Depression, anxiety, and post-traumatic stress disorder in pregnancies among Métis people in Alberta: a population-based cross-sectional study

Abstract

Background

Despite high stress from historical trauma, systemic discrimination, and limited access to culturally-safe healthcare, the perinatal mental health of the Métis, Canada’s second-largest Indigenous group, remains poorly understood. We evaluated the prevalence and sociodemographic and clinical factors associated with depression, anxiety, and post-traumatic stress disorder (PTSD) in Métis pregnancies.

Methods

This cross-sectional study used population-based retrospective cohort data of all pregnancies resulting in live births (≥ 22 weeks’ gestation) in Alberta, Canada, from 2006 to 2016. The Métis cohort was identified through linkage between the Identification Registry of the Otipemisiwak Métis Government of the Métis Nation within Alberta and administrative health databases. Diagnoses of depression, anxiety, and PTSD were identified in administrative health databases using validated case-finding algorithms. Age-standardized prevalence of depression, anxiety, and PTSD were compared between Métis and non-Métis pregnancies. Multilevel, multivariable Poisson regression models examined sociodemographic and clinical factors associated with these conditions in Métis pregnancies. Adjusted prevalence ratios (aPR) with 95% confidence intervals (CI) were reported, adjusting for relevant sociodemographic and clinical characteristics.

Results

The study analyzed 8,163 Métis and 487,294 non-Métis pregnancies. The age-adjusted prevalence of depression in Métis pregnancies was 3.3% (95% CI 2.7, 3.9), while the prevalence of anxiety was 3.5% (95% CI 2.9, 4.2). Depression was 1.44 times (95% CI 1.25, 1.64) and anxiety 1.31 times (95%CI 1.15, 1.49) more prevalent in Métis than non-Métis pregnancies. PTSD prevalence was low in both groups. Among Métis pregnancies, urban residence, pre-existing medical conditions, and smoke or substance use in pregnancy were associated with higher depression and anxiety prevalence.

Conclusions

Depression and anxiety during pregnancy were more common in Métis than non-Métis pregnancies. Collaboration with Métis organizations is critical to developing culturally-relevant detection and treatment for mental health challenges.

Supplementary Information

The online version contains supplementary material available at 10.1186/s44263-026-00259-w.

Background

Mental health problems during pregnancy including depression, anxiety, and post-traumatic stress disorder (PTSD) are major public health concerns. Globally, depression and anxiety affect up to 7.1% and 8.5% of pregnant individuals, respectively [1, 2]. In Canada, one large prospective cohort study reported depressive symptoms in 9.7% of pregnant individuals and anxiety symptoms in 15.4% [3], while a population-based administrative data study from a single province reported a combined rate of 13.5 for any mood or anxiety disorder during pregnancy [4]. PTSD is less frequently diagnosed in community-based samples (3.3%) [5], but can be higher in those facing socioeconomic disadvantage, trauma, intimate partner violence, and/or pregnancy complications [5]. Untreated, these exposures can also increase rates of obstetric complications, preterm birth, low birth weight, and developmental challenges [6].

Mental health prevalence varies across racial and ethnic groups among women, with synthesis evidence documenting heterogeneity in depression and anxiety estimates across groups and settings [7]. In the perinatal period, systematic reviews similarly report racial/ethnic differences in the prevalence and correlates of antenatal mental health conditions [8, 9]. This broader evidence base provides context for understanding perinatal mental health among Indigenous populations.

Evidence indicates that Indigenous populations face a disproportionate mental health burden compared to their non-Indigenous counterparts [10, 11]. They are also more likely to experience adverse social and structural determinants of health, such as poverty, substance use, and limited access to medical services, which are linked to perinatal depression and anxiety [10]. Indeed, Indigenous individuals in the CANZUS countries (Canada, Australia, Aotearoa/New Zealand, and the United States) are 62% more likely to experience perinatal mental health challenges than non-Indigenous populations [11]. These challenges and their consequences may be compounded by a higher prevalence of adversities linked to social and structural disadvantage, which can increase symptom severity and limit access to timely, culturally safe supports, with potential short- and long-term, including intergenerational, impacts for children and families [10–12]. However, Indigenous populations are heterogeneous, and distinctions-based, group-specific evidence on mental health during pregnancy is needed.

In Canada, Métis people, one of three Indigenous groups recognized in Canada’s Constitution Act, have a distinct heritage, language, and culture. Métis are the second-largest Indigenous group in Canada (624,220 people in 2021), compared with 1,048,405 First Nations people and 70,545 Inuit [13]. Despite resilience, Métis individuals continue to face colonial legacies, socioeconomic disadvantage, intergenerational trauma, and systemic discrimination, all of which may influence perinatal mental health. Exclusion from historic treaties, land dispossession, ineligibility for Non-Insured Health Benefits, and racism in healthcare also contribute to distinct health patterns [14].

Perinatal mental health determinants in Indigenous populations in Canada have been examined primarily through Indigenous versus non-Indigenous comparisons, with fewer studies reporting distinction-based findings (e.g., First Nations and Inuit) [11]. In contrast, research on Métis individuals is limited, a gap that reflects, in part, differences in how Indigenous identity is captured in data sources and eligibility for programs that can constrain Métis-specific maternal health analyses. Findings from the Canadian Maternity Experiences Survey [15] suggest Métis individuals have higher rates of pre-pregnancy depression [16], and postpartum depressive symptoms than non-Indigenous individuals [17]. However, these results rely on self-reporting, introducing potential bias and limiting generalizability. Additionally, key risk factors such as limited social or partner support, history of abuse, chronic health conditions, pregnancy complications, previous mental illness, unplanned pregnancies, high perceived stress, and adverse life events [18, 19] remain largely unexamined in Métis pregnancies.

Despite growing recognition of the importance of perinatal mental health, critical gaps remain in understanding Métis experiences. The lack of Métis-specific epidemiological data hinders efforts to assess mental health burden and develop equitable, evidence-based healthcare policies. This study aimed to: 1) estimate the prevalence of depression, anxiety, and PTSD among Métis pregnancies in Alberta and compare these to non-Métis pregnancies, and 2) identify sociodemographic and clinical factors associated with these diagnoses in Métis pregnancies.

Methods

Study design and population

This cross-sectional analysis used linked data from a population-based retrospective cohort of singleton and multifetal pregnancies resulting in live births (≥ 22 weeks’ gestation) in Alberta, Canada, between April 1, 2006, and March 31, 2016. Alberta has Canada’s second-largest Métis population (~ 127,000 individuals) [12] and the highest proportion of Métis women [13]. The study was conducted in collaboration with the Otipemisiwak Métis Government of the Métis Nation within Alberta (MNA), which contributed to study design, data access, and interpretation. A formal research agreement guided this partnership, and the team included Métis co-authors and academic researchers. The study adhered to the six principles of ethical Métis research (reciprocal relationships, respect, safe and inclusive environments, diversity, Métis relevance and context considerations) [20]. Métis perspectives informed study interpretation and reporting. The study followed the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) guidelines [21] (Supplementary Material 1: STROBE Checklist).

Data sources

Access to de-identified data was facilitated through Alberta Health (the provincial ministry and data custodian for provincial health administrative data), Alberta Health Services (the provincial health authority and data custodian for relevant clinical/health system data), and the MNA under an information-sharing agreement. The Alberta Perinatal Health Program registry provided maternal and newborn clinical data. The Alberta Perinatal Health Program registry captures maternal and newborn clinical and demographic data for all births in the province, including those occurring in hospitals or attended by registered midwives at home. The MNA Identification Registry documents Métis citizenship for approximately 46,000 individuals in the province [22]. The Discharge Abstract Database [23] and the Alberta Physician Claims Assessment [24] databases provided hospitalization and physician claims data using International Classification of Diseases, Ninth (ICD-9) [25], and Tenth Revision enhanced Canadian version (ICD-10-CA) [26] diagnostic codes.

Definition of the study cohorts

Métis pregnancies were identified through probabilistic linkage between the MNA Identification Registry and the Alberta Health Care Insurance Plan based on full name, sex and date of birth (96% accuracy). Once Métis status (yes/no) was assigned in the Alberta Health Care Insurance Plan, records were deterministically linked to the Alberta Perinatal Health Program registry via personal health numbers to classify pregnancies as Métis or non-Métis. Non-Métis pregnancies served as the reference group for comparing the prevalence of depression, anxiety, and PTSD in the Métis cohort.

Depression, anxiety, and PTSD definitions

Depression, anxiety, and PTSD were identified using validated administrative case definitions developed and evaluated in prior work [27–29]. We flagged relevant ICD-9/ICD-10-CA codes in the Discharge Abstract Database and the Alberta Physician Claims Assessment database and applied the validated algorithm requiring one hospitalization or two physician claims within 10 months before delivery, which improves specificity and reduces false positives from rule-out diagnoses, coding error, or isolated single claims [27–29]. For depression ICD-9 codes included major depressive disorder, single episode (296.2), major depressive disorder, recurrent episode (296.3), bipolar disorder with depressive episode (296.5), dysthymia (300.4), adjustment disorders (309.x, excluding 309.81 for PTSD), and depressive disorder not otherwise specified (311). ICD-10-CA codes included major depressive disorder, single episode (F32.x), major depressive disorder, recurrent (F33.x), dysthymia (F34.1), mixed anxiety and depressive disorder (F41.2), adjustment disorder with depressed mood (F43.2), schizoaffective disorder, depressive type (F20.4), and bipolar disorder with depressive episodes (F31.3–F31.5). In the validation study, this case-finding algorithm for depression showed sensitivity 61.4%, specificity 94.3%, positive predictive value 69.7%, and negative predictive value 92% [27].

Anxiety was identified using a validated administrative case definition [28]. ICD-9 codes included generalized anxiety disorder (300.0) and other anxiety states (300.2). Corresponding ICD-10-CA codes included phobic anxiety disorders (F40.x) and other anxiety disorders (F41.x, excluding mixed anxiety and depressive disorder [F41.2]) [28]. In the validation study, this case-finding algorithm for anxiety had sensitivity 30.1%, specificity 88%, positive predictive value 53.2%, and negative predictive value 73.5% [28], reflecting the sensitivity–specificity trade-off of administrative case definitions for conditions often managed in outpatient settings [30, 31]. PTSD was identified using a validated administrative case definition [29] including ICD-9 code for post-traumatic stress disorder (309.81) and ICD-10-CA code for reaction to severe stress and adjustment disorders, specifically post-traumatic stress disorder (F43.1x) [29]. In the validation study, this case-finding algorithm for PTSD had sensitivity 97%, specificity 76%, positive predictive value 47%, and negative predictive value 99% [29].

Sociodemographic and clinical characteristics

Sociodemographic and clinical characteristics were selected based on their theoretical relevance to perinatal mental health disorders during pregnancy [18, 19]. They included maternal age (categorized as < 20, 20–34, ≥ 35 years based on perinatal risk profiles) [32], place of residence (urban, rural based on population density and urban centre proximity), parity (nulliparous, or 1, 2, ≥ 3 prior live births), pregnancy type (single, multifetal), high pre-pregnancy weight (> 91 kg; as defined in the perinatal registry), presence (yes/no) of pre-existing medical conditions (i.e., diabetes mellitus, heart disease, hypertension, renal disease, severe asthma, lupus, epilepsy, or Crohn’s disease), pregnancy-related medical conditions (i.e., gestational hypertension, gestational diabetes, preeclampsia, and anemia), smoking (registry-coded as anytime during pregnancy), alcohol or substance use during pregnancy (registry-coded as ≥ 3 drinks on any occasion or ≥ 1 drink per day throughout pregnancy; substance use coded yes/no), and prenatal care adequacy (no care, inadequate, intensive, intermediate, adequate), categorized using the Revised-Graduate Prenatal Care Utilization Index based on timing of initiation and number of prenatal visits, accounting for gestational age [33, 34]. Socioeconomic status was assessed using the Pampalon Material and Social Deprivation Index [35], a census-derived, small-area measure that is widely used in Canadian health research as a proxy for individual circumstances when individual-level socioeconomic status (SES) variables are not available in administrative health data. The index classifies residential areas into quintiles from the least (Q₁) to the most deprived (Q₅) and captures material (income, education, and employment) and social deprivation (living alone, not married, and single-parent families). Maternal socioeconomic status was assigned using postal codes at delivery to link individuals to their residential material and social deprivation quintiles.

Statistical analysis

Sociodemographic and clinical characteristics were summarized using descriptive statistics. Missing values were not imputed. Data for variables with 20 or fewer observations were suppressed per privacy guidelines [36]. The distribution of maternal and neonatal characteristics between Métis and non-Métis children were compared using Chi-square tests, with statistical significance set at p < 0.05.

The period prevalence of depression, anxiety, and PTSD for Métis and non-Métis pregnancies was calculated using the total number of cases as the numerator and the total number of pregnancies as the denominator, and standardized by age using the 2010 Canadian age distribution for females as the reference population.

To compare prevalence between Métis and non-Métis pregnancies, multilevel, multivariable Poisson regression models with robust standard errors were used to estimate crude and adjusted prevalence ratios (PR) with 95% confidence intervals (CI). Adjusted models included maternal age, multifetal pregnancy, any pre-existing medical condition (at least one of heart disease, renal disease, diabetes, hypertension, severe asthma, lupus, epilepsy, and/or Crohn’s disease), any pregnancy-related medical condition (at least one of gestational hypertension, gestational diabetes, preeclampsia, and/or anemia), smoking or alcohol and/or substance use during pregnancy (combined), place of residence, and socioeconomic deprivation. Correlated covariates (i.e., pre-existing and pregnancy-related medical conditions and substance use during pregnancy) were combined to address multicollinearity. Adjusted prevalence ratios (aPR) with 95% CI were reported. The multilevel structure clustered pregnancies by individuals to account for the dependence created by individuals with more than one pregnancy during the study period. In addition, within the Métis cohort only, separate multilevel Poisson regression models were fit for each outcome (depression, anxiety, and PTSD) to examine associations with sociodemographic and clinical factors. These Métis-only models were mutually adjusted for all covariates listed above, and results are presented as aPRs with 95% CIs. All statistical analyses were conducted using STATA, Release 15 (StataCorp. College Station; TX).

Results

Of 495,457 live births in Alberta (2006–2016), 8,163 (1.6%) were to Métis pregnancies and 487,294 (98.4%) to non-Métis pregnancies (Fig. 1).

Fig. 1.

Flow diagram of Métis and non-Métis pregnancies

Table 1 summarizes the clinical and sociodemographic characteristics of Métis and non-Métis pregnancies. Most Métis pregnancies occurred at 20–34 years-old (81.7%) and involved primigravidas (42.2%). Over one-third of Métis pregnancies occurred in rural areas, while nearly half were in the two most deprived areas. Pre-existing medical conditions such as diabetes, heart disease, hypertension, and renal disease were rare (< 1.3%). Gestational hypertension and gestational diabetes occurred in 5.4% and 5.2% of Métis pregnancies, respectively, while preeclampsia and anemia were present in 1.4% and 0.8%. Most Métis pregnancies were smoke-free (68.8%) and had no reported alcohol or substance use (93.1%).

Table 1.

Sociodemographic and clinical characteristics of Métis and non-Métis pregnancies in Alberta (2006–2016)

Characteristic	Métis (N = 8,163) % (n)	Non-Métis (N = 487,294) % (n)	p-value1
Age group			< 0.001
< 20	8.5 (690)	3.4 (16,512)
20–34	81.7 (6,668)	78.2 (381,327)
≥ 35	9.9 (805)	18.4 (89,448)
Missing	0 (0)	0.001 (7)
Pre-pregnancy weight > 91 kg			< 0.001
Yes	14.3 (1,168)	9.2 (44,727)
No	84.8 (6,923)	90.2 (439,297)
Missing	0.9 (72)	0.7 (3,270)
Pre-existing diabetes			0.111
Yes	1.3 (109)	1.2 (5,596)
No	97.8 (7,982)	98.2 (478,431)
Missing	0.9 (72)	0.7 (3,267)
Pre-existing heart disease			0.776
Yes	0.1 (< 20)2	0.1 (340)
No	99.1 (8,086)	99.3 (483,687)
Missing	0.9 (72)	0.7 (3,267)
Pre-existing hypertension			0.014
Yes	1.2 (99)	1.0 (4,623)
No	97.9 (7,992)	98.4 (479,404)
Missing	0.9 (72)	0.7 (3,267)
Pre-existing renal disease			0.210
Yes	0.1 (< 20)2	0.1 (450)
No	98.9 (8,080)	99.2 (483,577)
Missing	0.9 (72)	0.7 (3,267)
Other pre-existing medical conditions3			0.142
Yes	8.0 (650)	7.5 (36,772)
No	91.2 (7,441)	91.8 (447,252)
Missing	0.9 (72)	0.7 (3,270)
Pregnancy type			0.523
Single	96.7 (7,897)	96.6 (470,787)
Multifetal	3.3 (266)	3.4 (16,507)
Parity			< 0.001
0 (Nulliparous)	42.2 (3,448)	42.5 (207,304)
1	31.1 (2,538)	34.3 (167,143)
2	15.6 (1,277)	14.0 (68,271)
≥ 3	10.5 (860)	8.8 (42,729)
Missing	0.5 (40)	0.4 (1,847)
Gestational hypertension			0.781
Yes	5.4 (437)	5.4 (26,486)
No	93.8 (7,147)	93.9 (457,541)
Missing	0.9 (72)	0.7 (3,267)
Gestational diabetes			0.039
Yes	5.2 (427)	5.8 (28,159)
No	93.9 (7,664)	93.6 (455,868)
Missing	0.9 (72)	0.7 (3,267)
Preeclampsia			0.157
Yes	1.4 (116)	1.3 (6,088)
No	96.9 (7,907)	97.3 (474,258)
Missing	1.7 (140)	1.4 (6,948)
Anemia during pregnancy			0.327
Yes	0.8 (64)	0.9 (4,329)
No	98.3 (8,027)	98.4 (479,698)
Missing	0.9 (72)	0.7 (3,267)
Prenatal care adequacy			< 0.001
Adequate	36.1 (2,943)	38.2 (185,943)
Intermediate	41.2 (3,364)	37.8 (184,416)
Intensive	3.8 (309)	3.8 (18,384)
Inadequate	16.2 (1,323)	16.5 (80,602)
No care	2.7 (224)	3.7 (17,949)
Smoking in pregnancy			< 0.001
Yes	30.4 (2,479)	14.3 (69,651)
No	68.8 (5,612)	85.0 (414,376)
Missing	0.9 (72)	0.7 (3,267)
Alcohol or substance use in pregnancy			< 0.001
Yes	6.0 (492)	3.0 (14,387)
No	93.1 (7,599)	96.4 (469,639)
Missing	0.9 (72)	0.7 (3,268)
Place of residence			< 0.001
Rural	36.9 (3,014)	23.7 (115,471)
Urban	62.0 (5,061)	75.4 (367,394)
Missing	1.1 (88)	0.9 (4,429)
Material Index Quintile			< 0.001
Q1-least deprived	10.0 (819)	18.6 (90,665)
Q2	16.7 (1,360)	18.9 (92,160)
Q3	18.9 (1,544)	18.8 (91,795)
Q4	22.5 (1,837)	18.4 (89,719)
Q5-most deprived	26.4 (2,156)	20.0 (97,523)
Missing	5.5 (447)	5.2 (25,432)
Social Index Quintile			< 0.001
Q1-least deprived	12.2 (994)	13.0 (63,384)
Q2	14.6 (1,189)	18.9 (92,103)
Q3	20.3 (1,660)	21.8 (105,992)
Q4	27.2 (2,218)	21.9 (106,929)
Q5-most deprived	20.3 (1,655)	19.2 (93,454)
Missing	5.5 (447)	5.2 (25,432)

¹p-values based on Chi-square test excluding the missing data category

²Exact number for variables with fewer than 20 observations is suppressed

³Other pre-existing medical conditions include severe asthma, lupus, epilepsy, or Crohn’s disease

Depression

Depression was diagnosed in 265 Métis pregnancies and 9,968 non-Métis pregnancies. The age-standardized prevalence of depression in Métis pregnancies was 3.3% (95% CI 2.7%, 3.9%) versus 2.2% (95% CI 2.1%, 2.3%) in non-Métis pregnancies. After adjusting for covariates, depression was 1.44 times more prevalent in Métis pregnancies compared to non-Métis pregnancies (aPR 1.44, 95% CI 1.25, 1.64) (Table 2).

Table 2.

Crude and age-specific standardized period prevalence of mental health outcomes for Métis and non-Métis pregnancies in Alberta (2006–2016)

Outcome		Métis N = 8,163	Non-Métis N = 487,294
Depression	Number of cases	265	9,968
	Prevalence % (95%CI)	3.3 (2.9, 3.6)	2.1 (2.0, 2.1)
	Age-adjusted prevalence % (95%CI)	3.3 (2.7, 3.9)	2.2 (2.1, 2.3)
	Crude PR (95%CI)	1.61 (1.40, 1.84)	Reference
	Adjusted PR1 (95%CI)	1.44 (1.25, 1.64)	Reference
Anxiety	Number of cases	272	11,934
	Prevalence % (95%CI)	3.3 (2.9, 3.7)	2.5 (2.4, 2.5)
	Age-adjusted prevalence % (95%CI)	3.5 (2.9, 4.2)	2.5 (2.4, 2.6)
	Crude PR (95%CI)	1.37 (1.20, 1.56)	Reference
	Adjusted PR1 (95%CI)	1.31 (1.15, 1.49)	Reference
PTSD	Number of cases	< 20	146
	Prevalence % (95%CI)	0.07 (0.01, 0.12)	0.03 (0.02, 0.04)
	Age-adjusted prevalence % (95%CI)	NA2	NA2

CI = confidence interval; NA = not applicable; PR = prevalence ratio; PTSD = Post-traumatic stress disorder;

¹ Adjusting for maternal age, multifetal pregnancy, any pre-pregnancy medical condition (at least one of heart disease, renal disease, pre-pregnancy diabetes, pre-pregnancy hypertension, severe asthma, lupus, epilepsy, and/or Crohn’s disease), any pregnancy medical condition (at least one of gestational hypertension, gestational diabetes, preeclampsia, and/or anemia), smoking, alcohol, and/or substance use during pregnancy (combined), urban residence, and material and social deprivation

² For PTSD, due to the small number of cases (n < 20) in the Métis group, age-standardization and PR were not calculated

Among Métis pregnancies, pre-pregnancy weight > 91 kg (aPR 1.50, 95% CI 1.10, 2.05), having at least one pre-existing medical condition (aPR 2.64, 95% CI 1.95,3.55), smoking, alcohol, or substance use during pregnancy (aPR 1.63, 95% CI 1.25, 2.12), and urban residence (aPR 1.64, 95% CI 1.18, 2.27) were associated with depression (Table 3). Supplementary Material 2 presents crude and adjusted estimates of associations (PR) between sociodemographic/clinical characteristics and depression in Métis pregnancies (Supplementary material 2: Table S1).

Table 3.

Associations of sociodemographic and clinical characteristics with depression and anxiety during pregnancy in the Métis cohort

Characteristic	Depression	Anxiety
	aPR (95% CI)	aPR (95% CI)
Age group
< 20	0.69 (0.39, 1.23)	0.69 (0.39, 1.20)
20–34	1.00 (ref)	1.00 (ref)
≥ 35	1.15 (0.76, 1.73)	1.06 (0.72, 1.56)
Pre-pregnancy weight > 91 kg
No	1.00 (ref)	1.00 (ref)
Yes	1.50 (1.10, 2.05)	1.16 (0.82, 1.62)
Multifetal pregnancy
No	1.00 (ref)	1.00 (ref)
Yes	1.71 (0.83, 3.53)	2.52 (1.40, 4.53)
Parity
Nulliparous (0)	1.00 (ref)	1.00 (ref)
1	1.12 (0.85, 1.48)	0.85 (0.62, 1.16)
2	1.00 (0.69, 1.45)	1.20 (0.84,1.70)
≥ 3	0.90 (0.56, 1.46)	1.82 (1.25,2.66)
Any pre-existing medical condition1
No	1.00 (ref)	1.00 (ref)
Yes	2.64 (1.95, 3.55)	1.71 (1.24, 2.36)
Any pregnancy-related medical condition2
No	1.00 (ref)	1.00 (ref)
Yes	1.06 (0.73, 1.55)	1.44 (1.02, 2.03)
Prenatal care adequacy
Adequate	1.00 (ref)	1.00 (ref)
Intensive	1.39 (0.79, 2.46)	1.74 (1.00, 2.78)
Intermediate	0.93 (0.70, 1.24)	0.80 (0.59, 1.07)
Inadequate	1.09 (0.75, 1.58)	0.81 (0.56, 1.17)
No care	0.70 (0.28, 1.70)	0.93 (0.46, 1.89)
Smoke, alcohol or substance use during pregnancy
No	1.00 (ref)	1.00 (ref)
Yes	1.63 (1.25, 2.12)	1.57 (1.21, 2.03)
Place of residence
Rural	1.00 (ref)	1.00 (ref)
Urban	1.64 (1.18, 2.27)	2.10 (1.51, 2.91)
Material Deprivation Quintile
Q1- least deprived	1.00 (ref)	1.00 (ref)
Q2	1.04 (0.64, 1.69)	0.97 (0.62, 1.54)
Q3	1.17 (0.72, 1.88)	0.72 (0.45, 1.15)
Q4	1.29 (0.80, 2.06)	0.95 (0.61, 1.50)
Q5- most deprived	1.01 (0.62, 1.65)	0.93 (0.59, 1.48)
Social deprivation quintile
Q1- least deprived	1.00 (ref)	1.00 (ref)
Q2	1.06 (0.63, 1.77)	1.22 (0.75, 2.00)
Q3	1.09 (0.68, 1.75)	0.84 (0.51, 1.37)
Q4	1.07 (0.68, 1.69)	1.20 (0.76, 1.87)
Q5- most deprived	1.17 (0.72, 1.87)	1.35 (0.85, 2.14)

aPR = adjusted prevalence ratio (for all variables listed in this table using a multivariable model); CI = confidence interval; ref = reference; Q = quintile; aPR in bold when 95% CI excludes the value of 1.00

¹Any pre-existing medical condition: at least one of heart disease, renal disease, diabetes, hypertension, severe asthma, lupus, epilepsy, and/or Crohn’s disease

² Any pregnancy-related medical condition: at least one of gestational hypertension, gestational diabetes, preeclampsia, and/or anemia

Anxiety

Anxiety was diagnosed in 272 Métis pregnancies and 11,934 non-Métis pregnancies. The age-standardized prevalence of anxiety in Métis pregnancies was 3.5% (95% CI 2.9%, 4.2%) versus 2.5% (95% CI 2.4%, 2.6%) in non-Métis pregnancies. After adjustment, anxiety was 1.31 times more prevalent in Métis pregnancies (aPR 1.31, 95% CI 1.15, 1.49) (Table 2).

Among Métis pregnancies, six factors were associated with anxiety. Three overlapped with depression: having at least one pre-existing medical condition (aPR 1.71, 95% CI 1.24, 2.36), smoking, alcohol, or substance use during pregnancy (aPR 1.57, 95% CI 1.21, 2.03), and urban residence (aPR 2.10, 95% CI 1.51, 2.91). Three were specific to anxiety: multifetal pregnancy (aPR 2.52, 95% CI 1.40, 4.53), parity ≥ 3 (aPR 1.82, 95% CI 1.25, 2.66), and having at least one pregnancy-related medical condition (aPR 1.44, 95% CI 1.02, 2.03) (Table 3). Additional file presents crude and adjusted estimates of associations (PR) between sociodemographic/clinical characteristics and anxiety in Métis pregnancies (Supplementary material 3: Table S2).

PTSD

PTSD was diagnosed in 146 non-Métis pregnancies and fewer than 20 Métis pregnancies. The prevalence of PTSD was 0.07% (95% CI 0.01%, 0.12%) among Métis pregnancies versus 0.03% (95% CI 0.02%, 0.04%) in non-Métis pregnancies. Due to the small number of observations in the Métis group (n < 20), counts were suppressed to minimize residual disclosure risk, and PTSD age-standardized prevalence was not calculated because age-standardization would be statistically unstable (i.e., some age strata had zero or very small cell counts) (Table 2).

Discussion

This study addresses a significant gap in the literature on Indigenous mental health in Canada by providing a population-based study of mental health disparities among Métis individuals during pregnancy using validated case definitions of mental disorders. The findings revealed that depression was 1.44 times more prevalent and anxiety 1.31 times more common in Métis pregnancies. These findings align with a meta-analysis showing perinatal mental health challenges among Indigenous populations globally were 1.62 times higher than in non-Indigenous populations [11].

No previous studies have used record-linked administrative health data to assess mental health outcomes during pregnancy in Indigenous populations [10, 11], limiting direct comparisons. While some studies have examined postpartum depression among Indigenous peoples in the CANZUS countries [37], research on mental health conditions during pregnancy remains scarce.

Despite a higher prevalence of depression, anxiety, and PTSD in Métis compared to non-Métis pregnancies, overall estimates were lower than those reported in perinatal general population studies using administrative health data [4, 38–40]. A Manitoba-based study reported a 13.5% prevalence of perinatal mood and anxiety disorders [4], while a US study found 6.9% prevalence of depressive disorders [38]. Differences in case definitions likely explain these discrepancies. Some studies classified conditions based on a single physician visit [38], whereas our study required at least two, improving specificity but reducing sensitivity. Additionally, broader diagnostic categories and the use of antidepressant prescriptions as a proxy for depression [38] may have contributed to higher prevalence elsewhere. Similarly, among studies using administrative health databases, PTSD prevalence estimates during pregnancy range from 0.5%−1.7% [39, 40].

Key factors associated with depression and anxiety in Métis pregnancies included pre-existing medical conditions, substance use, and urban residence. Additional risk factors for anxiety were multifetal pregnancy, higher parity, and pregnancy-related medical conditions. These findings highlight the importance of considering Métis social determinants, life stressors, and intergenerational experiences, not just biological pathways.

It is important to note the bidirectional relationship between mental and physical health conditions. Chronic illness can increase depression and anxiety risk, while mental health challenges may worsen physical health through behavioural (reduced self-care) and physiological (chronic inflammation) pathways [19, 41]. Pregnancy-related medical conditions were uniquely linked to anxiety, suggesting that medical complications could heighten psychosocial distress, particularly when culturally safe prenatal care is lacking. Integrating culturally responsive mental health care into maternal health services could help address these concerns.

The association between substance use and mental health during pregnancy also warrants attention. Substance use frequently co-occurs with depression and anxiety, but causal pathways are complex. Pre-existing mental health conditions may increase substance use as a coping mechanism [42], while chronic substance use may worsen mental health through neurobiological changes or stress exposure (e.g., intimate partner violence, financial instability) [42]. Métis individuals have reported using smoking in pregnancy to manage distress [43], highlighting the need for integrated mental health and substance use interventions that are culturally safe and non-stigmatizing.

Urban residence was associated with more depression and anxiety in Métis pregnancies. This contrasts with research suggesting lower mental health challenges among Indigenous adults in urban settings versus those in rural areas [44]. One explanation is that healthcare access in urban areas increases detection and diagnosis. However, urban living may also weaken connections to land, kinship networks, and Métis cultural practices, which are essential to identity, and well-being. This reinforces the need for culturally adapted interventions incorporating Métis ways of knowing, land-based healing, and community-centered approaches.

Our study findings should be interpreted alongside Métis-led qualitative evidence that offers context for perinatal mental health among Métis women, and which is generally strength-based and centered on Métis self-determination, relationality, and cultural continuity [45]. Qualitative studies describe barriers that may contribute to distress during pregnancy, including discrimination and heightened scrutiny across system, apprehension about judgment, mistrust in services, and limited access to culturally safe supports [46–48]. At the same time, they emphasize the benefits of cultural connection, kinship and community support, and respectful, relationship-centered care, especially when services are Métis-informed and culturally safe [46, 47, 49, 50]. This qualitative literature provides plausible context for our results and supports culturally safe, Métis-led approaches to mental health care during pregnancy [45].

Taken together, these findings point to the importance of collaborative, community-driven mental health services for Métis individuals during pregnancy. Pregnancy presents a special opportunity for detection and intervention, as many individuals are motivated to adopt healthier behaviours during this time. Integrating mental health screening with culturally safe pregnancy care could enhance early detection and prevention. Expanding culturally tailored perinatal mental health programs can help mitigate mental health disparities and improve outcomes for Métis individuals, families, and communities.

This study has several strengths. Using large administrative health databases enabled a population-based approach, improving representativeness and generalizability, with the use of validated case definitions for depression, anxiety, and PTSD, ensuring systematic and reliable classification of mental health diagnoses as a particular strength. The case-finding algorithms had high specificity [27–29], effectively minimizing false positives and ensuring accurate classification of unaffected individuals. Another strength was the collaboration with the MNA, aligning research with community priorities and advancing equity-driven data collection. This study also responds to United Nations recommendations and global calls for disaggregated Indigenous health data [51].

Some limitations must be acknowledged. Administrative health databases, while comprehensive, were not designed for research, limiting the depth of information on covariates and unmeasured confounders. Additionally, this dataset reflects individuals who engaged with the healthcare system, potentially underestimating undiagnosed mental health conditions during pregnancy. Relatedly, although we used validated case definitions that prioritize specificity, sensitivity –particularly for depression and anxiety– can be limited, reflecting the sensitivity–specificity trade-off inherent in administrative health data case-finding approaches [30, 31]. This likely results in under-ascertainment of cases; if diagnostic capture differs by Métis status or geography, some differential misclassification is possible. Another key methodological issue is that case definitions included both new and pre-existing diagnoses recorded within 10 months before delivery. While this captures overall mental health burden, it does not distinguish between new-onset cases and pre-existing conditions. Thus, we could not assess incidence or symptom onset. However, this method aligns with prior research using administrative data [52, 53], ensuring comparability across studies. Future research incorporating longitudinal data before conception could further explore mental health trajectories from preconception to postpartum. Finally, while results from 2006–2016 pregnancies may have limited generalization to the present, they provide a valuable benchmark for future studies on mental health trends in Métis populations.

Conclusions

Métis individuals during pregnancy experience higher depression, anxiety, and PTSD prevalence than their non-Métis peers. Key factors associated with increased prevalence in Métis pregnancies include pre-existing medical conditions, substance use, and urban residence. These findings suggest Métis individuals may benefit from systematic mental health screening during pregnancy, particularly where support services are available. Co-occurring factors could serve as screening indicators, facilitating tailored interventions. More broadly, this presents an opportunity for culturally safe, community-driven initiatives to address modifiable risk factors and reduce disparities. Culturally responsive interventions co-developed with Métis communities are essential to improving perinatal mental health. Future research should explore the intersection of structural determinants, cultural connections, and perinatal mental health, to inform policy and programs. Policymakers should also evaluate existing health legislation to remove systemic barriers, integrate culturally responsive care, and ensure Métis communities have equitable access to perinatal mental health resources.

Abbreviations

aPR

Adjusted prevalence ratios

CANZUS

Canada, Australia, Aotearoa/New Zealand, and the United States

CI

Confidence interval

ICD

International Classification of Diseases

Kg

Kilograms

MNA

Otipemisiwak Métis Government of the Métis Nation within Alberta

NA

Not applicable

PR

Prevalence ratio

STROBE

STrengthening the Reporting of OBservational studies in Epidemiology

PTSD

Post-traumatic stress disorder

Authors’ contribution

SO, RB, KB, RJV-L, and MBO contributed to the conceptualization and methodology of the study. JS-L was responsible for data curation. JS-L and MBO conducted data collection and linkage. Formal analysis was performed by SO and JS-L, and all authors –SO, RB, KB, RJV, JS-L, and MBO– contributed to the interpretation of results. MBO secured funding for the study. SO, JS-L, and MBO led the writing of the original draft. MBO is the guarantor for this study; had access to all study data and takes responsibility for the contents of the article. All authors read and approved the final manuscript.

Funding

Canadian Institutes of Health Research (Grant 389286), the Alberta Women’s Health Foundation through the Women’s and Children’s Health Research Institute and the Canada Research Chairs Program (950–232833).

Data availability

Data cannot be publicly shared due to privacy and ethical restrictions under Alberta’s Health Information Act. The coded dataset is securely held by Alberta Health and Alberta Health Services (AHS), the legal custodians. While eligible researchers may request confidential access to Alberta administrative data through AHS (Provincial Research Data Services/Health System Access for Research: https://www.albertahealthservices.ca/research/page8579.aspx), the study’s analytic dataset includes linkage with the Métis Nation of Alberta population registry under a formal research agreement and therefore cannot be released for external access.

Declarations

Ethics approval and consent to participate

This study was approved by the University of Alberta Human Research Ethics Board (Pro00078176), which approved a waiver of consent for this secondary analysis of de-identified administrative health data. The research conformed to the principles of the Helsinki Declaration.

Consent for publication

Not applicable. The study used de-identified administrative data.

Competing interests

The authors declare no competing interests.