[ 10 August 2004 ]

Swedish study which claims to have found link between asthma and phthalates is flawed says industry

European Council for Plasticisers and Intermediates (ECPI) statement

A Swedish study, conducted by a group of scientists who claim to have found a link between asthma and the use of phthalate plasticisers, is flawed says the plasticiser industry.

The study entitled "The association between asthma and allergic symptoms in children and phthalates in house dust: a nested case-control study" was published online by Environmental Health Perspectives, on July 15, 2004.

The Scientists, led by Carl-Gustaf Bornehag of the Swedish National Testing and Research Institute, report what they call statistically significant associations between butylbenzyl phthalate (BBP) levels in dust and the induction of rhinitis and eczema and between di(2-ethylhexyl) phthalate (DEHP) levels in dust and asthma. They assert that the different associations "can be explained by a combination of chemical physical properties and toxicological potential." and that the results have "global implications."

However, the study makes only an extremely weak correlation between the incidence of asthma and the use of phthalate plasticisers. Furthermore, the European Council for Plasticisers and Intermediates (ECPI) believes this study is flawed as it does not appear to take into consideration a number of key factors which are widely known and accepted as being significant contributors to asthma and other respiratory diseases.

Causality

The authors state that there has been an increase in asthma in the developed world over the past 30 years, suggesting an environmental cause. This may well be true but does not necessarily mean that this increase is due to increasing exposure to man-made chemicals.

In fact, quite the opposite may be true. The prevailing theory (which is well supported by epidemiological evidence) is that the increase in immune disorders is paradoxically related to the dramatic reduction in infectious diseases, greatly diminishing the ability of our immune systems to build up necessary resistance.

It is known, for example, that the asthma rate is lower in East Germany than West Germany although the former is much more polluted than the latter.

One could therefore suggest that if phthalates have a role in asthma it is more likely as a contributor to our cleaner environment than as a prime cause.

Epidemiology

  • The principal problem with this study is that it is not really a case-control study. The authors identified children with and without asthma and then investigated "environmental" causes. One obvious selection bias that this could introduce is that the use of vinyl flooring and wall covering is widely recommended for children with asthma as a means of reducing dust. It is common practice for parents of allergic/asthmatic children to modify the home environment to minimise "dust catchers" such as rugs, carpets, quilts, cloth furniture etc, and to replace them with items that have smooth surfaces and are easy to clean. Given that these alternative materials are frequently made from soft PVC containing phthalates, it would not be surprising to find associations between allergy/asthma symptoms and phthalate levels in dust.
  • The levels of phthalates found in dust particles in this study are not abnormal. Indeed they are comparable with the levels found in several other studies.
  • There are other potential confounders that the authors did not take into consideration. With respect to rhinitis specifically, it has been known for many years that the incidence is inversely related to birth order and size of family suggesting that exposure to childhood diseases actually reduces the risk of rhinitis. This has been noticed in developed countries where there have been significant reductions in family size since the development of oral contraceptives.
  • The authors indicate within the introduction that the cases and controls were actually selected from an early cross-sectional (i. e. snapshot in time) survey. Therefore, cases had asthma/rhinitis/eczema at the start of the study, before any environmental samples were collected. This raises the real possibility that the phthalate-disease associations were a result of case status, rather than a cause.
  • Parents filled out a questionnaire on children between 1 and 6 years old. To be an asthma case they had to have had an attack during the previous year. They were then questioned 1.5 years later and had to have had the same symptoms. 6 months later they were medically examined and their bedroom dust analysed ( they were then 3 – 8 years old). According to the authors the symptoms usually start during the first year of life – so the phthalate levels in bedroom dust are being analysed 2 – 7 years after the first symptoms. Surely the dust will have changed in that time.
  • There is considerable potential for selection bias in this study, given that approximately 1100 cases and 1100 controls were asked to participate, but only approximately 200 (20%) of each did. Those who participate in studies are usually inherently different than those who do not, which is why epidemiologic studies try to maximize participation (e. g. 70% or better). The authors provide insufficient material to determine whether or not this substantial loss of study subjects would bias the study and do not even discuss this in detail. Instead, they have relegated discussion of potential selection biases to a yet-to-be published paper.
  • Although the authors provide insufficient information on possible selection biases, they do indicate that families were more likely to participate if the child had more symptoms, if there was no smoking, or if socio economic status was higher. These are important characteristics that could reasonably create bias. For example, these findings suggest that cases were more likely to participate than controls, given that children had to have at least two symptoms to be defined as a case. This suggests differential bias (i. e. differential selection between cases and controls), which is a much greater concern than random error.
  • The authors reported phthalate concentrations (i. e. grams phthalate per gram dust), but did not report or control for total amount of dust, which is a factor that might influence wheeze or rhinitis. Furthermore, the authors did not appear to gather information on allergen (e. g. house dust mite) concentrations in dust. This is a significant flaw since dust mites are widely recognised as an important contributor to asthma. Additionally, it could bias the study if dust/allergen concentrations increased as phthalate concentrations increased, or vice-versa.
  • There are many risk factors that have been associated with asthma, allergy, wheeze, or rhinitis, including environmental tobacco smoke, socio economic status, pet ownership, birth order, and allergen exposure. The authors controlled only for some of these, therefore drawing conclusions based on incomplete data.

In the July 24 issue of Science News Danish Toxicologist Gunnar Damgaard Nielsen of the National Institute of Occupational Health in Copenhagen is quoted as saying: "whether there is a causal relationship between phthalates and promotion of asthma is not clear." In its analysis, Bornehag's group didn't include factors that influence the abundance of allergens, such as a home's humidity and cleanliness. Furthermore, Nielsen notes, vinyl flooring is often used in buildings with other cheap materials, some of which may independently promote the growth of allergy-causing moulds and dust mites.

To underscore the complexity of identifying environmental triggers for allergies, Nielsen also points out that vinyl flooring is popular in part because it's easy to clean, and cleaning reduces dust and any allergy-triggering material it might carry.

The plasticiser industry contributed to the cost of this study because we were originally led to believe that it was going to be conducted in a thorough and comprehensive manner. We are very disappointed that this was not the case and that some of the individual authors are now attempting to exploit the results in such a way as to give a completely wrong impression about its significance.

A further study has subsequently been carried out by academics at Erlangen University and scientists from the German EPA (part of German environmental study on children). Urine samples from 254 children were analysed for phthalate breakdown products and their house dust analysed for phthalates at the same time. No correlation was found between level of phthalates on house dust and the intake of phthalates by children living in that house.

Please also see:

For further information please contact:

Tim Edgar
European Council for Plasticisers and Intermediates
Avenue E Van Nieuwenhuyse 4,
B-1160 Brussels, Belgium

Telephone: 0032 2 676 7363
Mobile: 0032 475 37 66 93

To return to the archive press release index click here.