Saturday, May 7, 2011

Activity vs Disability in Low-Back Pain Assessed

Back Pain It is often assumed that patients with disability related to low-back pain (LBP) will have reduced physical activity, and treatment emphasizes maintaining or gradually increasing activity. Now, researchers have found that there is no significant correlation between physical activity level and disability in acute/subacute LBP. In chronic LBP, as might be expected, persons with higher levels of disability are less likely to be physically active, and whether increased activity would be helpful for them is largely unknown.

Reporting in the journal Pain, an international team of investigators conducted a meta-analysis to examine the relationship between physical activity and disability in LBP [Lin et al. 2011]. Their literature search included studies up to May 2010 that measured both disability (eg, with the Roland Morris Disability Questionnaire) and physical activity (eg, by accelerometry) in patients with non-specific LBP.

The search identified 3,213 records and 18 studies were eligible for inclusion; correlation coefficients from individual studies were pooled using a random-effects model. Results showed a statistically non-significant and negligible relationship between physical activity and disability in acute or subacute (<3 months) LBP (r = −0.08, 95% confidence interval −0.17 to 0.002). However, for chronic (>3 months) LBP, there was a statistically significant, moderate, and negative relationship (r = −0.33, 95% confidence interval −0.51 to −0.15). The negative correlation coefficient, or r, reflects the inverse relationship between decreasing physical activity and increasing disability.

Overall, the study suggests that persons with acute or subacute LBP appear to vary widely in their levels of physical activity independently of their pain-related disability. Whereas, persons with chronic LBP and high levels of disability are also likely to have low levels of physical activity, as might be expected.

COMMENTARY: The authors note that an assumption of physical activity and disability being negatively correlated — whereby persons with LBP who have high levels of disability also have low levels of activity — had not been previously confirmed. Disability has been defined as an umbrella term for decreased capacity to perform, or altered performance of, normal activities of daily living. Physical activity is generally associated with bodily movement beyond a resting state and/or the execution of a task or action. Therefore, where disability focuses on what persons cannot do, physical activity focuses on what persons are able to do or actually do in daily living.

This study found that some persons with acute/subacute LBP will be able to maintain a reasonable level of physical activity despite reporting pain-related disability, but some reporting low levels of disability also will have limitations in their physical activity. In short, there appears to be no firm association of physical activity and disability in these patients. And, the authors suggest that further research is needed to determine whether a low level of physical activity in acute/subacute LBP is a prognostic factor for developing more chronic pain.

Treatments recommended for LBP typically promote physical activity to aid recovery and reduce disability: in acute LBP, persons are advised to stay active and avoid bed rest; in chronic LBP, active approaches, such as cognitive behavioral therapy and exercise, are recommended over more passive treatments. However, this current research raises questions about the extent to which providing interventions that increase levels of physical activity in patients with LBP will reduce pain-related disability.

NOTES about meta-analysis: This approach gathers together the best available studies and treats the accumulated data as a single, larger body of evidence for analysis. While meta-analysis is a high level of evidence, it is only as good as the quality of the individual studies that are combined. In this present meta-analysis, Lin et al. do not comment on the evidentiary quality of the studies that they included in their analyses.

Of the 7 studies combined to examine acute/subacute LBP, 4 were clearly non-significant statistically (P >0.05) and the other 3 were only marginally significant (P ≈0.05); hence, the pooled 95% confidence interval (−0.17 to 0.003) clearly includes the null effect (r=0.0) and is not statistically significant. Therefore, one may confidently also conclude that there most likely was no clinically important association of physical activity and disability in these patients.

Whereas, for chronic LBP, there was a heterogeneous mix of 14 studies for analysis, with 6 showing a robust and statistically significant negative correlation between physical activity and disability (P >0.01), and 2 others having less pronounced but still significant associations. The remaining 6 studies were not statistically significant, but several of the studies were small in scale and likely underpowered to detect significant effects. The 95% confidence interval (r = −0.15 to −0.51), which does not include the 0.0 null effect and is statistically significant, suggests there is a clinically meaningful inverse association — increasing disability related to decreasing activity — ranging from a small to large effect. However, toward the lower end of the range (−0.15), the clinical importance of the outcome might be questioned and further research would help to confirm the strength of the association.

For a better understanding of these statistical concepts, see prior UPDATES articles in our “Making Sense of Pain Research” series discussing the interpretation of P-values [here] and confidence intervals [here].

REFERENCE: Lin CC, McAuley JH, Macedo L, et al. Relationship between physical activity and disability in low back pain: A systematic review and meta-analysis. PAIN. 2011(Mar)152(3):607-613 [abstract here].