Identification of women at high risk of postpartum psychiatric episodes: A population-based study quantifying relative and absolute risks following exposure to selected risk factors and genetic liability

Abstract

Background:

We quantified relative and absolute risks of postpartum psychiatric episodes (PPE) following risk factors: Young age, past personal or family history of psychiatric disorders, and genetic liability.

Methods:

We conducted a register-based study using the iPSYCH2012 case-cohort sample. Exposures were personal history of psychiatric episodes prior to childbirth, being a young mother (giving birth before the age of 21.5 years), having a family history of psychiatric disorders, and a high (highest quartile) polygenic score (PGS) for major depression. PPE was defined within 12 months postpartum by prescription of psychotropic medication or in- and outpatient contact to a psychiatric facility. We included primiparous women born 1981–1999, giving birth before January 1st, 2016. We conducted Cox regression to calculate hazard ratios (HRs) of PPE, absolute risks were calculated using cumulative incidence functions.

Results:

We included 8174 primiparous women, and the estimated baseline PPE risk was 6.9% (95% CI 6.0%–7.8%, number of PPE cases: 2169). For young mothers with a personal and family history of psychiatric disorders, the absolute risk of PPE was 21.6% (95% CI 15.9%–27.8%). Adding information on high genetic liability to depression, the risk increased to 29.2% (95% CI 21.3%–38.4%) for PPE.

Conclusions:

Information on prior personal and family psychiatric episodes as well as age may assist in estimating a personalized risk of PPE. Furthermore, additional information on genetic liability could add even further to this risk assessment.

1 |. INTRODUCTION

1.1 |. Postpartum psychiatric episodes

Following childbirth, women are at increased risk of experiencing postpartum psychiatric episodes (PPE),^1–3 and postpartum depression is one of the most common mental disorders occurring after childbirth.^3–5 Rarer and more severe episodes of PPE include psychosis and mania.^3,6 Across symptom severity and different diagnoses, PPE affects 10%–15% of women giving birth, with the exact prevalence depending both on the definition of timing (ranging from months to 1 year postpartum) and severity of the disorder.^3,5–7

1.2 |. Risk factors for postpartum psychiatric disorders

Established risk factors for PPE include a personal and parental history of psychiatric disorders prior to pregnancy as well as young age.^2,5,6,8–10 A common denominator for these risk factors is they are relatively easy to identify and can be assessed preconceptionally or at the early stages of pregnancy with simple questions. In recent years, studies have quantified the genetic risk for PPE, incorporating single measures of genetic vulnerability for a given disorder (polygenic scores/PGSs), indicating a shared genetic etiology between mood disorders and PPE.^11,12 Risk of developing PPE is therefore inherently influenced by a combination of risk factors as it is observed for other psychiatric disorders; yet, no studies have explored how personal risk factors and genetic vulnerability alone or in combination affect both the relative and absolute risks of PPE.

1.3 |. Risk measurement

In epidemiology, most studies report relative risk estimates representing a risk ratio between two defined groups. Relative risk can be challenging both to convey to patients and readers and implement in clinical use, as the interpretation can easily overestimate the actual importance.^13,14 In comparison, absolute risk of a given outcome provides the proportion of individuals who will develop the disorder of interest. In addition, absolute risk can serve as a more pragmatic metric aiding understanding the true number of individuals affected and furthermore represents clinically relevant information for both healthcare professionals and patients.^13,15 Therefore, the present study aimed to quantify both relative and absolute risks of PPE following exposure to risk factors including young age, personal and family history of psychiatric disorders as well as genetic liability.

2 |. MATERIALS AND METHODS

2.1 |. Study design and population

We conducted a register-based study based on a study population identified from a population-based dataset defined in the Danish Civil Registration System (CRS).¹⁶ The population-based cohort included every singleton born in Denmark between May 1st, 1981 and December 31st, 2005, with a registered mother, who was alive and residing in Denmark by their first birthday (N = 1,472,762).¹⁷ Among those individuals, the iPSYCH2012 study sample was collected, in which psychiatric cases were identified as individuals with a recorded diagnosis of either schizophrenia, affective disorders, bipolar disorder, autism spectrum disorders, or attention-deficit/hyperactivity disorders through December 31st, 2012 in the Danish Psychiatric Central Research Register (N = 57,377).^17,18 Further, a random sample of 30,000 individuals was drawn from the entire population. With this, the iPSYCH2012 contains 86,189 individuals, including both psychiatric cases and a random sample of controls. For the current study, we included all primiparous women from the iPSYCH2012 case-cohort sample born in Denmark between May 1st, 1981, and December 31st, 1999, by Danish-born parents, N = 8174. All women included in the study cohort gave birth to their first child before December 31st, 2015. Note, throughout this article, we, for fluency, refer to pregnant and birthing individuals as “female,” “women,” or “mothers” while acknowledging that not all individuals included in our study choose these labels.

2.2 |. Data sources

2.2.1 |. National registries

The Danish Civil Registration System (CRS)¹⁶ was established in 1968 and includes data and a unique personal identification number (CPR number) on all individuals born in Denmark and individuals residing in the country. The use of CPR allows linkage to all other Danish national registries. Further, this register contains information on, for example, gender and vital status.

The Danish Psychiatric Central Research Register (PCR)¹⁸ was established in 1969 and includes information on all admissions to Denmark’s psychiatric hospitals from this point on. From 1995, data on all outpatient contacts have been added to the register.

Since 1995, all reimbursed prescription drugs sold in Danish pharmacies have been registered in the Danish National Prescription Register (DNPR).¹⁹ The register contains information on the Anatomical Therapeutic Chemical (ATC) classification codes and the date of prescriptions dispensed. Combined with the PCR data, this register was used in this study to identify mild–moderate episodes of postpartum psychiatric disorders (see details of this definition below).

2.2.2 |. Genetic data

Since May 1981, practically all newborns in Denmark have been screened for phenylketonuria, and the dried blood spots used for testing are subsequently stored for diagnostic purposes, screening, quality control, and research.^17,20 For each of the individuals included in the iPSYCH2012 case-cohort sample, the genetic data were obtained from these samples and the material was whole-genome amplified using the Qiagen REPLI-g mini kit and genotyped with PsychChip v1.0 array from Illumina according to the manufacturer’s instructions.²¹ Information on markers not genotyped was attained by imputation using the 1000 genomes phase 3 as a reference panel.¹⁷ The quality control and imputations were done with the Ricopili pipeline.^12,22

PGS can be calculated for specific disorders, diseases, or traits that have been identified and is defined as a weighted sum of genetically associated variants, weighted by the effect sizes on the disorder. The weights are based on genome-wide association studies (GWAS).²³ We derived externally trained PGS for MDD based on SNP weights from GWAS summary statistics, excluding the iPSYCH sample,²⁴ using LDpred.²⁵ We also leveraged individual-level SNP data on individuals with MDD in the iPSYCH sample by deriving another set of internally trained PGSs using BOLT-LMM,²⁶ using a 5-fold cross validation scheme to avoid over-fitting the PGS. For both methods, the set of SNPs was restricted to variants with minor allele frequency >1%, missing values <10%, and SNPs overlapping with HapMap3 (http://www.hapmap.org), ending with a total of 717,871 SNPs for the BOLT-LMM PRS and 166,906 SNPs for the LDpred PRS. The final PGS for MDD was constructed as a meta-PGS to increase its prediction accuracy, that is, a linear combination of PGS obtained from external GWAS summary statistics and individual-level data.²⁷

2.3 |. Outcome: Postpartum psychiatric episodes

The outcome of interest was the first episode of PPE within 12 months of the first childbirth. A PPE episode was defined as either (1) mild–moderate PPE—redeemed prescription of psychotropic medication (ATC codes N05 or N06) or (2) severe PPE—inpatient or outpatient contact to a specialized psychiatric treatment facility with a main diagnosis within the ICD-10 F-chapter (F00-F99) excluding diagnoses of intellectual disability (F70-F79) or substance abuse (F10-F19) as we did not expect these episodes to be triggered by childbirth.

2.4 |. Exposures of interest: Personal, family, or genetic risk factors

We evaluated different established risk factors for PPE and analyzed the risk of developing PPE throughout the 12 months postpartum period individually for each of the included risk factors or in combination.

The investigated risk factors were:

Age at childbirth—

the variable was divided into quartiles, and being a young mother was defined as giving birth before the age of 21.5 years, the lowest quartile of age at the time of childbirth in our study population.

Previous psychiatric history—

We defined previous psychiatric history in women as having a record of redeeming psychotropic medication (ATC codes: N05 and N06) or in- or outpatient treatment at a psychiatric facility prior to the date of first childbirth (ICD-10: F00-F99, excluding F70-F79 and F10-F19).

Psychiatric family history was defined as an admission to a psychiatric facility (F00-F99) among either of the woman’s parents prior to the date of first childbirth. Note, information on redeemed psychotropic medication was not included in this definition as the National Prescription Registry was not established until 1995.

PGS for MDD—

The calculated PGS for MDD for each woman in the cohort was categorized into quartiles, and women within the highest quartile were defined as having a high PGS for MDD.

2.5 |. Statistical analyses

We applied inverse probability weights to take into account the sampling scheme and the overrepresentation of psychiatric cases in the iPSYCH2012 case-cohort sample: The controls were identified in a population without a recorded diagnosis of schizophrenia, affective disorders, bipolar disorder, autism spectrum disorders, or attention-deficit/hyperactivity disorders and were weighted with the inverse probability of being chosen for the random sub-cohort (~54), meaning that each non-case individual counted approximately 54 times in the analyses. Inverse probability weighting is often used in secondary traits and phenotype analysis.²⁸ The precise weights were calculated based on the complete female Danish background population of singletons born May 1st, 1981-December 31st, 1999, to Danish born parents, who were alive and residing in Denmark on their first birthday, and who themselves gave birth to their first child before December 31st, 2015. Similarly, weights were also assigned to cases (individuals with a registered diagnosis of schizophrenia, affective disorders, bipolar disorder, autism spectrum disorders, or attention-deficit/hyperactivity disorders). However, as their probability of being chosen was equal to 100% (per definition in the iPSYCH study design), their corresponding weight was 1. After the assignment of the inverse probability weights, cases and non-case controls were analyzed as one combined sample.

The women included in the study sample were followed from the date of first childbirth and until 12 months postpartum, the date of PPE episode (our defined outcome), death, emigration, or December 31st, 2016, whichever came first.

Using Cox regression in Stata 15.1, we conducted survival analysis with time since childbirth as the underlying time axis, producing the outcome measure of hazard ratios (HRs) and 95% confidence intervals (95% CIs), estimating (unadjusted) relative risks of PPE. The absolute risks of PPE were calculated as cumulative incidence functions using competing risk regression based on Fine and Gray’s proportional subhazards model, accounting for the competing outcome of death.²⁹ To obtain standard errors and confidence bands for the absolute risks, we used bootstrap sampling. Taking the case-cohort design into account, non-cases cohort members were sampled at random with replacement and weighted as described above in the analyses, while cases were retained for each repetition.³⁰ After 1000 repetitions, estimated absolute risk and lower and upper limits of the confidence bands were then calculated as the median and the 2.5th and 97.5th percentiles, respectively, and smoothed using Epanechnikov kernel-weighted local polynomial smoothing.

In the final analysis of absolute risks, crude cumulative incidence for PPE in the cohort was presented, whereas for psychiatric history and each combination of risk factors, adjusted cumulative incidences were reported by adjustment for all remaining risk factors.

2.6 |. Sensitivity analyses

We conducted a sensitivity analysis to validate the impact of the included personal risk factors (not including genetic information) in the entire Danish background population. All investigated risk factors (previous history of psychiatric disorders, being a young mother, and having a family history of psychiatric disorders) were defined as described above. The inclusion criteria for the cohort in the sensitivity analyses were defined in the same way as in our study design: We included all primiparous women born in Denmark between May 1st, 1981, and December 31st, 1999, by Danish-born parents, giving birth to their first child before December 31st, 2015. Note, these analyses were done to confirm representativeness of our results by comparing measures of absolute risks in the sample for the present study with the entire background population.

According to Danish regulations, informed consent is not required when conducting population register-based studies.

3 |. RESULTS

Among 86,189 individuals included in the iPSYCH2012 case-cohort, we identified 8174 women born in Denmark between May 1st, 1981 and December 31st, 1999, to Danish-born parents, who gave birth to their first child before December 31st, 2015, and with available information on the MDD PGS. Among the 8174 mothers, 2169 women had PPE within 12 months of childbirth. A total of 1850 mothers with PPE had a previous history of psychiatric disorders at the time of childbirth, while 319 women experienced their first-ever onset of psychiatric disorder in the postpartum year.

3.1 |. The relative risk of PPE

Table 1 shows the total number of women and number of PPE cases by included risk factors and the corresponding unadjusted HRs and 95% CI. While all investigated risk factors significantly increased the relative risk of PPE within 1 year after childbirth, the highest risk estimate was observed for previous history of psychiatric disorder in the mother compared to mothers without previous psychiatric history, HR 6.53, 95% CI: 4.76–8.97. The other identified risk factors all increased the relative risk of PPE approximately two times: being a young mother (1st quartile:<21.5 years) HR 1.65, 95% CI: 1.10–2.47 (compared to 4th age quartile); family history of psychiatric disorders HR 1.72, 95% CI: 1.22–2.41 (compared to no family history); and PGS for MDD (highest quartile) HR 2.42, 95% CI: 1.63–3.60 (compared to the lowest quartile).

TABLE 1.

Number of women exposed to individual risk factors and the relative risk of postpartum psychiatric episodes (PPE) given by hazard ratios (HRs) and 95% confidence intervals (95% CIs).

Risk factor	N (Women total)	N (PPE cases)	HR (95% CI)
Previous history of psychiatric disorders
Yes	5632	1850	6.53 (4.76–8.97)
No	2542	319	1.00 (Ref.a)
Young mother at first childbirth
1st quartile: <21.5 years	2031	507	1.65 (1.10–2.47)
2nd quartile: 21.6–24.4 years	2043	564	1.39 (0.94–2.06)
3rd quartile: 24.5–27.2 years	2040	564	1.14 (0.78–1.68)
4th quartile: >27.3 years	2060	534	1.00 (Ref.a)
Family history of psychiatric disorders
Yes	6141	1768	1.72 (1.22–2.41)
No	2033	401	1.00 (Ref.a)
High PGSb for MDDc
1st quartile	2043	454	1.00 (Ref.a)
2nd quartile	2044	559	1.55 (1.03–2.33)
3rd quartile	2043	562	1.47 (0.97–2.22)
4th quartile	2044	594	2.42 (1.63–3.60)

^aReference.

^bPolygenic score.

^cMajor depressive disorder.

3.2 |. The absolute risk of PPE

3.2.1 |. Isolated effects of PPE risk factors

Figures 1–4 illustrate the absolute risk of PPE within 1 year among women exposed to each of the investigated risk factors individually, compared to women not exposed to this particular risk factor.

FIGURE 1.

Cumulative incidence of postpartum psychiatric episodes (PPE) within the first postpartum year in primiparous women with and without a previous history of psychiatric disorder. Shaded areas represent 95% confidence intervals.

FIGURE 4.

Cumulative incidence of postpartum psychiatric episodes (PPE) within 1 year after childbirth among primiparous women in the four polygenic score (PGS) quartiles for major depressive disorder (MDD). Shaded areas represent 95% confidence intervals.

Figure 1 shows having a diagnosis of psychiatric disorders prior to childbirth corresponds to a risk of developing a PPE in the year following childbirth of 17.4% (95% CI 14.8%–19.9%). In comparison, women without a previous history of psychiatric disorders had a 2.9% (95% CI 2.1%–3.7%) risk of developing PPE. Approximately 9.1% (95% CI 6.7%–11.6%) of women in the lowest quartile of age (<21.5 years) at first childbirth were diagnosed with PPE (Figure 2). Family history of psychiatric disorder defined as having one or two parents with a diagnosis of psychiatric disorder corresponded to an 8.0% (95% CI 6.8%–9.1%) risk of developing a PPE (Figure 3), while having a PGS for MDD in the highest quartile of the study population meant a risk of PPE within the defined postpartum year of 10.8% (95% CI 8.5%–13.6%), Figure 4.

FIGURE 2.

Cumulative incidence of postpartum psychiatric episodes (PPE) within 1 year after childbirth in the four quartiles of age among primiparous women. First quartile: <21.5 years. Second quartile: 21.6–24.4 years. Third quartile: 24.5–27.2 years. Fourth quartile: >27.3 years. Shaded areas represent 95% confidence intervals.

FIGURE 3.

Cumulative incidence of postpartum psychiatric episodes (PPE) within the first postpartum year in primiparous women with and without a family history of psychiatric disorders. Shaded areas represent 95% confidence intervals.

3.2.2 |. Combination of risk factors

Figure 5 presents the results from the analysis of combined risk factors. The baseline risk of PPE within 12 months postpartum for all women included in the cohort was 6.9% (95% CI 6.0%–7.8%). Having a previous history of psychiatric disorders translated into 15.7% (95% CI 13.2%–18.4%) risk of PPE, adding the risk factors of being a young first-time mother (youngest quartile <21.5 years) increased the risk to 20.2% (95% CI 14.7%–25.9%), and further, a family history of psychiatric disorders resulted in a combined absolute risk of 21.6% (95% CI 15.9%–27.8%). Finally, adding the genetic risk factor of being in the highest quartile of PGS for MDD increased the absolute risk of PPE to a total of 29.2% (95% CI 21.3%–38.4%). While the absolute risk of PPE was 29% among the group of women with all investigated risk factors, these findings must be interpreted in the light of this group of cases representing only approximately 86 (4.0%) of the total PPE cases in the cohort (Figure 5).

FIGURE 5.

Cumulative incidence of postpartum psychiatric episodes (PPE) within the first postpartum year in primiparous women in the entire cohort (background population) or with different combinations of risk factors. Shaded areas represent 95% confidence intervals.

3.2.3 |. Sensitivity analyses conducted in the entire Danish background population

In sensitivity analyses conducted on the female background population, we found similar results to those calculated in the iPSYCH2012 sample, as absolute risk of PPE in the background population was 7.1% (95% CI 6.9%–7.2%). Furthermore, absolute risk of PPE among women with a history of psychiatric treatment (either with psychotropic medication or as an in- or outpatient at a psychiatric facility) was 18.8% (95% CI 18.4%–19.4%), being a young first-time mother (<21.5 years) accounted for an absolute risk of PPE of 10.0% (95% CI 9.6%–10.5%). Also, having a family history of psychiatric disorders contributed to an absolute risk of 10.9% (95% CI 10.4%–11.3%). The combination of all three personal risk factors amounted to an absolute risk of 24.4% (95% CI 22.3–26.6%) for PPE within 12 months after childbirth among primiparous mothers from the Danish background population (results not shown).

4 |. DISCUSSION

This study showed that primiparous women with a combination of three personal risk factors: (1) Previous history of psychiatric disorder, (2) Being a young mother, and (3) Having a family history of psychiatric disorders, have an absolute risk of 22% for a PPE episode within 1 year postpartum. Adding information regarding the defined genetic risk of a high PGS for MDD additionally increased the absolute risk of PPE within 1 year of childbirth to a total of 29%.

4.1 |. Importance of findings

4.1.1 |. Estimating risk of PPE

Many studies have established risk factors for developing PPE, including previous history of psychiatric disorders, family history of psychiatric disorders, but also primiparity, and low perceived partner support.^{3,5,6,9,31,32} However, these risk factors have been investigated and presented through results providing relative risk estimates, which are difficult to translate into clinically useful information. While the description of relative risks might allow us to identify which risk factors are of particular importance compared to others, they do not provide a perceptible risk evaluation for the individual women requesting advice on their risk of PPE. Prediction studies focused on postpartum psychiatric episodes can also assist identification of which risk factors are of particular importance. These include indicators of previous depression (identified either as diagnoses or medication use) or other mental disorders which has been identified in several PPE prediction studies and is in line with our own results.^33–35 Moving forward, our extended knowledge of individual and combined effects of risk factors will hopefully assist considerations toward personalized and targeted interventions aimed specifically at at-risk individuals, while it will also be relevant to consider how identified risk factors from PPE studies compare to studies aimed at predicting psychiatric disorders in general.

With the present study, we chose three well-established personal risk factors for PPE (previous history of psychiatric disorders, young maternal age, and family history of psychiatric disorders), which are all easy to assess in early pregnancy or even in women seeking preconceptional counseling. The presence of these three risk factors in combination corresponded to an absolute risk of PPE of 22% within 1 year after childbirth. We confirmed that the most substantial risk factor for PPE was a previous history of psychiatric disorders,^5,6,32 entailing a 17% risk for subsequent PPE, yet both being a young mother and having a family history of psychiatric disorders, seen isolated, also amounted to ~10% risk of PPE each. We also confirmed that having a higher PGS for MDD is associated with a higher risk of PPE.¹² In combination with the three personal risk factors described above, having a PGS for MDD in the highest quartile resulted in a combined risk of 29%. Considering the results presented in Figure 5, we note the above-mentioned differences in absolute risks are based on relatively few cases with overlapping confidence intervals, and replication of the results will be necessary in larger samples which also will allow diagnoses-specific analyses.

4.2 |. Applying PGS in clinical practice

With increasing access to genetic information, the usefulness of PGS in clinical psychiatric practice has been discussed.^36–38 However, PGS only captures some of the genetic risk in an individual, and additional information on family history of psychiatric disorders will still be relevant to obtain from the patients when evaluating the genetic risk component.^37,38 Furthermore, the genetic risk of psychiatric disorders only represents part of an individual’s combined risk of developing psychiatric disorders, as the overall risk of PPE is composed of genetic, environmental, and personal risk factors.³⁸ Therefore, PGS, at the current stage, could be considered as a supplement in the overall risk assessment of patients, although direct plans for implementation are premature.

4.3 |. Strengths and limitations

The data in this study were based on the iPSYCH2012 study sample, one of the largest data sources combining genetic and environmental information on psychiatric disorders.¹⁷ The design of the iPSYCH2012 study sample, selected from a national, representative population, strengthens the generalizability of our results to the general Danish population. Furthermore, we obtained data from the Danish national registers, ensuring follow-up information of high validity on all individuals in the study sample.^16,18,19

There are inherent methodological limitations to this study: We included a broad definition of PPE with both mild–moderate cases (redeemed psychotropic medication) and severe cases (inpatient or outpatient treatment at a psychiatric facility), prevalent cases (mothers with a previous history of psychiatric disorders) and incident cases (first onset in the postpartum period), and we extended the time period of potential PPE onset to 12 months postpartum. This resulted in a heterogeneous group of PPE; but was done to ensure statistical robustness of the study. In the current study, we defined a crude dichotomous measure of previous psychiatric history (yes/no) for the women in the cohort

The study population in the present study was relatively young, with the oldest participants being 34 years old. This was due to the origins and sampling strategies of the dataset. Also, due to the inclusion criteria of the iPSYCH2012 cohort, all mothers were Danish-born, and a large proportion of our PPE cases had a previous psychiatric history prior to delivery (85%). These characteristics jointly influence the generalizability of the results to older/other populations.

Younger maternal age has been identified as a risk factor for postpartum psychiatric episodes,³² but we could not investigate more detailed associations and a possible U-shaped association between maternal age and PPE risk due to the relatively young age of our cohort. Furthermore, since our study cohort was relatively young, the women included in our study were not past their reproductive age. We, consequently, only included primiparous mothers in the analyses.

For the course of this study, we investigated a few known and easily identifiable risk factors of PPE: being a young mother, having a previous psychiatric history, having a family history of psychiatric disorders, and a genetic vulnerability for MDD. However, we acknowledge that other and alternative risk factors could be of relevance for the risk of developing PPE and moving forward, these could be added and investigated in future studies

To conclude, young mothers with a preexisting personal psychiatric history and parental history of psychiatric disorders have a 22% absolute risk of developing mild, moderate, or severe PPE within 1 year postpartum. Hence, it may be possible to identify a vulnerable group of women at substantial risk of PPE with three simple measures easily evaluated before or in early pregnancy. Furthermore, adding measures of genetic vulnerability for MDD elevates the combined absolute risk of PPE within 1 year of childbirth to 29%. All these observations must be considered acknowledging the baseline risk of PPE within 12 months after first childbirth in the entire cohort was 7%, and that the group of women with a particularly high risk of PPE only represents approximately 4.0% of the total PPE cases in the cohort.

Summations

• A young mother with a preexisting personal and familial history of psychiatric disorders has a 22% absolute risk of developing a mild to severe postpartum psychiatric episodes (PPE). Furthermore, adding measures of genetic vulnerability for depression elevates the combined absolute risk to 29%. If replicated, this result could support estimating a personalized risk of psychiatric episodes following childbirth.

Limitations

• We investigated few known and easily identifiable risk factors of postpartum psychiatric episodes (PPE): Being a young mother, having a personal or family psychiatric history, and a genetic vulnerability for depression, but we acknowledge that this approach limits generalizability. Other and alternative risk factors could be of relevance for the risk of developing PPE and moving forward, these could be added and investigated in future studies.

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no new data were created or analyzed in this study.