|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Departments of Medicine (Dr. Goldstein) and Physical Therapy (Dr. Brooks), University of Toronto; the Respiratory Medicine Program (Dr. Ferreira), West Park Hospital, Toronto, Ontario; and Centre de Pneumologie (Dr. Lacasse), Hopital Laval, Ste-Foy, Quebec.
Correspondence to: Ivone M. Ferreira, MD, PhD, Respiratory Medicine Program, West Park Hospital, 82 Buttonwood Ave, Toronto, Ontario, M6M 2J5, Canada; e-mail: ivoneferreira{at}hotmail.com
| Abstract |
|---|
|
|
|---|
Purpose: To conduct a meta-analysis of randomized controlled trials (RCTs) to clarify whether nutritional supplementation (caloric supplementation for at least 2 weeks) improved anthropometric measures, pulmonary function, respiratory muscle strength, and functional exercise capacity in patients with stable COPD.
Methods: RCTs were identified from several sources, including the Cochrane Airways Group register of RCTs, a hand search of abstracts presented at international meetings, and consultation with experts. Two reviewers independently selected trials for inclusion, assessed quality, and extracted the data. Within each trial and for each outcome, we calculated an effect size. The effect sizes were then pooled by a random-effects model. Homogeneity among the effect sizes was also tested.
Results: From 272 references, nine RCTs were ultimately included. Six articles were considered as high quality. Only two studies were double blinded. For each of the outcomes studied, the effect of nutritional support was small: the 95% confidence intervals around the pooled effect sizes all included zero. The effect of nutritional support was homogeneous across studies.
Conclusion: Nutritional support had no effect on improving anthropometric measures, lung function, or functional exercise capacity among patients with stable COPD.
Key Words: COPD meta-analysis nutrition respiratory rehabilitation systematic review
| Introduction |
|---|
|
|
|---|
Several studies have shown an association between malnutrition and impaired pulmonary status among patients with COPD.8 9 10 Individuals with low weight have more gas trapping, lower diffusing capacity, and lower exercise capacity than those with similar pulmonary mechanics but with normal weight.8 In addition, malnourished patients with COPD had a higher mortality rate than those whose nutrition was adequate.9 Both a low body mass index (BMI) and the use of supplemental oxygen in the home are independent predictors of reduced survival among hypoxic patients with COPD.11
Several factors contribute to the impaired respiratory status in malnourished patients with COPD. As with healthy individuals, malnutrition impairs skeletal muscle function.10 It also results in a reduced diaphragmatic mass associated with a decrease in both strength and endurance of the respiratory muscles.10 12
Because the additional presence of malnutrition might be disadvantageous for some patients with COPD, nutritional support has been suggested as a useful management approach. Although in some ambulatory COPD patients, nutritional repletion resulted in improvements in respiratory and limb muscle function,13 14 15 this has not always been the case.16 17 Furthermore, it remains unclear whether improvements in respiratory and peripheral muscle function will result in better functional exercise capacity or improvements in health-related quality of life. A recent study by Schols and colleagues18 provided evidence that for some patients with COPD, the negative effect of low body weight can be reversed by appropriate therapy.
Several reviews have endorsed nutritional support in COPD.1 13 19 However, there have been no systematic reviews of the literature that might help summarize its effect. Therefore, we undertook a systematic review of randomized controlled trials (RCTs) of nutritional supplementation to identify whether it has been shown to influence anthropometric measures, pulmonary function tests, respiratory and peripheral muscle function, functional exercise capacity, or health-related quality of life in patients with COPD.
| Materials and Methods |
|---|
|
|
|---|
Study Selection
Two reviewers (I.M.F. and D.B.) independently selected trials
for inclusion in the review according to prior agreement regarding
definitions and interventions. Disagreement was solved after
consultation with a third reviewer (Y.L.). We used the following
criteria to select RCTs for inclusion in the meta-analysis:
Study Design:
Only RCTs were considered for inclusion in the
meta-analysis.
Nutritional Support:
Nutritional support was defined as any
caloric supplementation > 2 weeks.
Study Population:
To be selected, trials must have included
stable patients among whom at least 75% had COPD characterized by a
FEV1 < 70%, and < 15% reversibility after
bronchodilator. Subjects received oral, enteral, or parenteral
nutritional support vs placebo or vs their usual diet or other
treatment regimens such as anabolic substances. Studies involving
patients with COPD undergoing treatment in the ICU were excluded.
Types of Outcomes:
The primary measures of outcome were
anthropometric (body weight, lean body mass, BMI) and functional
exercise (timed walk test, submaximal or graded exercise). Other
measures included pulmonary mechanics (lung volumes, respiratory muscle
function), peripheral muscle function, and health-related quality of
life.
Data Extraction
The data and trial quality information were extracted
independently by two reviewers (I.M.F. and D.B.). Missing data from the
primary study reports was requested from the investigators.
Methodologic quality of the trials was assessed using a validated scale
(range, 0 to 5) applied to items that influence intervention efficacy.
The scale consisted of items pertaining to randomization, masking,
dropouts, and withdrawals.20
Data Analysis
In each trial and for each outcome, we calculated the treatment
effect from the difference between the preintervention and
postintervention changes in the treatment and control groups. We
standardized the resulting treatment effect to obtain an effect size by
dividing the treatment effects by the pooled SD of the postintervention
outcome measure in the treatment and control groups. The effect sizes
were weighted by the inverse of the population variance and combined
according to a random effects model, (ie, studies
represented a random sample from a larger population). Homogeneity
among studies was also tested. The pooled effect sizes plus the
corresponding 95% confidence intervals (CIs) were reported for each
outcome in SD units, and then converted back to natural units of the
most commonly used measure.21
22
A Priori Hypotheses Explaining Heterogeneity Among Studies
We identified a priori hypotheses concerning the
underlying differences in the studies that might explain heterogeneity
(ie, major differences in the apparent effect of the
intervention across studies). We hypothesized the following: (1)
undernourished patients benefit more from supplementation than
nourished patients; (2) the benefits of supplementation are linked to
the amount of supplementation; (3) the length of supplementation
determines the extent of the improvement; (4) supervised
supplementation results in better improvements in anthropometric
measures and exercise capacity than unsupervised supplementation; and
(5) nutritional support combined with exercise training is better than
supplementation alone; and (6) the results of the trials are also
determined by their methodologic quality.
| Results |
|---|
|
|
|---|
= 0.89; 95% CI, 0.82 to
0.96). These were read in detail. Of these studies, 34 were excluded
(
= 0.94; 95% CI, 0.85 to 1.00). Both reviewers agreed on all
except four reports that were therefore sent to a third reviewer (Y.L.)
for arbitration. Reports were excluded for the following reasons:
uncontrolled study,9
11
23
24
25
26
no
randomization,27
28
investigations into weight loss
without nutritional support,29
30
31
32
33
34
35
36
37
38
39
nutritional assessment
without nutritional support,3
8
40
41
42
43
44
45
46
47
48
subjects not
meeting eligibility criteria,49
50
51
52
and second report in
the same trial.53
The reports were then classified
according to the type and duration of supplementation and the presence
of anabolic substances. Only two articles were retrieved in which
anabolic steroid supplementation was reported.54
55
In an
additional article56
and abstract,57
growth
hormones were administered. Articles in which the influence of one meal
was reported58
59
60
61
were excluded, as were articles that
reported supplementation of < 2 weeks.62
63
64 Of the reports that dealt with supplementation for > 2 weeks, a total of 10 articles were eligible for inclusion.14 15 16 17 65 66 67 68 69 In four instances, the authors did not respond to repeated requests for further information. In two of these studies, two reviewers independently obtained the relevant information by expanding published figures,16 17 whereas the remaining two studies could not be included.68 69 Table 1 shows a summary of the included articles. We were able to include a total of nine RCTs, since one trial55 was considered as two studies, the data having been reported separately in undernourished and nourished patients. For one crossover trial,17 only the data from the first part of the study was included.
|
A total of 277 subjects (144 study and 133 control) were included in the analysis. Of these, two studies included an inpatient component and seven were entirely outpatient based. All studies except one used oral supplementation. The primary results of the meta-analysis are shown (Table 2 ). For each of the outcomes studied, the effect of nutritional support was small: the 95% CIs around the pooled effect sizes all included zero. Changes in the 6-min walk test did not exceed the minimum clinically important difference (defined as the smallest difference perceived as important by the average patient). The effect of nutritional support was homogeneous across studies (p values ranged from 0.19 to 1.00).
|
| Discussion |
|---|
|
|
|---|
Studies of nutritional support in which the observer is blinded and
impartial to the group allocation present challenges in both design and
execution. Controlling for nutritional support in the outpatient
setting is especially difficult, as only partial supervision is
possible. If the supplements are associated with side effects such as
bloating, fullness, or dyspnea, subjects are tempted to reduce the
supplement or alternatively to reduce their usual nutritional intake
such that their caloric intake falls.17
Only six of the
studies reviewed were considered to be of high quality as reflected by
their achieving a score of
2 in Jadads study design assessment
scale,72
and only two studies were double blinded.
The absence of a response to nutritional support likely reflected the
multifactorial mechanisms in COPD by which an increased energy
expenditure is not balanced by an adequate dietary intake. The judgment
of energy expenditure might have been incorrect, and the subjects may
have had limited ability to ingest the required energy intake. Thus,
there is always the possibility that the intervention itself was not of
sufficient magnitude to produce an effect. However, since the effect
size was consistently small despite varied amounts of nutritional
support, we think that the results are unlikely to reflect a failure to
intervene. Other explanations for a poor response to nutritional
support have included an increased metabolism, diet-induced
thermogenesis, tissue hypoxia, and medications such as corticosteroids.
Studies evaluating the influence of elevated levels of soluble tumor
necrosis factor-
receptors, have been reported.5
6
7
Saudny-Unterberger et al,50
in a study of patients during
an acute exacerbation of COPD, found no change in weight, hand grip, or
respiratory muscle strength as a result of nutritional
supplementation.50
Inevitably, the meta-analysis involved a retrospective analysis that was subject to the limitations of any retrospective study. However, the primary scientific studies from which the data was collected were all prospective randomized and controlled.73 Although a small total number of subjects (144 in the experiment group and 133 in the control group) is sometimes associated with a bias toward overestimating a treatment effect, this issue does not apply to the current study as the results were negative.
Weight was the most frequently reported outcome. As alterations in weight can be due to changes in fat,23 and given that many patients with COPD have a reduced fat-free mass, nutritional support would more accurately be assessed against lean body mass. However, this outcome was reported in only two of the included studies.55 65 In one of these studies, the data was presented as percent change lean body mass. This unit of measure could not be included in the analysis, and the raw data was not supplied.
The meta-analysis would have been enriched with the addition of information regarding health-related quality of life. Unfortunately, too few studies reported dyspnea or sense of well-being for these measures to be included as outcomes. During the 3 months of dietary supplementation, Efthimiou et al14 did note improvement in breathlessness and general well-being that fell gradually over the subsequent 3 months. In contrast, Otte et al,67 (using a different scale) did not identify changes in well-being associated with supplementation.
The meta-analysis would also have been enriched by a larger number of subjects. With only nine trials included (six of high quality), we tried to maximize the available data by contacting experienced investigators in this field. Only half of those contacted responded. However, the two randomized trials excluded because of incomplete data would likely not have altered the conclusion of the meta-analysis, as one was positive and one was negative.
The study groups tended to be small, except for the two trials by Schols et al55 that together accounted for almost 50% of the total sample. Measures of weight and anthropometry were available for each of the trials, whereas fewer trials included exercise (three trials) or respiratory muscle strength (four trials).
Notwithstanding the above points, the meta-analysis did increase the statistical power of the primary end points, in comparison with any of the individual trials, thus narrowing the CIs around the estimates of the magnitude of the treatment effect. It also served to resolve uncertainty when reports disagreed.74
The implications of this meta-analysis will be troubling for clinicians who note the harmful effects of weight loss in COPD and the associated reduced scores for health-related quality of life,75 the higher risk of infection, and the higher mortality.9 Post hoc analysis of COPD patients who have gained weight has suggested a decrease in mortality,18 and at least one study has reported that improved immune function occurred as a result of nutritional support.66 The administration of an investigational immunostimulant reduced the frequency and duration of hospitalization as well as the number of deaths in subjects with COPD.76
In summary, this meta-analysis of nutritional support did not identify improvements in anthropometric measures or functional exercise capacity among patients with stable COPD. Alternative approaches such as the combination of nutritional support with anabolic substances (such as anabolic steroids or growth hormone) may be of value to undernourished patients and should be explored.54 55 56 57 Further studies should include the influence supplementation on health-related quality of life.
| Acknowledgements |
|---|
| Footnotes |
|---|
This study was made possible due to funding from FAPESP, Fundacao de Amparo a pesquisa do Estado de Sao Paulo, Brazil (I.M.F.), the Canadian Lung Association/MRC/Glaxo Welcome Postdoctoral fellowship (D.B.), and West Park Hospital Foundation.
Received for publication May 18, 1999. Accepted for publication September 9, 1999.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
G. P Currie and J G. Douglas Non-pharmacological management. BMJ, June 10, 2006; 332(7554): 1379 - 1381. [Full Text] [PDF] |
||||
![]() |
A. G. N. Agusti Systemic Effects of Chronic Obstructive Pulmonary Disease Proceedings of the ATS, November 1, 2005; 2(4): 367 - 370. [Abstract] [Full Text] [PDF] |
||||
![]() |
A.M.W.J. Schols Nutritional and metabolic modulation in chronic obstructive pulmonary disease management Eur. Respir. J., November 2, 2003; 22(46_suppl): 81S - 86s. [Abstract] [Full Text] [PDF] |
||||
![]() |
M C Steiner, R L Barton, S J Singh, and M D L Morgan Nutritional enhancement of exercise performance in chronic obstructive pulmonary disease: a randomised controlled trial Thorax, September 1, 2003; 58(9): 745 - 751. [Abstract] [Full Text] [PDF] |
||||
![]() |
M D L Morgan and J R Britton Chronic obstructive pulmonary disease * 8: Non-pharmacological management of COPD Thorax, May 1, 2003; 58(5): 453 - 457. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. L. Croxton, G. G. Weinmann, R. M. Senior, R. A. Wise, J. D. Crapo, and A. S. Buist Clinical Research in Chronic Obstructive Pulmonary Disease: Needs and Opportunities Am. J. Respir. Crit. Care Med., April 15, 2003; 167(8): 1142 - 1149. [Abstract] [Full Text] [PDF] |
||||
![]() |
A.G.N. Agusti, A. Noguera, J. Sauleda, E. Sala, J. Pons, and X. Busquets Systemic effects of chronic obstructive pulmonary disease Eur. Respir. J., February 1, 2003; 21(2): 347 - 360. [Abstract] [Full Text] [PDF] |
||||
![]() |
M D L Morgan Preventing hospital admissions for COPD: role of physical activity Thorax, February 1, 2003; 58(2): 95 - 96. [Full Text] [PDF] |
||||
![]() |
E. Prescott, T. Almdal, K.L. Mikkelsen, C.L. Tofteng, J. Vestbo, and P. Lange Prognostic value of weight change in chronic obstructive pulmonary disease: results from the Copenhagen City Heart Study Eur. Respir. J., September 1, 2002; 20(3): 539 - 544. [Abstract] [Full Text] [PDF] |
||||
![]() |
British Thoracic Society Standards of Care Subcomm Pulmonary rehabilitation Thorax, November 1, 2001; 56(11): 827 - 834. [Full Text] [PDF] |
||||
![]() |
G. Akner and T. Cederholm Treatment of protein-energy malnutrition in chronic nonmalignant disorders Am. J. Clinical Nutrition, July 1, 2001; 74(1): 6 - 24. [Abstract] [Full Text] [PDF] |
||||
![]() |
I. M. Ferreira, D. Brooks, Y. Lacasse, and R. S. Goldstein Nutritional Intervention in COPD : A Systematic Overview Chest, February 1, 2001; 119(2): 353 - 363. [Abstract] [Full Text] [PDF] |
||||
![]() |
M C Steiner and M D L Morgan Enhancing physical performance in chronic obstructive pulmonary disease Thorax, January 1, 2001; 56(1): 73 - 77. [Full Text] |
||||
![]() |
L. Ezzell and G. L Jensen Malnutrition in chronic obstructive pulmonary disease Am. J. Clinical Nutrition, December 1, 2000; 72(6): 1415 - 1416. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |