To examine the effects of exercise on bone mineral density in postmenopausal women.

MEDLINE was searched from January 1975 to December 1995, and Current Contents from January 1993 to December 1995, using the following keywords either alone or in combination: 'exercise', 'humans', 'physical activity', 'fitness', 'bone', 'bone density', 'bone mineral density', 'postmenopausal', 'women', 'female', 'bone mass' and 'osteoporosis'. The search was limited to full-length papers published in the English language.

The author also cross-referenced all retrieved review articles and original investigations.

A funnel plot was used to test for publication bias.

Study designs of evaluations included in the review

Clinical randomised exercise trials where the studies reported changes in regional bone mineral density, where exercise was the primary intervention in postmenopausal women, and a comparative non-exercise group was included.

Specific interventions included in the review

Exercise (aerobic training in 6 studies, and strengthening exercises in 5 studies) ranging from 15 to 90 minutes in duration, 2 to 7 times a week over an average period of approximately 7 to 39 months. The control group had no exercise. Some of the studies combined exercise with oestrogen or calcium supplements. Four studies had both groups taking hormone replacement therapy (HRT), 6 studies had no HRT for either group, and 1 study had no information about HRT status.

Participants included in the review

The participants were postmenopausal women whose mean age ranged from 43 to 76 years.

Outcomes assessed in the review

Changes in regional bone mineral density were measured at the femur, lumbar, radius or forearm, and ultradistal ankle sites. These changes were measured using dual-energy X-ray absorptiometry (5 studies), dual-photon absorptiometry (3 studies), single-photon absorptiometry only (1 study), and computed tomography scan (1 study). One study used both single- and dual-photon absorptiometry.

How were decisions on the relevance of primary studies made?

One author examined the methods and results sections separately under coded conditions.

The author did not state that they assessed validity.

The author used a coding sheet to extract up to 288 pieces of information for each study. The author was blinded to the author(s), journal and year of publication. The author coded the same studies twice, two weeks apart, to establish the intra-rater reliability. The coding sheet was then modified until there was 100% agreement between both coding sessions.

How were the studies combined?

The treatment effect changes in bone density, weighted by sample size, were combined using computer statistical packages to give an overall treatment effect change for all groups. The statistical methods used to combine the studies were not reported. Significance was set at a p-value of less than or equal to 0.05.

With the exception of changes in regional bone mineral density, all other comparisons between exercise and non-exercise groups were examined using independent t-tests. If normality and/or equal variance were violated, then Mann-Whitney rank-sum tests were used.

Subgroup analyses were also performed for mode of training, site of bone mineral density assessed, site-specificity of bone mineral assessment, method of assessing bone mineral density, and pharmacologic intervention.

Spearman's correlation tests were used to assess differences between treatment effect changes in terms of regional bone mineral density and study characteristics, physical characteristics, bone density assessment, training programme characteristics, and initial bone mineral density.

How were differences between studies investigated?

The chi-squared Q statistic was used to test for heterogeneity.

Eleven randomised controlled trials with a total of 719 participants (370 in the exercise group and 349 in the non-exercise group) were included.

The funnel plot used to examine publication bias was symmetrical.

There was an overall treatment effect change in bone density for all groups of 0.27% (range: -17.10 to 17.30; 95% confidence interval: 0.16, 0.37).

Significant heterogenity was observed among treatment effects (Q=588.72, p=0.00). A total of 27 outliers were identified, but since there was no physiologic reason for their exclusion they remained in the analysis.

In the subgroup analysis, confidence intervals for the original treatment effects were significant for the following: training mode (strength only), site measured (lumbar and femur only), method of assessing bone density (dual-photon absorptiometry and dual-energy X-ray absorptiometry), exercise groups that were not matched with a non-exercise group that received calcium and/or HRT, and sites that measured bone density specific to the site loaded.

No important relationships were found in the correlation analysis.

The author states that the meta-analytic review suggested that exercise may slow the rate of bone loss in postmenopausal women, but it is premature to form strong conclusions regarding the effects of exercise on regional bone mineral density in postmenopausal women.

This was a reasonably complete systematic review. The author searched several databases and cross-checking original studies. However, the search was limited to English language reports, so it is unclear whether additional studies may have been missed. The inclusion criteria were stated, but the author did not assess either the validity or quality of the included studies. The data were extracted by the author alone, and although this task was repeated, it does not meet the criteria of independent selection or the extraction criteria.

The author states that there was significant heterogeneity between the studies in this review and that the sample sizes were small. Despite this, the studies were still combined; there were insufficient details of the statistical methods used to combine the studies. There was a jump from the description of the individual studies to the overall treatment effects and the confidence intervals CIs tabulated in the review.

The conclusions of the review, that the rate of bone loss for postmenopausal women can be slowed by exercise, should be viewed with great caution: there was only a partially significant finding that strength exercises provided this benefit in the group calculations, and other statistically-significant results were obtained by the exclusion of certain groups of participants.

The author states additional well-designed studies, which examinate the interactions between primary outcome measures and study design variables, are needed before a recommendation can be made regarding the efficacy of exercise as a non-pharmacologic therapy for maintaining and/or increasing regional bone mineral density in postmenopausal women.

Northern Illinois University, Research and Artistry grant.

Kelley G. Aerobic exercise and lumbar spine bone mineral density in postmenopausal women: a meta-analysis. J Am Geriatr Soc 1998;46:143-52.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.