Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Herbal remedies for asthma

Review
Results
Comment

Overheard in a post office recently was a comment from one person to another, with the effect that a grandchild was being properly treated with natural herbal remedies for its asthma, and wasn't the mother doing a good thing. Hard to bite one's tongue, sometimes, but without the exact evidence to hand, or even with it, silence is golden.

Given that about four million people in the UK have asthma, the question about efficacy of alternative therapies will be a frequently-asked question. How can it be answered? A systematic review has found all the randomised trials of herbal remedies [1]. How should we think about evaluating that evidence to reach our own conclusions when the question is asked of us?

Review


Searching for studies was very thorough. It consisted of using a range of electronic databases, including the Cochrane Library, with a limit at end 1999. Other files and researchers were interrogated, No language limitation was imposed. Any randomised trial of any herbal remedy was included.

Asthma diagnosis allowed use of recognised criteria or reversible airway constriction. Outcomes were lung function tests, symptoms, use of medicines and events like healthcare use, days lost from work or school, or steroid use. In the review, lung function tests were used, with a change of 15% or more considered clinically relevant.

Results


Seventeen randomised trials were found, six using traditional Chinese medicines, eight using traditional Indian preparations, and three others. No two studies used the same herbal remedy in the same preparation, as best we can tell. Only nine of these trials were described as double blind, and the overall methodological quality was poor. Using a scale of up to five points for methodological quality [2], 10 trials scored two points or fewer, four scored 3 and three scored 4. Nine of the 17 trials reported a clinically relevant improvement in lung function scores.

Comment


This brief rehearsal of the results does not do justice to a fine review that summarises each of the seventeen studies and is thoughtful. The review concludes that there is no definitive evidence for any herbal preparation, and the discussion on the potential for harm is informed and informative. But there is an opportunity to use the results to hone our skills about how to think beyond the bare result, even if it is statistically significant. This means thinking about quality, bias, validity and relevance.

What we might want to do is to consider what criteria evidence from randomised trials has to meet before becoming acceptable. We know, for instance, that studies with a quality score of 2 or less tend to be more positive in their results, a form of quality bias. So studies may have to have quality scores of three or more. Then there is the validity question. Asthma is a chronic disorder, so for relevance we might want to use only studies that have a duration of at least four weeks.

If we use only those criteria, what do the results look like in terms of the number of studies that meet them? The Table shows this. The number of studies with quality score of at least 3, and a duration of four weeks or more is three, all traditional Indian remedies.

Table: Randomised trials of herbal remedies for asthma

Table: Randomised trials of herbal remedies for asthma
Type of herbal remedy Total number of studies Number with quality score 3 or more And with duration 4 weeks or more And have result favouring herbal remedy
Chinese 6 0 0 0
Indian 8 5 3 1*
Other 3 2 0 0
* note that this study, published in 1974, had peak benefit at four weeks

Only one of these has positive outcomes. This study, published in 1974, had 123 bronchial asthmatics (age unreported) enrolled and measured lung function at various times over 12 weeks. The peak benefits were at four weeks.

Now we could have chosen other criteria of validity, like size (small trials may be biased). Or we could have chosen year of publication (we want up-to-date information) or place conducted (concerned about geographical bias). But we don't need them, and if we did use them, we'd get much the same answer.

The point, at the end, is this. The evidence for effectiveness for herbal remedies in asthma doesn't amount to a hill of beans. Those who promote them do a disservice to consumers. We can be confident that no-one has a clue whether they work or not, and that no conventional medicine would be licensed with the evidence available.

References:

  1. H Huntley, E Ernst. Herbal remedies for asthma: a systematic review. Thorax 2000 55: 925-929.
previous or next story in this issue