IgE-Mediated Reactions and Hyperreactivity in Pregnancy Rhinitis

Objectives  To determine whether respiratory allergy or hyperreactive nasal mucosa is exceptionally common in women with pregnancy rhinitis, and to evaluate other possible risk factors such as clinical asthma or rhinitis, smoking, age, parity, and sex of the child.

Patients and Methods  From an antenatal questionnaire study, 165 women, 83 (50%) of whom had had pregnancy rhinitis, were examined 6 months after delivery, and multiple antigen simulataneous testing chemiluminescent assay (MAST CLA) (10 airborne allergens) was performed. After histamine provocations, rhinostereometry and acoustic rhinometry were performed in 25 of them. Serum levels of soluble intercellular adhesion molecule-1 were determined 4 times during and once after pregnancy in 5 women with pregnancy rhinitis and 17 without pregnancy rhinitis.

Results  Thirty-nine women (24%) were sensitized to 1 or more allergen. The pregnancy rhinitis group showed significantly higher levels of IgE to house dust mites. There were also more smokers in the pregnancy rhinitis group. Clinical asthma or rhinitis, age, parity, and sex of the child did not differ significantly between the 2 groups. Mucosal swelling increased with rising concentrations of histamine, as measured with rhinostereometry, but there was no significant difference between the 2 groups in any of the variables. Serum soluble intercellular adhesion molecule-1 was not elevated in the pregnancy rhinitis group.

Conclusions  This study found no increased frequency of allergy in general in women who have had pregnancy rhinitis. However, IgE against house dust mite was more frequent in the pregnancy rhinitis group. Smoking seems to be a risk factor, but age, parity, sex of the child, and hyperreactive nasal mucosa do not. Soluble intercellular adhesion molecule-1 was not elevated during pregnancy rhinitis.

THE FIRST study on pregnancy rhinitis was reported a century ago by MacKenzie.¹ Although it is well recognized by most clinicians as a separate clinical entity, the few scientific reports on the subject have not comprised a clinical definition. Based on a study of daily nasal congestion symptom scores of 23 pregnant women, we define pregnancy rhinitis as a nasal congestion present the last 6 or more weeks of pregnancy without other signs of respiratory tract infection and with no known allergic cause, disappearing completely within 2 weeks after delivery.² This is the basis for our further studies of its cause.

Elevated estrogen level has been the most common pathophysiologic explanation to pregnancy rhinitis. This is based mainly on biopsy studies on nasal mucosa in pregnant women,³ and from women taking contraceptive pills,⁴ and the fact that the early high-estrogen contraceptive pills produced nasal congestion as a side effect. The presence of estrogen receptors in nasal mucosa has been demonstrated.⁵ Indirect effects of estrogen, mediated by histamine for example, is an alternative explanation model.⁶ Besides estrogen, theories concerning progesterone⁷ and local vasoactive intestinal polypeptide release⁸ have been suggested.

In an earlier study, we found that serum levels of placental growth hormone were elevated in women with pregnancy rhinitis, whereas estradiol and progesterone were not.⁹ This is in conformity with our findings that objectively registered nasal blockage does not always increase in the course of pregnancy. In fact, it may even decline, even though estradiol levels increase dramatically.²

Since the electron microscopic findings of the nasal mucosa from pregnant women with nasal symptoms were identical to those in allergic rhinitis,³ we hypothesized that allergic subjects, with or without clinical symptoms, may be more prone to develop pregnancy rhinitis. Or, if allergy is not present, that nasal hyperreactivity may be a predisposing factor.

The primary objective of this study was to determine whether respiratory allergy or hyperreactive nasal mucosa was more common in women with pregnancy rhinitis than in women without this condition. Secondary objectives were to evaluate other possible risk factors for developing pregnancy rhinitis, such as clinical asthma or rhinitis, smoking, age, parity, and sex of the child.

An additional question was whether serum levels of soluble intercellular adhesion molecule-1 (sICAM-1) was elevated in women with pregnancy rhinitis, as elevated levels have been shown in perennial,¹⁰ as well as in seasonal allergic rhinitis.¹¹

All participants gave their informed consent. The study was approved by the Ethics Committee of Sahlgrenska University Hospital, Göteborg, Sweden.

The women were recruited from a questionnaire study performed during a 1-year period (September 1995 to September 1996) in 10 maternity health care centers in the southern part of County of Bohuslän (Bohuslandstinget), Sweden, where every woman who registered for a first visit for pregnancy (N=1798) was asked to participate. A total of 838 questionnaires were collected. The women were followed up with regular maternity health care checkup visits until 6 weeks after delivery. An individual analysis of the data of every woman was made regarding how she met the diagnostic criteria of pregnancy rhinitis.²

A subpopulation consisting of 83 women with pregnancy rhinitis and 82 women who clearly did not have the diagnosis was included in the present study. The demographics are shown in Table 1, top. Six months after delivery (±1 month), the women completed a standard allergy questionnaire (ALK, Kungsbacka, Sweden), serum samples were collected and stored at −70°C, and an ear, nose, and throat examination was performed (E.E.).

In vitro testing for 10 common airborne allergens (birch, timothy grass, mugwort, horse, dog, cat, Dermatophagoides pteronyssinus, Dermatophagoides farinae, Alternaria, Cladosporium) was performed using the multiple antigen simulataneous testing chemiluminescent assay (MAST CLA) system (MAST Immunosystems, Mountain View, Calif). The serum samples were analyzed in one batch. A CLA class greater than or equal to 1, corresponding to an IgE concentration of more than 0.70 IU/mL, was considered positive, as recommended by the manufacturer. In cases with a clinical history of allergy, where CLA class was 0, serum samples were also tested with the corresponding fluorescence enzyme immunoassay (CAP FEIA) (Pharmacia & Upjohn, Stockholm, Sweden).

Of the 165 women, a subgroup consisting of 12 women with pregnancy rhinitis and 13 women without was included in the study of nasal hyperreactivity. The demographics are shown in Table 1, bottom. On the same visit, before the allergy procedure, unilateral provocation was performed, at 10-minute intervals, with 0.14 mL of histamine dihydrochloride solution (phosphate buffer with 0.9% benzyl alcohol as preservative; Apoteksbolaget, Umeå, Sweden) in increasing concentrations: 1 mg/mL, 2 mg/mL, and 4 mg/mL. The diluent has been used separately in similar provocations.¹² An aspirator was connected to a 1-mL syringe, and the solution was delivered to the mucosa of the inferior turbinate, without touching it.

Rhinostereometry is an optical method used to measure differences in mucosal swelling with an accuracy of 0.2 mm.¹³ It has been used in histamine provocations using the same concentrations as in the present study,¹⁴ and a limit of 0.4-mm ipsilateral congestion 5 minutes after provocation with 2-mg/mL histamine has been proposed to differentiate hyperreactive subjects from controls.¹⁵ We used a Rhinostereometer type S (Rhinomed, Lidingö, Sweden).

After 30 minutes of acclimatization to the indoor climate, baseline measurements were established on a stable level. Further registrations were made 5 and 10 minutes after each provocation. The difference from baseline was used for analyses. All measurements were made by one of us (E.E.), unaware of women's pregnancy rhinitis status.

Acoustic rhinometry is a computerized method used to measure cross-sectional areas and volumes in different depths of the nose, based on reflected acoustic impulses.¹⁶ Hilberg et al¹⁷ have demonstrated that allergic subjects show a greater decrease in minimal cross-sectional area (MCA) on nasal histamine challenge than do nonallergic subjects. The MCA and the anterior volume (between the first and second notches) have been shown to correlate well with the corresponding measurements on computed tomography.¹⁸ We chose to use MCA, the area at the second notch ("dip 2") and the volume from the first notch to 3 cm further posterior ("volume"). Subjects were instructed to hold their breath with mouth shut to minimize influence of nasal breathing on nasal mucosa. Effort was made to avoid nasal valve distortion with the tube. We used the Rhinoklack 1000 (Stimotron, Ulm, Germany).

Baseline measurements were followed by three 5-minute postchallenge measurements, after the rhinostereometry.

As part of our previous study⁹ on 27 nonsmoking healthy women who were followed up from early pregnancy until 1 month after delivery, serum samples were collected on 4 visits during pregnancy (in gestational weeks 12, 22, 29, and 37) and 1 month after delivery, and stored at −70°C for later analysis of sICAM-1. None of the women had a history of respiratory allergy or chronic nasal or sinus problems. They were aged 22 to 38 years (mean, 28 years). Of the 22 women who completed the study, 2 groups were compared regarding serum levels of sICAM-1: 5 women with pregnancy rhinitis and 17 without the diagnosis. The sICAM-1 concentration was determined by an enzyme-linked immunosorbent assay (Serotec, Kidlington, England), with a detection limit of 0.3 ng/mL.

The Fisher permutation test was used in the analyses of CLA positivity for the different allergens, age, and parity in women with pregnancy rhinitis compared with women without pregnancy rhinitis. Fisher exact tests were used to compare the 2 groups regarding sensitization, clinical asthma or rhinitis with CLA positivity, smoking, and sex of the child as well as for the comparison of rhinostereometry, 5 minutes after 2-mg/mL histamine challenge.

For the analyses of rhinostereometry and acoustic rhinometry changes, mean regression coefficients were used in Fisher test for pair comparison regarding changes within the 2 groups, and regarding differences between the groups. The correlation between rhinostereometry and acoustic rhinometry variables was established using the Pitman test.

When comparing the mean serum values of sICAM-1 for the group of women with pregnancy rhinitis and the group without rhinitis, respectively, values were transformed to logarithms and 2-way analysis of variance for repeated measurements followed by the Student-Newman-Keuls test was used.

P<.05 was considered statistically significant.

There was an even distribution of births over the 12 months of the study. Of the 165 women, 39 (24%) were sensitized to any 1 or more of the 10 allergens tested. Overall, there was no statistically significant difference between the group who had had pregnancy rhinitis and the group who had not (29% and 18%, respectively; Fisher exact test, P=.15).

However, sensitization to D pteronyssinus and D farinae was more frequent in the former group (Fisher permutation test, P=.04), but there was no significant difference regarding any of the other allergens (Table 2).

There were no significant differences between the groups regarding CLA positivity and clinical asthma or rhinitis, age, parity, and sex of the child. Smoking was significantly more frequent in the group of women who had had pregnancy rhinitis (18% vs 6%; Fisher exact test, P=.003).

In 9 women with a clinical history of allergy, CLA was negative regarding the proposed allergens (birch, timothy grass, mugwort, horse, cat). CAP FEIA was also performed on the 12 serum samples, and was negative in all cases.

The number of cases with mucosal congestion greater than 0.4 mm, measured with rhinostereometry 5 minutes after provocation with 2-mg/mL histamine, was not significantly different between the 2 groups (Fisher exact test, P=.21).

Mucosal swelling increased significantly with rising concentrations of histamine, as measured with rhinostereometry within both groups (Figure 1), but with acoustic rhinometry an increase was observed only in the pregnancy rhinitis group, expressed as change in dip 2 (Figure 2). However, there was no significant difference between the 2 groups in any of the variables (Figure 1 and Figure 2).

When comparing rhinostereometry at 5 minutes and acoustic rhinometry variables MCA, dip 2, and volume, there was a significant correlation between dip 2 and volume only (Pitman test, R=0.79, P=.001).

The pregnancy rhinitis group had similar mean serum values of sICAM-1 as the group without the diagnosis (Figure 3). The values did not change significantly over time.

If histamine release is involved in the pathogenesis of pregnancy rhinitis, there is a possibility that subjects with subclinical allergy or subclinical nonspecific rhinitis are more susceptible to pregnancy-induced mucosal changes, and thus develop symptoms.

The overall sensitization to any 1 or more of the 10 allergens tested, with or without clinical asthma or rhinitis, was not increased in the group of women who had had pregnancy rhinitis compared with the group of women who had not. However, sensitization to D pteronyssinus and D farinae was more frequent in the former group (Table 2). Thus, it seems that the few women who have high levels of IgE against house dust mites are at greater risk to develop pregnancy rhinitis. The possibility of a false-positive result cannot be ruled out, as we tested for 10 different, although not absolutely independent, allergens. Thus, a subclinical allergic rhinitis that happened to deteriorate during pregnancy could have caused their symptoms, but they all clearly recovered after delivery, which indicates a pregnancy-related reaction. Their nasal symptoms were confined to blockage, and they did not report sneezing or discharge, which would be more likely if allergy alone was responsible. The fact that in all cases the women were sensitized to both house dust mite species clearly supports the notion that this allergen is a true contributory factor. Perennial allergens other than those presently studied may influence the nasal mucosa in pregnancy in a similar way, but that remains to be determined.

Smoking was significantly more frequent in the group of women who had had pregnancy rhinitis (18% vs 6%), but age, parity, and sex of the child were not significant factors. It is possible that smoking, as an irritant, adds to the pregnancy-induced changes in some women and thus gives nasal obstruction.

Possibly, the study population does not include enough severe cases of allergy or hyperreactivity, as their symptoms may be difficult to separate from those induced by pregnancy rhinitis. That is, in women with persistent nasal symptoms due to allergy or hyperreactivity before pregnancy, it is difficult to detect the influence of pregnancy-induced changes and thus differentiate between pregnancy rhinitis and the previous condition.

This may explain our slightly lower prevalence of 24% of in vitro–defined allergy compared with the 27% prevalence among women in our region, as reported by Plaschke et al.¹⁹ The difference between the results of radioallergosorbent test (CAP RAST) used in that study and MAST CLA is also difficult to interpret, although a thorough review shows that these tests are reasonably comparable.²⁰

A limit of 0.4-mm ipsilateral congestion, measured with rhinostereometry, 5 minutes after provocation with 2-mg/mL histamine has been proposed to differentiate hyperreactive subjects from controls.¹⁵ We compared the group of women who had had pregnancy rhinitis in this respect with those who had not. There was no difference between the groups regarding number of cases exceeding that limit. Neither did we find any difference between the groups regarding reactions to increasing concentrations of histamine, as measured by rhinostereometry or acoustic rhinometry (Figure 1 and Figure 2).

The rhinostereometric measurements demonstrate an increasing mucosal swelling with increasing concentrations of histamine in both groups (Figure 1). With the acoustic rhinometer, this was only detected by declining dip 2 in the pregnancy rhinitis group (Figure 2). This may mean that dip 2 is more sensitive to changes in mucosal swelling than is MCA, or volume. However, there was no significant correlation between rhinostereometry and any of the different acoustic rhinometry variables. The significant correlation between dip 2 and volume was expected, as the volume includes dip 2.

Serum levels of sICAM-1 have been shown to be elevated in persons with perennial allergic rhinitis.¹⁰ In individuals with seasonal allergic rhinitis, sICAM-1 was elevated in serum during a high-pollen season, and 6 weeks after.²¹ In a season with low pollen counts, elevated levels were found only during the early season.¹¹ The control groups in all these studies showed values similar to ours. We found no significant difference in serum sICAM-1 levels during pregnancy between the group of women with pregnancy rhinitis and the group without, and there was no difference in levels over time.

In conclusion, the present study does not show an increased frequency of allergy in general in women who have had pregnancy rhinitis. However, it seems that the few women who have high levels of IgE against house dust mites are more likely to develop pregnancy rhinitis. Smoking seems to be a risk factor for pregnancy rhinitis, but age, parity, sex of the child, and hyperreactive nasal mucosa do not. Soluble ICAM-1 was not elevated in patients with pregnancy rhinitis.

Accepted for publication June 8, 1999.

This study was financially supported by the County of Bohuslän (Bohuslandstinget), the Göteborg Medical Society, Göteborg, the Torsten and Ragnar Söderberg Foundation for Scientific Research, Stockholm, and the Herman Krafting Foundation for Medical Research, Trollhättan, Sweden.

This study was performed at the Department of Otorhinolaryngology, Mölndal Hospital, Mölndal, Sweden.

We thank Lars Belin, MD, PhD, for allergologic expertise, and Alvar Ellegård, PhD, for revising our English.

Reprints: Eva Ellegård, MD, Department of Otorhinolaryngology, Kungsbacka Hospital, S-43440 Kungsbacka, Sweden.