This month a group of Canadian doctors published their research on reevaluating asthma diagnoses among adults, and shockingly they found that in almost 1/3 of participants asthma could be ruled out through systematically removing medications. This post is going to explore this study in a little more detail than pop-sci, but without requiring the reader to wade through the impenetrability of academic writing.


Patients were recruited through randomly dialing landline and mobile phone numbers over a 3 year period, from January 2012 through to February 2015. The researchers aimed their recruitment at 10 metropolitan areas, intending to achieve an approximation of the Canadian adult population.

To be eligible patients had to be over 18, i.e. an adult, and to have been diagnosed with asthma in the last 5 years. The rationale behind the 5 year cut-off was to only look at modern diagnostic practices. This does mean that participants could have been diagnosed while in childhood.

A further set of exclusions include:

  • Long-term use of oral steroids
  • Pregnant and/or breastfeeding
  • Inability to perform spirometry or brochial challenge tests
  • Smoking history > 10 pack-years

So these require some explanation, as they definitely could alter the outcomes of the study. Steroids in this context are not the ones of Olympic or bodybuilding fame. These kinds of steroids are used to suppress inflammation, and are used in a wide array of conditions such as rheumatoid arthritis, ulcerative colitis, and importantly asthama. However oral steroid use in asthma is limited to quite severe forms according to British Thoracic Society guidelines (treatment is very similar on both sides of the Atlantic), therefore by excluding oral-steroid-using asthmatics, they’re excluding a group of severe asthmatics, and thus people who are unlikely to be found not actually have asthma.

Some spirometry volumes
Some spirometry volumes

Spirometry essentially involves breathing into a plastic tube, while the volumes in different phases of breathing are measured. Some normal volumes in breathing are pictured above. Another important figure not pictured is the Forced Expiratory Volume in 1 second (FEV1), or in other words the largest volume of air that can be breathed out in 1 second. FEV1 is a key figure, and is reduced by conditions that obstruct the airways (such as asthma). Normally spirometry is not a challenging test, so people unable to perform spirometry are also going to be quite sick.

Bronchial challenge testing involves inhaling either methacholine or histamine, and measuring the FEV1 afterwards. For our purposes all we need to know about methacholine and histamine is that they cause bronchoconstriction, which means narrowing the airways. The actual test involves steadily increasing the amount of drug inhaled until FEV1 ahs decreased by 20%. The rationale behind the test is that people with sensitive, or hyperreactive, airways, such as asthamtics, will need less inhaled drug to cause a 20% decrease in FEV1.

Unlike regular spirometry, bronchial challenge testing can be quite demanding. As well as not wanting to exacerbate existing severe conditions, the test also increases blood pressure, so is avoided in people who have aneurysms or recent stroke.

Moving on to the testing procedures, this study used two different methods to confirm asthma. Initially patients had their FEV1 measured before and after taking the bronchodilator albuterol. For those familiar with asthma medications, this is the blue inhaler, and used to relieve symptoms in the short term.

For the purposes of this study, an increase in FEV1 of 12% or greater (and a minimum of 200mL) was diagnostic of asthma. From what I can find this figure seems relatively well supported (McCormack and Enright), as to is the use of bronchodilator spirometry for diagnosing asthma. It’s important to note that the participant was on all of their regular medication at this point.

After a diagnosis of asthma was confirmed the pariticipant was removed from the trial group, and so did not progress to the next stage of testing.

The later forms of testing all take the same form, performing bronchial challege tests after having some of their medication removed. In this study if a methacholine concentration of less than 8 mg/mL caused a 20% decrease in FEV1 then the subject was assumed to have asthma. Again this value seems sensible and well backed up in the literature (Hargreave et al).

This test was performed three times, the first while on all regular medications. From here patients were asked to halve all inhaled corticosteroids and long acting bronchodialtors, come off all anti-leukotriene medications, and be retested in 3 weeks. If this still didn’t provoke a confirmation of asthma they were asked to come off even the inhaled corticosteroids and long acting bronchodilators, and to be retested in another 3 weeks.

Looking at the biological half lives of these two sets of medications, they are short enough, in the order of hours, that a 3 week buffer period should be enough to remove interference with the test.

Finally as a catch all, participants still not confirmed to have asthma saw a respiratory specialist and were followed up with for 12 months.


After all rounds of testing 203 (33.1%) out of a total of 613 had a diagnosis of asthma ruled out.

Participants who were found to not have asthma had significantly better lung function, were less likely to be taking asthma medications, and reported lower rates of tests like spirometry and bronchial challenge testing being used in their diagnosis. Furthermore diagnoses made in primary care were more likely to be overruled.

Concerningly, of the 213 participants who made it to the specialist consultation, 12 were found to have cardiorespiratory conditions instead, indicating that their asthma diagnosis could be covering up life-threatening conditions.


Overall this study presents compelling evidence that asthma is overdiagnosed among Canadian adults. In practice this is unlikely to be as high as the 33.1% found in this study in part due to the exclusions.

This is noted by the authors who discuss that the study may be biased towards the milder group of asthmas, through the exlusion of long-term oral corticosteroid users. Certainly the results indicate that diagnosis was more likely to be in error if the participant was only an infrequent user of asthma medication, implying that milder asthma is more likely to be the target of misdiagnosis.

Another point of interest is that asthma is a condition that can resolve with time, certainly in children (Robertson) and to some extent in adults. Such is the tendency of overmedication in asthma, that most guidelines recommend reducing levels of medication after 3 months of good control and seeing how well tolerated the reduced doses are.

One of the difficulties of asthma is that it’s largely a diagnosis made on patient history, and its severity can be dependent on a wide array of factors. An allergen present at diagnosis may not be around when the diagnosis is reevaluated for example. This would show up as a misdiagnosed asthma that nonetheless would be problem should the person be exposed to the allergen.

What this means for patients

The message of this paper is clear, if you have been diagnosed in the last 5 years with asthma, it may have been a mistake, especially if you use your medications infrequently and can’t remember having spirometry tests.

Fortunately this can all be part of your treatment review you should be having after 3 months of good control according to British and American Thoracic Society guidelines, and may be worth raising with your doctor.