of the intervention groups increased calorie intake by means of a high-protein milkshake (150 kcal) because of the extra energy expenditure required during the strength training session. In one study [20], all the patients were eating a hospital-controlled diet of 2500 calories per day [16], and daily calorie intake was in the range of 1800–2500 kcal/day depending on the patient. In the remaining two articles [21,22], there were no specific food calories mentioned for the anorexic nervosa patients, but all participants were prescribed individual, nutritionally balanced diets. The consumption of energy increases with the increase of intensity and duration of physical activity [21]. Therefore, levels of physical activity should be matched with dietary intake.

Intervention effects of physical activity on skeletal muscle system

Compared with other eating disorders, fat, bone, and muscular tissues of anorexic nervosa patients are seriously affected and result in myopathy because of severe protein malnutrition and reduction in muscular strength [16,17,29]. Therefore, an appropriate intervention measure for increasing muscular strength is necessary except for adequate nutrition. The Chantler study [20] demonstrated that a light resistance training program of eight weeks increased the knee and elbow strength of anorexic patients in the experimental group. The del Valle study demonstrated that six dynamic repetitions of maximum strength significantly improved with training for the seated lateral row test, but no actual improvement solely attributable to the low to moderate-intensity strength training intervention was noted in the remaining strength/functional tests [16]. In another study, legpress, bench-press, and lateral row tests improved significantly after 8 weeks of high-intensity resistance training compared with controls, and improvements were still maintained after the subsequent 4-week detraining period. Two days/week frequency seemed to be suitable for strength maintenance programs, whereas 3 or more days/week were necessary for strength gaining programs [17]. One study [21] suggested that high-intensity resistance training led to significant differences in the percentage of change in arm muscle area when the intervention group and control group were compared. Similarly, significant SMM gains [18] were found in the intervention group and were sustained over time. There was another study in which routine care plus 20 jumps twice daily failed to influence markers of bone turnover in adolescents with AN but was well tolerated [22]. Acute bone loss was observed when healthy young men were placed on strict bed rest with a reduction of 1%-2% in whole body bone mass after only a month [30]. Thus, we must keep in mind that prolonged abstinence from exercise may contribute to decreased bone mass, whereas high aerobic exercise loads may also decrease bone mass density [19,20] and exacerbate the osteoporosis of AN patients. Only appropriate physical activity has the potential to increase bone mineral density and motor performance skills [23]. This may explain why the intervention measures in 5 of the included 7 RCTs included resistance training instead of aerobic exercise.

Intervention effects of physical activity on quality of life

Regular physical activity is also important in the prevention and treatment of mental disorders [31,32]. Vincent Thien’s study [19] indicated that the exercise group showed an improvement from baseline measures in all aspects of quality of life. In contrast, the control group showed a decrease in all aspects of quality of life from their respective baseline values, although the differences were not statistically significant. One study indicated that patients with AN were less depressed after an 8-week resistance training program [33]. In addition, strength training can improve the ability to carry out activities of daily living and contribute to an improvement in the overall health and well-being of anorexic patients [20]. The time for restoring vital-sign stabilization was significantly reduced in the intervention group [22]. Previous research [34] indicated that being given the opportunity to exercise during treatment increased overall compliance with the entire treatment program, including adherence to meal plans. A study [35] suggested that participating in this exercise mode had the potential to increase physical capacity and overall health status in children and adolescents. This study was carried out in healthy children and adolescents. Alin Andries’ study suggested that the EDI-2 scores were not significantly different between two groups with different densities of physical activity [21]. Exercise training appeared to significantly increase cardiovascular endurance [28]. Appropriate physical activity had a positive effect on mood, depression and self-esteem that might help to treat and prevent eating disorders [36].

Strengths and Limitations

This is the first systematic review and meta-analysis that only included RCT studies and was undertaken to assess the intervention effects of physical activity in patients with AN. In the meantime, the review also examined the intervention effects of physical activity on the skeletal and muscular system, stable vital-sign and quality of life in patients with AN. Moreover, most of the included studies were of high quality. Randomization sequence generation and attrition bias in all studies except one were characterized as low risk. However, this systematic review is limited by few studies and small sample sizes. In particular, only one RCT included more than 20 patients in each group [22]. Although we conducted an extensive database review, it is possible that articles were missed during the search due to publication and selective reporting biases.

Further randomized controlled trials with larger sample sizes and longer follow-up periods need to be performed to draw firmer conclusions. In addition, to establish guidelines for physical activity, future research should also explore the type, frequency, intensity and duration of effective and safe physical activity interventions for AN patients. Nutritional requirements under a certain amount of exercise also deserve additional research.

Conclusion

Thus far, exercise as a treatment for AN is still controversial. Most doctors are unwilling to prescribe exercise in that they worry that physical activity may worsen an already compromised state of health because of the increased caloric expenditure. However, our systematic review and meta-analysis found that there were no significant differences in weight, BMI, percentage of body fat and lean mass between the intervention group and the control group for patients with AN. Appropriate physical activity interventions might have beneficial outcomes for skeletal muscle systems, such as bone mineral density, muscular strength and agility. Moreover, appropriate physical activity interventions could shorten the time to vital-sign stabilization and improve quality of life. Therefore, the best practice may be to allow anorexic patients to perform a certain amount of physical activity and obtain extra nutrition to make up for additional energy expenditure due to exercise so that the treatment goals of weight gain are not compromised. Of course, exercise should be performed under scientific guidance.

Acknowledgment

The authors are grateful to Miss Zhang Wen for her help during the literature search. The authors also wish to express thanks to Miss Zhirong Wu and Yu Shi for their helping hands in reference management. This project (2016XQHLYG-01) is supported by a Clinical Research Fund from the Second Affiliated Hospital of Army Medical University.

Author Contributions

Study design: Q.Z. Data collection: L.C., X.S. and L.D.

Data analysis: R.X., Q.Z. Writing: Q.Z., R.X., L.C., and D.F.

References

1. James Morrison (2014) DSM-5 Made Easy [276-292] The Guilford Press, 72 Spring Street, New York, USA.
2. Smink FR, Van Hoeken D, Hoek HW (2012) Epidemiology of Eating Disorders: Incidence, Prevalence and Mortality Rates. Curr Psychiatry Rep 14(4): 406-414.
3. Harris EC, Barraclough B (1998) Excess Mortality of Mental Disorder. Br J Psychiatry 173: 11-53.
4. Hudson JI, Hiripi E, Pope HJ, Kessler RC (2007) The Prevalence and Correlates of Eating Disorders in the National Comorbidity Survey Replication. Biol Psychiatry 61(13): 348-358.
5. Agras WS (2001) The Consequences and Costs of the Eating Disorders. Psychiatr Clin North Am 24: 371-379.
6. Fairburn CG, Harrison PJ (2003) Eating Disorders. Lancet 361: 407-416.
7. O Brien KM, Vincent NK (2003) Psychiatric Comorbidity in Anorexia and Bulimia Nervosa: Nature, Prevalence, and Causal Relationships. Clin Psychol Rev 23(1): 57-74.
8. Konstantynowicz J (2011) Thigh Circumference as a Useful Predictor of Body Fat in Adolescent Girls with Anorexia Nervosa. Ann Nutr Metab 58(3): 181-187.
9. Bulik CM, Reba L, Siega-Riz AM, Reichborn-Kjennerud, T Anorexia Nervosa (2005): Definition, Epidemiology, and Cycle of Risk. Int J Eat Disord 37 Suppl S2-S9, S20-S21.
10. Hausenblas HA, Cook BJ, Chittester NI (2008) Can Exercise Treat Eating Disorders? EXERC SPORT SCI REV 36(1): 43-47.
11. Zunker C, Mitchell JE Wonderlich SA (2011) Exercise Interventions for Women with Anorexia Nervosa: A Review of the Literature. Int J Eat Disord 44(7): 579-584.
12. MoherD, Liberati A, Tetzlaff J, Altman DG (2009) Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. Ann Intern Med 151(4): 264-269, W64.
13. Higgins JP, ADGP (2011) The Cochrane Collaboration’S Tool for Assessing Risk of Bias in Randomized Trials. BMJ 343: d5928.
14. Guyatt GH (2008) GRADE: An Emerging Consensus on Rating Quality of Evidence and Strength of Recommendations. BMJ 336(7650): 924-926.
15. Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003) Measuring Inconsistency in Meta-Analyses. BMJ 327(7414): 557-560.
16. Del-Valle MF (2010) Does Resistance Training Improve the Functional Capacity and Well Being of Very Young Anorexic Patients? A Randomized Controlled Trial. J Adolescent Health 46(4): 352-358.
17. Fernandez-del-Valle M (2014) Resistance Training Enhances Muscular Performance in Patients with Anorexia Nervosa: A Randomized Controlled Trial. Int J Eat Disord 47(6): 601-609.
18. Fernandez-del-Valle M, Larumbe-Zabala E, Morande-Lavin G, Perez Ruiz M (2016) Muscle Function and Body Composition Profile in Adolescents with Restrictive Anorexia Nervosa: Does Resistance Training Help? Disabil Rehabil 38(4): 346-353.
19. Thien V, Thomas A, Markin D, Birmingham C (2000) Pilot Study of a Graded Exercise Program for the Treatment of Anorexia Nervosa. Int J Eat Disorder 28(1): 101-106.
20. Chantler I, Szabo CP, Green K (2006) Muscular Strength Changes in Hospitalized Anorexic Patients After an Eight Week Resistance Training Program. Int J Sports Med 27(8): 660-665.
21. Andries A, Gram B, Stoving R (2015) Effect of Dronabinol Therapy on Physical Activity in Anorexia Nervosa: A Randomised, Controlled Trial. Eating and weight disorders: EWD 20(1): 13-21.
22. Martin S, Bachrach L, Golden N (2017) Controlled Pilot Study of High-Impact Low-Frequency Exercise on Bone Loss and Vital-Sign Stabilization in Adolescents with Eating Disorders. J Adolescent Health 60(1): 33-37.
23. Vancampfort D, Vanderlinden J, De-Hert M, Soundy A, Adámkova M (2014) a Systematic Review of Physical Therapy Interventions for Patients with Anorexia and Bulemia Nervosa. Disabil Rehabil 36(8): 628-634.
24. Ng LW C, Ng DP, Wong WP (2013) Is Supervised Exercise Training Safe in Patients with Anorexia Nervosa? A Meta-Analysis. Physiotherapy (United Kingdom) 99(1): 1-11.
25. Davis C (1997) The Prevalence of High-Level Exercise in the Eating Disorders: Etiological Implications. Compr Psychiat 38(6): 321-326.
26. Kostrzewa E (2013) Longitudinal Changes in the Physical Activity of Adolescents with Anorexia Nervosa and their Influence on Body Composition and Leptin Serum Levels after Recovery. Plos One 8(10): e78251.
27. Carei TR, Fyfe-Johnson AL, Breuner CC, Brown MA (2010) Randomized Controlled Clinical Trial of Yoga in the Treatment of Eating Disorders. J Adolesc Health 46(4): 346-351.