The association between asthma and perinatal mental illness: a population-based cohort study

Abstract

Background

Asthma is a risk factor for mental illness, but few studies have explored this association around the time of pregnancy. We studied the association between asthma and perinatal mental illness and explored the modifying effects of social and medical complexities.

Methods

In a population-based cohort of 846 155 women in Ontario, Canada, with a singleton live birth in 2005–2015 and no recent history of mental illness, modified Poisson regression models were constructed to examine the association between asthma diagnosed before pregnancy and perinatal mental illness, controlling for socio-demographics and medical history. We explored the modifying effects of social and medical complexities using relative excess risk due to interaction. Additional analyses examined the association between asthma and perinatal mental illness by timing and type of mental illness.

Results

Women with asthma were more likely than those without asthma to have perinatal mental illness [adjusted relative risk (aRR) 1.14; 95% (confidence interval) CI: 1.13, 1.16]. Asthma was associated with increased risk of diagnosis of mental illness prenatally (aRR 1.11; 95% CI: 1.08, 1.13) and post-partum (aRR 1.17; 95% CI: 1.15, 1.19) and specifically diagnoses of mood and anxiety disorders (aRR 1.14; 95% CI: 1.13, 1.16), psychotic disorders (aRR 1.20; 95% CI: 1.10, 1.31) and substance- or alcohol-use disorders (aRR 1.24; 95% CI: 1.14, 1.36). There was no effect modification related to social or medical complexity for these outcomes.

Conclusions

Women with asthma predating pregnancy are at slightly increased risk of mental illness in pregnancy and post-partum. A multidisciplinary management strategy may be required to ensure timely identification and treatment.

Key Messages.

• Asthma has been associated with mental illness in the general population; however, few studies have explored the association between asthma and perinatal mental illness.

• We identified increased risk of perinatal mental illness in association with asthma after controlling for socio-demographic variables, history of other chronic disease and remote history of mental illness.

• There was no evidence of effect modification by indicators of social and medical complexity.

• The results provide evidence for providing multidisciplinary care for pregnant women with asthma that attends to their mental health needs.

Introduction

Perinatal mental illness affects a significant proportion of the population, with up to 18% of women experiencing a mood or anxiety disorder,¹ 5.5% experiencing a substance-use disorder² and 0.06% experiencing a psychotic disorder³ in pregnancy or the post-partum period. Adverse effects of perinatal mental illness extend beyond the mother, impacting her offspring and family. Perinatal mental illness has been associated with an increased risk of preterm birth, child developmental and behavioural problems, and adolescent depression in the offspring.^4,5 As such, it is important to identify risk factors for perinatal mental illness to inform intervention development and implementation.

Asthma is one of the most common chronic diseases, affecting 20% of the population.⁶ Outside of pregnancy, asthma has been linked with elevated risk of mental illness, with studies from the USA, Canada and elsewhere showing higher risks of incident mood and anxiety, psychotic and substance-use disorders among adults with asthma compared with those without.^7–12 It is hypothesized that mental illness could result from chronic stress associated with managing asthma symptoms.¹³ Alternatively, the similar immune-mediated mechanisms underlying asthma and mental illness suggest a single exposure could lead to both conditions via chronic inflammation or endocrine pathways.¹³

However, only three studies have explored this association in obstetric populations. Two Canadian population-based cohort studies showed elevated risk of perinatal mental illness with chronic respiratory diseases, including asthma.^14,15 In contrast, an Australian study did not find evidence of an association between prenatal asthma and perinatal mental illness.¹⁶ Given that pregnancy and the post-partum period are times of elevated stress, further research is needed to understand this association, including the possible presence of high-risk groups who could benefit most from intervention. Social and medical complexities could exacerbate the association between asthma and perinatal mental illness. Social complexities such as poverty could act as additional sources of stress, leading to a heightened response to the effects of asthma perinatally.¹⁷ Similarly, medical complexities, such as asthma co-morbidities and other inflammation-mediated chronic diseases, could lead to greater stress or inflammation and a resulting excess risk of perinatal mental illness.¹⁸ Further research is needed to explore these factors to identify vulnerable groups and inform the mental healthcare of women with asthma in the perinatal period.

Our aims were to: (i) examine the association between asthma and perinatal mental illness and (ii) explore indicators of social and medical complexity that could exacerbate the strength of this association.

Methods

Study design and setting

This was a population-based cohort study in Ontario, Canada, that utilized health administrative data. There are no direct costs to Ontario’s 14.6 million residents for essential healthcare services. The study population included women aged 15–49 years with singleton live births conceived between 1 April 2005 and 31 March 2015 and delivered in hospital. We excluded women with a recent history of mental illness in the 2 years before conception to capture new or recurring mental illness in the perinatal period.

Data were accessed and analysed at ICES (formerly the Institute for Clinical Evaluative Sciences) in Toronto, Ontario. ICES is a publicly funded, not-for-profit research institute that links administrative health and socio-demographic data sets for primary and specialist physician visits, emergency-department visits and hospital admissions. Ontarians have a medical insurance identifier that is used to access all healthcare services; an encrypted version is used as the unique encoded identifier to link data sets. Physician-visit data are recorded using physician-billing claims; hospital databases use the Canadian Coding Standards for the International Statistical Classification of Diseases and Related Health Problems, 10th revision or Diagnostic and Statistical Manual of Mental Disorders, 4th edition. ICES data sets are highly valid, complete and reliable.¹⁹

ICES is a prescribed entity under section 45 of Ontario’s Personal Health Information Protection Act.²⁰ Section 45 authorizes ICES to collect personal health information, without consent, for the purpose of analysis or compiling statistical information with respect to the management of, evaluation or monitoring of, the allocation of resources to or planning for all or part of the health system. Projects conducted under section 45, by definition, do not require review by a research ethics board. This project was conducted under section 45 and approved by ICES’s Privacy and Legal Office.

Exposure

Women with a singleton live birth were identified in the MOMBABY data set. This data set links maternal and newborn records for hospital births (>98% of births in Ontario). Date of conception was calculated by subtracting the gestational age from the date of birth. Gestational age is generally based on dating from a first-trimester ultrasound.²¹ The exposed group comprised women with at least two physician visits or one hospitalization due to asthma in the 2 years before pregnancy, using a validated ICES algorithm for identifying asthma,²² and the unexposed group comprised women with no asthma diagnoses during this time period. Outpatient physician visits were retrieved from the Ontario Health Insurance Plan (OHIP) Database, and hospitalizations were retrieved from the Canadian Institute for Health Information Discharge Abstract Database (CIHI-DAD).

Outcome

The primary outcome was perinatal mental illness, defined by one or more physician mental-illness diagnoses, psychiatrist visits, emergency-department visits with a mental-illness diagnosis or hospitalizations with a mental-illness diagnosis during pregnancy and up to 365 days postpartum¹⁹ (Supplementary Table S1, available as Supplementary data at IJE online). Mental-illness data were sourced from OHIP and CIHI-DAD, as well as the National Ambulatory Care Reporting System for emergency-department visits and the Ontario Mental Health Reporting System for hospitalizations in psychiatric institutions. Perinatal mental illness was further categorized by timing and type. Prenatal mental illness was defined as mental illness first diagnosed during pregnancy and post-partum mental illness was defined as mental illness first diagnosed within 365 days of delivery. Types of perinatal mental illness were mood or anxiety disorders, psychotic disorders, substance-use disorders, self-harm and other mental illness; these were not mutually exclusive because women could have more than one diagnosis.

Covariates

Covariates were selected a priori based on the literature. These were mothers’ age, parity, neighbourhood income quintile, material deprivation quintile, rurality, immigrant/refugee status, asthma co-morbidities, other inflammation-mediated chronic diseases and remote history of mental illness. Neighbourhood income, material deprivation and rurality were determined by linking individuals’ postal codes with census data. Neighbourhood income was based on the average income per single-person equivalent in a dissemination area and material deprivation was derived from the Ontario Marginalization Index, based on six indicators reflecting factors such as the proportion of the population aged ≥20 years without a high-school diploma and that received income from government transfer payments.²³ Rural communities were defined by <10 000 residents. Immigrant/refugee status indicated whether women were long-term residents, non-refugee immigrants or refugee immigrants, based on the Immigration, Refugees and Citizenship Canada Permanent Residents Dataset. We captured asthma co-morbidities, defined by at least one of either chronic hypertension,²⁴ diabetes,²⁵ obesity,²⁶ rheumatoid arthritis²⁷ or sleep apnoea²⁸ in the 2 years before conception. We also captured other inflammation-mediated chronic diseases that could be associated with mental illness: cardiovascular disease,^29–32 chronic migraine,³³ chronic obstructive pulmonary disease,³⁴ inflammatory bowel disease,³⁵ systemic lupus erythematosus,³⁶ multiple sclerosis,³⁷ psoriasis³⁸ and thyroid disease³⁹ (Supplementary Table S2, available as Supplementary data at IJE online). Remote history of mental illness was defined as a mental-illness diagnosis in at least one physician visit, emergency-department visit or hospitalization between database inception (1988 for hospitalizations, 1991 for physician visits and 2000 for emergency-department visits) and 2 years before conception.

Statistical analyses

Bivariate analyses were conducted by asthma status; standardized differences were used to assess differences between groups.⁴⁰ To address the first objective, the main model regressed perinatal mental illness against asthma using modified Poisson regression with a sandwich error term.⁴¹ Unlike logistic regression, modified Poisson regression allows direct calculation of robust relative risks (RRs) in the presence of common outcomes and performs better than log binomial regression.⁴¹ Generalized estimating equations accounted for multiple pregnancies in individual mothers in the study period. We handled confounding by the covariates shown in Supplementary Figure S1 (available as Supplementary data at IJE online) in two ways. Our main approach was to use multivariable modelling to adjust for age, parity, neighbourhood income, material deprivation, rurality, immigrant/refugee status, asthma co-morbidities and other chronic diseases, with remote history of mental illness added in a second step. Additionally, to ensure the robustness of the multivariable model, we used propensity scores to match women with and without asthma on the covariates. We used nearest-neighbour greedy matching algorithms with a caliper width of 0.20 standard deviations of the logit of the propensity score⁴² and matched women with and without asthma, without replacement, at a ratio of 1:5 (Supplementary Table S3, available as Supplementary data at IJE online).

To address the second objective, we explored effect modification by neighbourhood income, material deprivation, asthma co-morbidities and other inflammation-mediated chronic diseases on the association between asthma and perinatal mental illness by calculating additive interaction using relative excess risk due to interaction (RERI).⁴³ RERI is a more meaningful measure of interaction in epidemiology than multiplicative interaction.⁴⁴ It provides insight into the effect of one exposure within the strata of a second exposure, and its presence demonstrates that the number of cases attributed to two exposures is more than the sum of the cases from each exposure separately, i.e. it provides an indication of potential vulnerable groups to target for interventions.^44,45 RERI is defined as:

$$\mathit{\backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash ,\backslash ,RERI}={RR}_{11}-\mathrm{}{RR}_{10}-\mathrm{}{RR}_{01}+1\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{}\mathrm{\backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash \; \backslash ,\backslash ,}\left(1\right)$$

where RR₁₁ is the relative risk of the outcome when both factors are present; RR₁₀ is the relative risk of the outcome when only the first factor is present; and, RR₀₁ is the relative risk of the outcome when only the second factor is present. A RERI value of >0 indicates the presence of effect modification.

We conducted several additional analyses: we repeated our regression models according to the timing and type of perinatal mental illness. We also used an outcome definition requiring two or more physician visits for perinatal mental illness to increase specificity.

Analyses were conducted in SAS 9.4.

Results

There were 858 004 women who met the inclusion criteria, of whom 1.4% were excluded because they were missing data on one or more covariates. Therefore, 846 155 women were included in this study, of whom 13.7% had asthma. As compared with those without asthma, women with asthma were more likely to be 15–24 years old, be long-term Canadian residents and have a remote history of mental illness (Table 1).

Table 1.

Baseline characteristics of women with and without asthma

Characteristic	Asthma n = 116 203 (13.73%)	Without asthma n = 729 952 (86.27%)	Standardized difference
	n (%)	n (%)
Age			0.22*
15–24	28 055 (24.14)	114 757 (15.72)
25–35	72 001 (61.96)	477 845 (65.46)
35–49	16 147 (13.90)	137 350 (18.82)
Multiparous	60 290 (51.88)	411 281 (56.34)	−0.09
Neighbourhood income quintile			0.00
Q1 (lowest income)	24 163 (20.79)	149 632 (20.50)
Q2	23 581 (20.29)	147 526 (20.21)
Q3	24 223 (20.85)	151 426 (20.74)
Q4	25 017 (21.53)	158 034 (21.65)
Q5 (highest income)	19 219 (16.54)	123 334 (16.54)
Material deprivation index			0.03
Q1 (least deprived)	21 884 (18.83)	137 849 (18.88)
Q2	22 630 (19.47)	137 502 (18.84)
Q3	22 713 (19.55)	141 269 (19.35)
Q4	22 074 (19.00)	144 839 (19.84)
Q5 (most deprived)	26 902 (23.15)	168 493 (23.08)
Rural residence	13 468 (11.59)	72 969 (10.00)	0.05
Immigrant status			0.51*
Long-term resident	105 734 (90.99)	524 620 (71.87)
Non-refugee immigrant	8507 (7.32)	178 203 (24.41)
Refugee immigrant	1962 (1.69)	27 129 (3.72)
Asthma co-morbidity	3799 (3.27)	19 279 (2.64)	0.04
Other chronic disease	10 830 (9.32)	52 195 (7.15)	0.08
Remote history of mental illness	71 379 (61.43)	334 065 (54.23)	0.32*

^*Standardized differences of >0.10 are considered clinically different.

Asthma co-morbidities: chronic hypertension, diabetes, obesity, rheumatoid arthritis or sleep apnoea.

Other inflammation-mediated chronic diseases: cardiovascular disease, chronic migraine, chronic obstructive pulmonary disease, inflammatory bowel disease, systemic lupus erythematosus, multiple sclerosis, psoriasis and thyroid disease.

About 20.1% of women with asthma, compared with 15.4% of those without asthma, had perinatal mental illness, corresponding to an adjusted relative risk (aRR) of 1.22 [95% confidence interval (CI): 1.20, 1.23] in the first adjusted model. After additional adjustment for remote history of mental illness, the aRR was 1.14 (95% CI: 1.13, 1.16) (Table 2). Findings were nearly identical when we used propensity score matching as an alternative method of controlling for confounding (RR : 1.14, 95% CI: 1.13, 1.15).

Table 2.

Unadjusted and adjusted relative risks (RRs) of perinatal mental illness in pregnancy or up to 1 year post-partum by asthma status

		Unadjusted	Adjusted Ia	Adjusted IIb
	n (%) with outcome	RR (95% CI)	RR (95% CI)	RR (95% CI)
No asthma	112 213 (15.37)	1.00 (referent)	1.00 (referent)	1.00 (referent)
Asthma	23 396 (20.13)	1.31 (1.29, 1.33)	1.22 (1.20, 1.24)	1.14 (1.13, 1.16)

^aModel adjusted for age, parity, neighbourhood income quintile, material deprivation quintile, rurality, immigrant/refugee status, asthma co-morbidities and other inflammation-mediated chronic diseases.

^bModel additionally adjusted for remote history of mental illness.

There was no effect modification by indicators of social or medical complexities [neighbourhood income quintile RERI: −0.013 (95% CI −0.049, 0.022); material deprivation quintile RERI: −0.023 (95% CI −0.056, 0.011); asthma co-morbidities RERI: 0.046 (95% CI −0.041, 0.14); other inflammation-mediated chronic diseases RERI: −0.0078 (95% CI −0.061, 0.046)] (Table 3). However, women with asthma with co-morbidities (aRR: 1.35; 95% CI: 1.28, 1.43) or other inflammation-mediated chronic diseases (aRR: 1.33; 95% CI: 1.28, 1.37) had the highest risk of perinatal mental illness—higher than women with only asthma or only other conditions (Table 3).

Table 3.

Additive interaction between asthma and the modifiers, neighbourhood income quintile, material deprivation quintile, asthma co-morbidities and other inflammation-mediated chronic disease, on the risk of perinatal mental illness in pregnancy or up to 1 year post-partum

	Perinatal mental illness
Asthma—neighbourhood income quintile
No asthma—higher income	1.00
No asthma—lower income aRR (95% CI)	1.05 (1.04, 1.07)
Asthma—higher income aRR (95% CI)	1.15 (1.13, 1.17)
Asthma—lower income aRR (95% CI)	1.19 (1.16, 1.22)
RERI (95% CI)a	−0.013 (−0.049, 0.022)
RERI p-value	0.48
Asthma—material deprivation index
No asthma—less deprived	1.00
No asthma—more deprived aRR (95% CI)	1.02 (1.00, 1.03)
Asthma—less deprived aRR (95% CI)	1.15 (1.13, 1.17)
Asthma—more deprived aRR (95% CI)	1.14 (1.11, 1.17)
RERI (95% CI)a	−0.023 (−0.056, 0.011)
RERI p-value	0.18
Asthma—asthma co-morbidities
No asthma—no co-morbidities	1.00
No asthma—co-morbidities present aRR (95% CI)	1.16 (1.13, 1.20)
Asthma—no co-morbidities aRR (95% CI)	1.14 (1.13, 1.16)
Asthma—co-morbidities present aRR (95% CI)	1.35 (1.28, 1.43)
RERI (95% CI)a	0.046 (−0.041, 0.14)
RERI p-value	0.32
Asthma—other chronic diseases
No asthma—no diseases	1.00
No asthma—diseases present aRR (95% CI)	1.19 (1.17, 1.21)
Asthma—no diseases aRR (95% CI)	1.15 (1.13, 1.16)
Asthma—diseases present aRR (95% CI)	1.33 (1.28, 1.37)
RERI (95% CI)a	−0.0078 (−0.061, 0.046)
RERI p-value	0.79

^aRERI, relative excess risk due to interaction; RERI reference value = 0.

Models adjusted for age, parity, neighbourhood income quintile, material deprivation quintile, rurality, immigrant/refugee status, asthma co-morbidities, other inflammation-mediated chronic disease and remote history of mental illness.

Lower Income: 1st quintile; higher income: 2nd–5th quintiles.

More deprived: 5th quintile; less deprived: 1st–4th quintiles.

Asthma co-morbidities: chronic hypertension, diabetes, obesity, rheumatoid arthritis or sleep apnoea.

Other inflammation-medidated chronic diseases: cardiovascular disease, chronic migraine, chronic obstructive pulmonary disease, inflammatory bowel disease, systemic lupus erythematosus, multiple sclerosis, psoriasis and thyroid disease.

About 7.5% of women with asthma had prenatal mental illness compared with 6.1% of women without asthma—an aRR of 1.11 (95% CI: 1.08, 1.13) in the final adjusted model. Among women with asthma, 15.2% had post-partum mental illness compared with 11.0% of women without asthma—an aRR of 1.17 (95% CI: 1.15, 1.19) in the final adjusted model (Table 4).

Table 4.

Unadjusted and adjusted relative risks (RRs) of perinatal mental illness among women with and without asthma by type and timing of mental illness

			Unadjusted	Adjusted Ib	Adjusted IIc
	n (%) with outcome among women without asthmaa	n (%) with outcome among women with asthmaa	RR (95% CI)	RR (95% CI)	RR (95% CI)
Timing of mental illness
Prenatal	44 529 (6.10)	8654 (7.45)	1.22 (1.19, 1.25)	1.18 (1.16, 1.21)	1.11 (1.08, 1.13)
Post-partum	80 024 (10.96)	17 616 (15.16)	1.38 (1.36, 1.40)	1.26 (1.24, 1.28)	1.17 (1.15, 1.19)
Type of mental illness
Mood or anxiety disorders	109 183 (14.96)	22 620 (19.47)	1.30 (1.28, 1.32)	1.22 (1.20, 1.23)	1.14 (1.13, 1.16)
Psychotic disorders	2581 (0.35)	607 (0.52)	1.48 (1.35, 1.62)	1.32 (1.20, 1.44)	1.20 (1.10, 1.31)
Substance-use disorders	2122 (0.29)	653 (0.56)	1.93 (1.77, 2.11)	1.35 (1.24, 1.48)	1.24 (1.14, 1.36)
Self-harm	377 (0.05)	90 (0.08)	1.49 (1.18, 1.87)	1.23 (0.97, 1.56)	1.17 (0.92, 1.50)
Other mental illness	4016 (0.55)	1116 (0.96)	1.74 (1.63, 1.86)	1.34 (1.25, 1.44)	1.24 (1.16, 1.32)

^aNumber of women with outcome across exposure groups. Numbers will not add up to the total due to overlap of women with both prenatal and post-partum mental illness and/or multiple types of mental illness.

^bModel adjusted for age, parity, neighbourhood income quintiles, material deprivation index, rurality, immigrant/refugee status, asthma co-morbidities and other inflammation-mediated chronic diseases.

^cModel additionally adjusted for remote history of mental illness.

About 19.5% of women with asthma had a mood or anxiety disorder and 0.08–1.0% had other types of mental illnesses, whereas among women without asthma, 15.0% had a mood or anxiety disorder and 0.05–0.6% had other types of mental illness, corresponding to aRRs of 1.14 (95% CI: 1.13, 1.16) for mood and anxiety disorders, 1.20 (95% CI: 1.10, 1.31) for psychotic disorders, 1.24 (95% CI: 1.14, 1.36) for substance-use disorders and 1.24 (95% CI: 1.16, 1.32) for other mental illness in final adjusted models.

Increasing the specificity of the perinatal mental-illness definition resulted in a stronger association between asthma and perinatal mental illness in the final adjusted model (aRR: 1.30; 95% CI: 1.28, 1.33).

Discussion

This population-based study found an increased risk of perinatal mental illness among women with asthma compared with those without asthma. This relationship, though somewhat attenuated, remained after controlling for remote history of mental illness and other important covariates. The risk was consistently elevated when examining perinatal mental illness by timing and type. There was no evidence of effect modification by indicators of social or medical complexity in this population. These results support the need for multidisciplinary care for asthma and mental health management during the perinatal period.

Our results are in line with previous studies.^14–16 Two large Canadian population-based cohort studies found increased risk of perinatal mental illness with asthma¹⁵ and chronic respiratory diseases.¹⁴ A smaller Australian study of 120 women found no association between prenatal asthma and perinatal depression and psychological well-being, but did report a moderate association between poor post-partum asthma control and psychological well-being.¹⁶ The results of the Australian study are not directly comparable to ours since we defined asthma status before pregnancy and asthma symptoms may change during pregnancy.⁴⁶ We did not identify any studies that examined the association between asthma and other perinatal mental disorders, nor did we identify studies that examined the association between asthma and prenatal vs post-partum mental illness. Notably, we identified a 20% and 24% increased risk of psychotic and substance-use disorders, respectively, among women with asthma. These results point to the importance of expanding future studies to include other types of perinatal mental illness. The results, however, align with studies in the general population that have reported elevated risk of psychosis¹² and substance-use disorders⁹ in adults with asthma.

To our knowledge, this is the first study to have examined the modifying effects of indicators of social and medical complexities on the association between asthma and perinatal mental illness. We did not observe evidence of effect modification. This could be due to the lack of information on individual socio-economic status and under-reporting of certain chronic conditions in health administrative data, such as obesity and sleep apnoea, that may mask population heterogeneity. However, notably, despite the lack of effect modification, women with asthma and additional asthma co-morbidities or other inflammation-mediated chronic diseases had the highest risks of perinatal mental illness of all subgroups in the effect modification analyses, suggesting that these women could benefit most from intervention.

Whereas little research has examined the association between asthma and perinatal mental illness, several mechanisms of action have been proposed to explain the established link between asthma and mental illness in the general population. One mechanism hypothesizes mental health consequences of asthma resulting from stress due to having a chronic disease.¹³ Evidence also supports a bidirectional association between asthma and mental illness, suggesting an additional biological pathway. A common exposure could lead to both conditions.¹³ For example, there appears to be shared genetic susceptibility between asthma and psychosis or depression.^47,48 Studies have reported an inflammation-mediated response driven by cytokines IL-1, IL-2, IL-6 and tumour necrosis factor that could link asthma and mental illness in a cyclical manner.^49–51 Another theory proposes deregulation of the hypothalamus–pituitary–adrenal axis through stress and subsequent deregulation of cortisol could explain the link between asthma and anxiety.^49,52 Importantly, the association between asthma and perinatal mental illness remained in our study after excluding women with a recent history of mental illness and controlling for remote history of mental illness, suggesting particular vulnerability to mental illness among women with asthma in the perinatal period. This is an important finding, with implications specifically for perinatal care.

Limitations

Although our exposure and outcome were identified using verified operational definitions based on diagnostic codes, some misclassification bias may exist. Although we observed equivalent results using multivariable adjustment and propensity score matching, unmeasured confounding could also impact the results. We were unable to control for some variables not normally measured in administrative health data. We had no information on smoking. We were also unable to account for body mass index and had to rely on a diagnostic code for obesity, which is underreported in administrative data.^26,28 Maternal education and individual-level income were also not measured. We attempted to control for socio-economic status by measuring neighbourhood-level income and material deprivation based on a linkage of the individual’s postal code with census data.

Clinical implications

Our results provide evidence for tailoring multidisciplinary approaches for asthma management to pregnancy. A focus on collaboration between primary care physicians, respirologists and psychiatrists is recommended, with the ultimate goal of preventing perinatal mental illness among women with asthma. Collaborative care is a team-based approach to care and a review recommends proactive consultation with psychiatrists under this model to prevent and/or treat mental disorders in patients with chronic respiratory conditions.⁵³ Also recommended is the involvement of a care manager to identify and address barriers to care and to coordinate multidisciplinary strategies. The inclusion of care managers may be particularly important when working with vulnerable groups, such as those of lower socio-economic status,⁵⁴ or those with complex medical profiles, such as women with asthma and other co-morbidities, who may require additional support to benefit from mental healthcare.

Conclusion

Our population-based study shows a positive association between asthma and perinatal mental illness. Women with asthma were more likely to have mood and anxiety, psychotic and substance-use disorders in both the prenatal and post-partum periods compared with women without asthma. Our results speak to the importance of tailoring multidisciplinary approaches for the perinatal healthcare of women with asthma.

Supplementary data

Supplementary data are available at IJE online.

Ethics approval

Section 45 authorizes ICES to collect personal health information, without consent, for the purpose of analysis or compiling statistical information with respect to the management of, evaluation or monitoring of, the allocation of resources to or planning for all or part of the health system. Projects conducted under section 45, by definition, do not require review by a research ethics board. This project was conducted under section 45 and approved by ICES’s Privacy and Legal Office.

Funding

This study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health (MOH) and the Ministry of Long-Term Care (MLTC). This work was also supported by a grant from the Canadian Institutes of Health Research [Grant #376290 to H.K.B.]. This research was undertaken, in part, thanks to funding from the Canada Research Chairs Program to H.K.B. The analyses, conclusions, opinions and statements expressed herein are solely those of the authors and do not reflect those of the funding; no endorsement is intended or should be inferred.

Data availability

The data set from this study is held securely in coded form at ICES. Although data-sharing agreements prohibit ICES from making the data set publicly available, access may be granted to those who meet pre-specified criteria for confidential access, available at www.ices.on.ca/DAS. The full data-set creation plan and underlying analytic code are available from the authors upon request, understanding that the computer programs may rely upon coding templates or macros that are unique to ICES and are therefore either inaccessible or may require modification.

Disclaimer

Parts of this material are based on data and information compiled and provided by the Canadian Institute for Health Information (CIHI) and Immigration, Refugees and Citizenship Canada (IRCC). However, the analyses, conclusions, opinions and statements expressed herein are those of the author and not necessarily those of CIHI or IRCC.

Conflict of interest

Dr Vigod receives royalties for authorship of materials related to depression and pregnancy from UpToDate Inc. All other authors have no conflicts of interest.