Pollen in pregnancy could be linked to infant asthma...

'Pregnant women exposed to pollen shortly before birth are more likely to have children with severe asthma,' is the somewhat alarmist headline in the Daily Mail.

The news is based on a large Swedish study that aimed to investigate the relationship between exposure to pollen during pregnancy and the first few months of a child's life, and the risk of the child requiring hospitalisation for asthma in the first year of life.

The researchers found that high pollen exposure during the last trimester of a mother's pregnancy was associated with an increased risk of the child being admitted to hospital for asthma. While a statistically significant link between pollen exposure and hospitalisation for asthma was detected, the number of babies who required this was small: just 940 out of 110,381 (0.855%).

They also found that high pollen exposure during a child's first three months of life was associated with a reduced risk of an asthma-related hospital admission, but only in the children of mothers who smoked heavily.

While the evidence provided suggests that both of these associations are valid, this study provides little hard information on why either of these were detected, only theories.

The study design cannot prove that there is direct cause and effect at play (causal), as there may be other factors at work (confounders), such as the effect weather has on a mother's smoking habit.

While this is an impressive piece of research that must have involved a great deal of time and effort, somewhat frustratingly it raises more questions than it answers. The researchers have announced that they are planning further investigation into the issues raised by their study.

Where did the story come from?

The study was carried out by researchers from the Occupational and Environmental Medicine Department of the University of Umeå in Sweden, and other institutions. It was funded by the Swedish Research Council, Umeå University and individual research grants.

The study was published in the peer-reviewed journal Allergy, Asthma and Clinical Immunology.

Despite the scaremongering headline that suggests "severe asthma" in individual children is directly the result of pollen exposure, the Daily Mail did show caution in the main body of the story and quoted the authors' calls for more research.

What kind of research was this?

This was an ecological study investigating the relationship between seasonal exposure to pollen in the population and the seasonal rates of hospitalisation for asthma in infants under 12 months old.

Ecological studies can be useful for looking at possible associations between various lifestyle factors (such as exposure to pollen during pregnancy) and health outcomes (such as a child requiring hospitalisation due to asthma) at a population level.

They rarely allow researchers to show cause and effect, especially if both exposure and outcome are seasonal. They can usually only highlight possible associations, which then need to be followed up by further research.

The researchers explain in the introduction to the study that there is a great deal of uncertainty as to whether exposure to allergy-causing substances (allergens) such as pollen during pregnancy and early life either increases or decreases the risk of a child developing an allergic condition such as asthma.

What did the research involve?

The researchers included all vaginally-delivered babies in the greater Stockholm area from 1989 to 1996 (110,381 babies) using data obtained from the Swedish Medical Birth Registry and Inpatient Registry. Information collected from the registry included:

the child's date of birth
parity (the number of times a woman has given birth)
birth weight and length
gestational age
the baby's gender
smoking habits of the mother during pregnancy

Based on the date of the child's estimated date of conception and birth, the researchers looked up the average pollen levels in the city for three periods:

during the first 12 weeks of the mother's pregnancy (first trimester)
during the last 12 weeks of the mother's pregnancy (third trimester)
during the first 12 weeks of the child's life

Daily pollen levels for the time periods were obtained from the Swedish Museum of Natural History (measured at a single location in central Stockholm), while pollution levels were obtained from the City of Stockholm Environment and Health Administration.

The main outcome the researchers were interested in was any hospital admissions for asthma during the child's first year of life, using information obtained from the Swedish Inpatient Registry between 1989 and 1997.

Details on the number of admissions for lower respiratory tract illness were also used as an indication of the amount of exposure to pollens in the first three and six months of life.

The researchers analysed their results using three different statistical methods and adjusted the results for infant gender, gestational age, maternal smoking and season of birth.

They first looked at increases of pollen levels dependant on the season of birth for each of the three measured periods.

They then adjusted their results for pollution levels, and also adjusted for rates of hospitalisation in the first three months of the child's life and in the period from three to six months of life.

The researchers also assessed whether the effect of pollen exposure varied between the years the study took place by comparing the highest 25% of exposed children with the remaining children for the same year.

What were the basic results?

Of the 110,381 children, 940 (0.85%) were hospitalised due to asthma in their first year of life.

The main findings of this study were:

there was high variation in the daily pollen levels of the years studied (for instance, a peak in 1993 and lows in 1998 and 1994)
children born between February and July had the lowest risk of hospital admission for asthma, while children born between October and November had the highest risk. The researchers say this "season of birth effect" was present for some years but not others

When looking at the highest 25% of children exposed versus the remaining children:

exposure to high pollen levels in the last 12 weeks of a mother's pregnancy was associated with an increased risk of a child being admitted to hospital for asthma (adjusted odds ratio (aOR) 1.35, 95% confidence interval (CI) 1.07 to 1.71)
exposure to high pollen levels in the first three months of a child's life was associated with a reduced risk of admission to hospital due to asthma among children whose mothers were identified as heavy smokers (reported as smoking more than 10 cigarettes per day) (aOR 0.52, 95% CI 0.33 to 0.82)
there was no association between high pollen exposure in the first three months of a child's life and hospital admission for asthma in children of non-smoking mothers (aOR 0.96, 95% CI 0.75 to 1.24)

How did the researchers interpret the results?

The researchers conclude that high levels of pollen exposure during late pregnancy were unexpectedly associated with a higher risk of hospitalisation for asthma within the first year of life.

In addition, children exposed to high levels of pollen in infancy had a reduced risk, but only in the children of mothers who were heavy smokers.

The researchers provide a number of possible reasons for the association, including:

exposure to pollen during late pregnancy may make a baby's immune system oversensitive to the effects of pollen, making them more likely to develop a more severe form of asthma
there could be other seasonal factors responsible other than pollen exposure, such as vitamin D exposure
in terms of smoking, pollen is higher during the spring and summer months so smoking mothers may have been more likely to be outdoors, meaning their babies were exposed to less passive smoking

It should be stressed that all of the above are just unproven theories.

Conclusion

This large study provides some evidence of an association between exposure to pollen during pregnancy and the risk of a child being admitted to hospital for asthma. Importantly, it does not provide evidence that exposure to pollens during pregnancy leads to childhood asthma.

Despite the author's efforts to adjust their results for confounders, it is always possible that other factors, such as a family history of asthma, influenced the results. Additional limitations include:

It is likely that a number of the children admitted for "asthma" did not truly have asthma (misclassification) and in fact had another form of respiratory illness. This would have overestimated the rate of admissions, and if a large proportion were infectious cases the winter seasonal variation may have led to spurious results.
Conversely, children that did have asthma may not have had severe enough disease to require a hospital admission and instead were seen in an outpatient setting, such as a GP practice.
Hospital admissions for children was only measured in the first year of life. Admissions for asthma at older ages, when diagnosis of asthma is more reliable, would have provided more informative results.
The main limitation, though, is that pollen exposure and hospital admissions for asthma was not measured for individuals. These were taken from population statistics (city pollen counts and hospital admission rates), making this an ecological study, which cannot prove a cause and effect relationship.
As both rates of admission for chest problems in children and pollen counts are seasonal, much more research is needed into immune mechanisms if pollen is to be blamed for seasonal respiratory outcomes in children.

Hopefully, the upcoming research announced in the paper will go some way to addressing some of these limitations, and provide more useful information about the possible causes and risk factors for childhood asthma.

Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.