A Mendelian randomization approach for advancing healthcare innovation: From pregnancy hypertension to schizophrenia

Abstract

Background:

Schizophrenia, a severe mental disorder, is strongly associated with perinatal factors, including pregnancy hypertension, which affects fetal neurodevelopment. The causal relationship between pregnancy hypertension and schizophrenia remains unclear.

Objective:

To investigate the causal relationship between pregnancy hypertension and schizophrenia using a two-sample Mendelian randomization (MR) study.

Methods:

Pregnancy hypertension (cases/controls: 7686/115,8993) was considered as exposure and schizophrenia (cases/controls: 76755/243,649) as outcome. A two-sample MR study genetically estimated associations, with the inverse variance-weighted method as primary analysis. A two-step MR study explored potential mediators. Sensitivity tests assessed pleiotropy, heterogeneity, and stability.

Results:

Four single nucleotide polymorphisms (SNPs) were identified as genetic instruments for pregnancy hypertension (P < 5 × 10^-8, r² < 0.001). Pregnancy hypertension increased the risk of schizophrenia (odds ratio = 1.098, 95% CI = 1.025–1.173, P = 0.007), while reverse MR analysis found no significant link (odds ratio = 1.03, 95% CI = 0.980–1.082, P = 0.235). The two-step study revealed pregnancy hypertension elevated levels of 1,4-Dihydroxy-2-naphthoic acid (odds ratio = 1.351, 95% CI = 1.097–1.664, P = 0.004), which decreased schizophrenia risk (odds ratio = 0.97, 95% CI = 0.942–0.998, P = 0.040).

Conclusion:

Pregnancy hypertension is a high-risk factor for schizophrenia, with 1,4-Dihydroxy-2-naphthoic acid potentially mitigating this effect. These findings offer insights for the prophylaxis of schizophrenia and highlight the need for further research into underlying mechanisms.

Keywords

Pregnancy hypertension schizophrenia Mendelian randomization study perinatal factor

1. Introduction

Schizophrenia, a prevalent mental illness, is marked by diverse levels of cognitive deficits, emotional deviations, and behavioral irregularities, collectively impairing fundamental perceptual, reasoning, and judgmental processes.¹ This factor stands as the primary reason for worldwide disability, impacting about 1 out of every 100 people, which has a significant financial and societal cost because it manifests in early adulthood, causing impairment in social and vocational settings as well as a general reduction in life expectancy.² Most importantly, schizophrenia has significantly increased rates of suicide and mortality from all causes compared with the general population.³ Although the precise etiology of schizophrenia is unknown, schizophrenia is reported to be a multifaceted disease and its whole course could be affected by the interaction of environmental influences and genetic susceptibility,^4,5 potentially impacting normal brain development and maturation.^6–8 A increasing amount of studies suggest that the development of the fetal nervous system is significantly involved in the emergence of schizophrenia susceptibility.^9,10 Perinatal environmental factors, including infections, dietary deficits, mother stress, and obstetric difficulties, affect not only delivery outcome but also offspring neurodevelopment.¹¹ A meta analysis reveals a link between mother preeclampsia and a higher chance of children schizophrenia.¹² Furthermore, a different research demonstrates that there exists latent shared genes between preeclampsia and schizophrenia,¹³ which is the basis of further researches. Schizophrenia, most happened after adolescence, which leaves hysteresis in the diagnosis and treatment.¹⁴ On the other hand, current antipsychotics still have many limitations, such as severe neurological and metabolic side effects, sexual dysfunction, and agranulocytosis.¹⁵ Therefore, it is crucial to investigate the genuine relationship between schizophrenia and pregnancy disorders, as this could aid in the early detection and prevention of schizophrenia.

Worldwide, approximately 10% of pregnancies are impacted by pregnancy hypertension (PH), which plays a major role in the morbidity and mortality of mothers, fetuses, and newborns.¹⁶ Expectant mothers suffering from PH face risks of placental abruption, strokes, lung swelling, thromboembolic incidents, widespread intravascular clotting, and failure of multiple organs.^17,18 The risks to the fetus such as intrauterine growth retardation, premature birth and fetal distress are increasing during preeclampsia,¹⁹ which are also associated with low birthweight, prolonged high-level neonatal care, and postnatal death.²⁰ Additionally, PH may potentially have a profound influence on offspring's neurodevelopment.²¹ Studies reported that exposure to hypertensive disorders of pregnancy increases the risk of abnormal brain development and neurological diseases in children.¹¹ Many psychiatric disorders, such as schizophrenia and depression, have been found to be linked with PH.^22,23 Several retrospective cohort studies have identified an increased risk of motor and mental developmental abnormality in offspring related with maternal hypertensive disorders of pregnancy.^24–26 However, the association of PH and schizophrenia has not been identified on the genetic aspects.

Recently, it has been discovered that PH patients also have gut microbiota dysbiosis.²⁷ Wu's study has showed that dysfunction of gut microbiota can directly or indirectly lead to the occurrence of psychiatric disorders, such as schizophrenia, depression, autism, etc., through neural, immune, and gut endocrine signaling pathways.²⁸ The 1,4-Dihydroxy-2-naphthoic acid (1,4-DHNA), a biosynthesis of Escherichia, exhibit significant AhR (aryl hydrocarbon receptor) agonist activity in multiple cell types. The AhR is believed to play a crucial role in the connection between the gut bacteria and mental state which may concatenate PH and schizophrenia through environment factor.²⁹ Thus, it's interesting to explore and identify the potential relationship of PH, schizophrenia and gut microbiota.

Mendelian randomization (MR) represents a developing approach in epidemiology, genetically evaluating the potential causal connection between exposure and result via instrumental variables (IVs) through genetic variations [single nucleotide polymorphisms (SNPs)].³⁰ The development of disease is the result of multi-gene and multi-factor (environment, diet, etc.). Current genome-wide association analyses (GWASs) have identified millions of genetic variants linked to disease outcomes, providing the basis for MR analysis.³¹ The gene-outcome relationship is not influenced by typical confounding factors like postnatal environment, socioeconomic status, and behavioral habits. Consequently, the causal relationships inferred from MR studies are considered valid.³² Therefore, MR can avoid confounding factors in common clinical studies and determine the direction of causation in genetic correlation.³³ Our research involved both a two-sample and a two-stepMR analysis to evaluate the causal relationship and pinpoint possible intermediaries between PH and schizophrenia.

2. Materials and methods

2.1. Study design

Our MR study was conducted from October 15, 2023 to April 5, 2024, utilizing publicly accessible large-scale GWASs data to examine the genetic relationship and potential mediator between PH and schizophrenia through a two-sample MR analysis. Specifically, genetic variation as an instrumental variable for risk factors must satisfy the following conditions: (1) a reliable association with the risk factors under study (correlation hypothesis); (2) independence from any known or unknown confounding factors (independence assumption); (3) influencing outcomes solely through risk factors and not through any other direct causal pathways (excluding restrictive assumptions).³⁴

The MR analysis flowchart of PH on schizophrenia is illustrated in Figure 1. The overall impact of a given exposure on the result can be broken down into its direct and indirect consequences.³⁵ Thus, we also made an attempt to find latent mediator that could have an impact on the whole MR analysis. The direct effect of PH on schizophrenia is the effect that occurs without any mediator. The total effect of PH on schizophrenia includes both the direct effect and the effect mediated by any other factors. A two-step MR study was explored which include the casual effect of PH on mediator as first step and the casual effect of mediator on schizophrenia as second. Specifically, the aim is to determine whether PH is a causative factor of schizophrenia or schizophrenia itself has a causal impact on PH. Hence, we conducted bidirectional MR analyses, which means we first performed MR analysis from PH to schizophrenia, and the direction was reversed subsequently. The analysis of MR and clumping was performed using the packages “Mendelian Randomization” and “Two-Sample MR” (https://cran.r-project.org/package-MendelianRandomization) in R.

Figure 1.

Flowchart of the MR analysis. This design postulates a bidirectional association between PH and schizophrenia, disregards any association with confounding variables, and verified the patent mediating effect. The three assumptions of MR are as follows: (1) instrumental variables must be strongly associated with PH; (2) instrumental variables could not be linked with any confounders; and (3) instrumental variables must influence schizophrenia only through PH, not through other pathways. The dash lines indicate irrelevance, and the solid lines indicate relevance. Abbreviations: 1,4-DHNA, 1,4-Dihydroxy-2-naphthoic acid; MR, Mendelian randomization; PH, pregnancy hypertension.

2.2. Data resources

Summary statistics of PH were obtained from the FinnGen Biobank, which included a sample population of 123,579 European individuals. The distribution between cases and controls was 7686 and 115,893 respectively. The FinnGen Biobank is a notable collaboration between public and private entities, aimed at collecting and evaluating genetic and health data from half a million participants (https://www.finngen.fi/en, accessed on July 6th, 2024). Summary statistics of schizophrenia come from the Integrative Epidemiology Unit (IEU) dataset belonging to the Open GWAS project. This dataset contains 320,404 European individuals, with case/control ratios of 76755/243649. And data of 1.4-DHNA, also from the IEU dataset, contained 7738 European individuals. The IEU Open GWAS project is an extensively curated collection of complete GWAS summary datasets comprising approximately 347,679,150,230 genetic associations derived from over 50 thousand freely accessible or downloadable datasets. As our study utilized publicly available or published GWAS summary data that had already been approved by relevant institutional and ethics review boards, additional ethical approval was not required. Our study adhered to the STROBE-MR reporting guidelines for MR in observational studies. Table 1 provides detailed information about each GWAS dataset used in our investigation.

Table 1.
Details of GWAS data included in MR analyses.

Trait Consortium Ethnicity Sample size

PH FinnGen Biobank European 123,579

Gut bacteria pathway abundance (1,4-DHNA) UK biobank European 7738

Schizophrenia UK biobank Mixed 320,404

Trait	Consortium	Ethnicity	Sample size
PH	FinnGen Biobank	European	123,579
Gut bacteria pathway abundance (1,4-DHNA)	UK biobank	European	7738
Schizophrenia	UK biobank	Mixed	320,404

Abbreviations: 1,4-DHNA, 1,4-Dihydroxy-2-naphthoic acid; MR, Mendelian randomization; PH, pregnancy hypertension.

2.3. Genetic instrumental variables

In the investigation of MR analysis, SNPs pinpointed via GWASs serve as IVs to explore the causal relationship between exposures and outcomes. To ensure robustness, we employed stringent criteria for selecting genetic variations associated with exposure, including a significance threshold of P < 5 × 10⁻⁸ or P < 2 × 10⁻⁵, independent SNP pruning within a 10,000 kb window, and a criterion of r²< 0.001.³⁶ SNPs that were not available for the outcome or showed significant associations (P < 5 × 10⁻⁸) with the outcome were excluded to avoid violating assumptions. We also used the PhenoScanner website to remove SNPs associated with confounding factors that could influence the outcome. Exposure-related SNPs from the outcome datasets were selected based on minor allele frequencies greater than 0.01 after harmonizing their alleles’ orientation correctly. The strength of each SNP was assessed using F-statistics calculated by applying the formula: F = R² (N - K - 1) / (K (1 - R²)), where R² represents variance in exposures explained by genetic variance, K is the number of SNPs, and N is sample size. In order to mitigate weak instrument bias in MR analysis, only SNPs with an F-statistic exceeding 10 were chosen according to established guidelines.³⁷ Table 2 showed the final IVs of MR we used.

Table 2.
Details of the selected IVs in Two-step MR analysis.

Exposure Outcome Number of SNPs SNP F Value

PH 1,4-DHNA 4 rs13306561 44.24

rs167479 46.19

rs16998073 55.18

rs236721 48.98

1,4-DHNA Schizophrenia 26 rs10192555 41.82

rs10793056 41.72

rs10950493 41.57

rs11030125 45.10

rs11761901 48.89

rs12156434 38.46

rs12432473 36.25

rs139901010 41.07

rs207254 46.48

rs2834173 43.02

rs28433015 37.86

rs28634397 43.67

rs34011530 41.37

rs34278344 39.24

rs395671 39.05

rs4585628 41.83

rs61861701 55.10

rs62079926 47.64

rs62465767 41.47

rs63721662 41.55

rs6925364 38.80

rs73227800 38.36

rs7504312 38.40

rs76292847 39.24

rs7691920 43.41

rs9886448 41.67

rs9941912 50.53

rs996888 44.88

PH Schizophrenia 4 rs12901406 884.80

rs13306561 1012.22

rs16998073 1262.57

rs236721 1120.64

Abbreviations: 1,4-DHNA, 1,4-Dihydroxy-2-naphthoic acid; MR, Mendelian randomization; PH, pregnancy hypertension; single nucleotide polymorphisms SNP; instrumental variables, IVs.

2.4 Mediation analysis

After completing a bidirectional two-sample MR analysis to explore the link between PH and schizophrenia, we further carried out a two-step MR analysis to examine potential mediating pathways in the causal relationship. The total influence of any exposure on an outcome can be divided into direct and indirect effects. The indirect effect, also known as mediated effect, refers to the influence of PH on schizophrenia through a mediator. This can be calculated by multiplying the MR effect of PH on the mediator (denoted as ‘a’) with the MR effect of the mediator on schizophrenia after adjusting for the genetically determined PH's influence on schizophrenia.³⁸ Then the mediator proportion can be obtained by the ratio of indirect effect on whole effect.³⁹

2.5 Sensitivity analysis

We employed various MR techniques for conducting MR analysis, including inverse-variance weighted (IVW), weighted median, weighted mode, simple mode, and MR-Egger methodologies.⁴⁰ We utilized MR-Egger regression to investigate the potential pleiotropic effects of the chosen SNPs as IVs and assess whether horizontal pleiotropy was influencing our analysis.⁴¹ The Cochran's Q test was utilized to evaluate the presence of heterogeneity and determine if any IV influenced the results, ensuring consistency with the assumptions of MR. Additionally, a sensitivity analysis known as “leave-one-out” was conducted to identify significant SNPs contributing to heterogeneity among different modes of inheritance. Sensitivity analysis involved generating forest plots, scatter plots, and funnel plots. To enhance result robustness, outliers in MR analysis were identified and eliminated using Radial-MR before repeating the IVW estimate. Furthermore, F-statistics were computed to assess instrument strength in terms of correlation between IVs and exposure factors, while R² indicated the proportion of phenotypic variation explained.⁴¹ MR analysis was performed by R software (version 4.2.0), Two-Sample MR package (version 0.5.6), and Radial-MR package (version 1.0).⁴² Statistics were deemed significant at P < 0.05.

3. Results

3.1. Selection of IVs

In the MR analysis of PH on schizophrenia, we identified four SNPs that showed significant associations below the genome-wide threshold (P < 5 × 10-8, r²< 0.001, and kb < 10000). None of these SNPs strongly associated with any other confounding factors after alignment in Phenoscanner. Likewise, for the MR analysis of PH on 1,4-DHNA, we also obtained four SNPs followed the genome-wide threshold (P < 5 × 10-8, r2 < 0.001, and kb < 10000). Respectively, in the MR analysis of 1,4-DHNA on SCZ, we initially identified a set of 2 SNPs that did not reach significance according to the genome-wide threshold (P < 2 × 10⁻⁵, r²< 0.001, and kb < 10000). To identify potential outliers deviating significantly from others using IVW-radial method, we retrieved one SNP (rs80031880), resulting in a final set of 2 significant SNPs after data retrieval from the outcome GWAS and harmonization (Table 2).

3.2. Bidirectional two-sample MR analysis

3.2.1 Causal effect of PH on schizophrenia

A positive correlation was observed between genetically predicted PH and an increased risk of schizophrenia IVW showing an odds ratio (OR) = 1.098, 95% confidence interval (CI) = 1.025–1.173, P = 0.007) (Table 4). Our analysis using MR-Egger regression indicated no presence of horizontal pleiotropy (intercept-0.001, P-0.95), and Cochran's Q test revealed no heterogeneity (Table 3). Each method's SNP was represented by a black dot, with the x-axis depicting the impact of increasing SNP on exposure and y-axes depicting the impact on the outcome. The slope of each line indicated the potential strength of causal correlation. Notably, leave-one-out analysis showed no significant alteration in the causal association between PH and schizophrenia, suggesting that individual SNPs did not influence our causal estimation results. Figure 2 illustrates all findings obtained through various methods.

Table 3.
Heterogeneity and pleiotropy analysis in the MR study.

Heterogeneity Pleiotropy

Exposure Outcome Method Cochran's Q P Value Egger-Intercept P Value

PH 1,4-DHNA MRE 1.31 0.51 −0.10 0.45

IVW 2.17 0.53

1,4-DHNA Schizophrenia MRE 25.50 0.37 −0.004 0.59

IVW 25.80 0.41

PH Schizophrenia MRE 1.09 0.57 0.001 0.95

IVW 1.10 0.77

Schizophrenia PH MRE 223.75 0.17 −0.01 0.07

IVW 227.26 0.14

		Heterogeneity			Pleiotropy
PH	1,4-DHNA	MRE	1.31	0.51	−0.10	0.45
		IVW	2.17	0.53
1,4-DHNA	Schizophrenia	MRE	25.50	0.37	−0.004	0.59
		IVW	25.80	0.41
PH	Schizophrenia	MRE	1.09	0.57	0.001	0.95
		IVW	1.10	0.77
Schizophrenia	PH	MRE	223.75	0.17	−0.01	0.07
		IVW	227.26	0.14

Table 3 shows that our analysis using MR-Egger regression indicated no presence of horizontal pleiotropy, and Cochran's Q test revealed no heterogeneity. Abbreviations: 1,4-DHNA, 1,4-Dihydroxy-2-naphthoic acid; IVW, inverse variance weighted; PH, pregnancy hypertension; MR, Mendelian randomization; MR egger, MRE ; SM, simple mode.

Figure 2.

Causal effect of PH on schizophrenia via a two-sample MR analysis. A. Scatter plot showed the causality of PH on schizophrenia, with the slope of each line corresponding to the estimated MR effect per method; B. Leave-one-out plot displayed MR effect by eliminating individual SNPs one by one and combined result; C. Forest plot of individual and combined SNP MR-estimated effect sizes; D. Funnel plot check whether there is heterogeneity among individual genetic variants. Abbreviations: MR, Mendelian randomization; PH, pregnancy hypertension; SNP, single nucleotide polymorphism.

Figure 3.

Causal effect of schizophrenia on PH. A. Scatter plot showed the causality of schizophrenia on PH, with the slope of each line corresponding to the estimated MR effect per method; B. leave-one-out plot displayed MR effect by eliminating individual SNPs one by one and combined result; C. forest plot of individual and combined SNP MR-estimated effect sizes; D. funnel plot check whether there is heterogeneity among individual genetic variants. Abbreviations: MR, Mendelian randomization; PH, pregnancy hypertension; SNP, single nucleotide polymorphism.

Table 4.

An estimate of the relationship between PH and schizophrenia based on bidirectional MR estimates.

MR Analysis	Beta	SE	SNPs	Method	OR	95% CI	P value
PH on schizophrenia	0.093	0.035	4	IVW	1.098	1.025–1.176	0.007
	0.081	0.170	4	MRE	1.085	0.776–1.517	0.678
	0.078	0.041	4	WME	1.081	0.997–1.173	0.058
	0.064	0.055	4	WMO	1.067	0.956–1.190	0.327
	0.066	0.060	4	SM	1.069	0.950–1.202	0.348
Schizophrenia on PH	0.029	0.025	207	IVW	1.030	0.980–1.082	0.235
	0.203	0.099	207	MRE	1.225	1.001–1.491	0.043
	0.005	0.035	207	WME	1.005	0.937–1.079	0.868
	−0.073	0.117	207	WMO	0.929	0.738–1.169	0.533
	−0.054	0.117	207	SM	0.946	0.752–1.190	0.640

Abbreviations: CI, confidence interval; IVW, inverse variance weighted; MR, Mendelian randomization; MRE, MR egger; OR, odds ratio; PH, pregnancy hypertension; SM, simple mode; SNP, single nucleotide polymorphism; WME, weighted median; WMO, weighted mode.

3.2.2 Causal effect of schizophrenia on PH

The reverse MR analysis revealed the genetic predisposition to schizophrenia had no effect on PH. Therefore, our MR findings did not support a causal role for schizophrenia in the development of PH, avoiding the potential effects of reverse causality. Figure 3 displayed the estimates of all MR analysis methods. Table 4 showed the bidirectional MR analysis of PH and schizophrenia with all methods.

3.3. Two step MR analysis

We performed a two-step MR study to identify potential mediating pathways in the causal association between PH and schizophrenia. 1,4-DHNA was about to be a potential mediator between PH and schizophrenia after the two-step MR analysis targeting the gut microbiota. The MR analysis indicated the genetically predicted PH was positively associated with an increased risk of 1,4-DHNA [(IVW: OR = 1.351, 95% CI = 1.097–1.664, P = 0.004]. Furthermore, 1,4-DHNA was positively associated with a decreased risk of schizophrenia [(IVW: OR = 0.97, 95% CI = 0.942–0.998, P = 0.040). The finding of our analysis indicated that 1,4-DHNA could be a suppressive factor in the effect of PH on schizophrenia. The results of two-step MR study, obtained using all methods, are illustrated in Figure 4.

Figure 4.

Mediation analysis of the effect of PH on schizophrenia via potential mediators. This figure displays the causal effect of PH on schizophrenia via five MR methods, secondly shows the association of PH, 1,4-DHNA and schizophrenia. The error bars represent 95% CIs, P < 0.05 was considered significant. Abbreviations: CI, confidence interval; 1,4-DHNA, 1,4-Dihydroxy-2-naphthoic acid; IVW, inverse variance weighted; PH, pregnancy hypertension; MR, Mendelian randomization; MR egger, MRE; OR, odds ratio; SM, simple mode; WME, weighted median; WMO, weighted mode.

4. Discussion

Our MR study first identified that PH leads to increased rate of the occurrence of schizophrenia, which implied the effect of prenatal disorders on neuropsychiatric disorders in offspring. We further found PH increased the risk of 1,4-DHNA, but the increased 1,4-DHNA decreased the occurrence of schizophrenia via a two-step study, suggesting that 1,4-DHNA might be involved in the pathogenesis of schizophrenia induced by PH.

Schizophrenia is a complex and incurable disease, bothering the majority of patients and clinicians.⁴³ Despite schizophrenia usually manifesting in early adulthood, various pieces of evidence indicate that its development starts at an early stage in neurodevelopment.⁴⁴ Brain development can be hindered by genetic and initial environmental risk elements.⁴⁵ The schizophrenia polygenic risk score in patients with perinatal complications was five times higher than in healthy controls, indicating a correlation between genetic and obstetric risk factors. The initial stages of neurodevelopment are marked by the creation of synaptic links, which persists during one's early years.⁴⁶ Disturbed neurodevelopmental process can lead to widespread impairment of neural communication and cognitive deficits.^47,48

Numerous studies have demonstrated the connection between hypertension disorders during pregnancy and an increased risk of mental and behavioral disorders in offspring, such as autism, attention-deficit hyperactivity disorder, anxiety, distress and schizophrenia.^24,25,49 However, previous studies primarily focus on clinical observation and data analysis, which don’t comprehensively evaluate the roles of potential confounders, mediators, and moderators. Despite the aforementioned findings, it is still unknown how PH influences offspring neurodevelopment especially schizophrenia occurrence. Hence, within the context of this study, our work investigated the genetic link between PH and schizophrenia and tried to identify the underlying mechanism by using the MR analysis. By using the PhenoScanner website, we eliminated SNPs associated with other mental disorders, such as stress, depression and obsession, which could have impacted the outcome as confounding factors. After adjusting for confounding and other variables, we discovered that PH causally impact on schizophrenia. MR-Egger regression, which provides valuable assessment of whether horizontal pleiotropy distorts the study, was used to evaluate the potential pleiotropy of IVs selected. There was no horizontal pleiotropy, as well as heterogeneity in selected IVs suggesting that our findings may accurately reflect the genuine impact. Additionally, the reverse MR study identified that schizophrenia did not influence the occurrence of PH. Hence, we could conclude that PH causally affect and do increase the risk of offspring schizophrenia, which consisted with previous studies,^50,51 which provide a new insight into the biology of schizophrenia. However, the accurate mechanism is still obscure. For that, our research endeavors to identify potential mediating factors that may have an impact on this process.

According to current studies, a higher risk of schizophrenia was strongly correlated with PH genetic predictions, and this facilitation could be partly rescued by the gut bacterial pathway abundance. Emerging researches have highlighted differences in gut microbiota composition among individuals with various psychiatric disorders, including schizophrenia.^52,53 Besides the composition, the metabolites of gut microbiota are closely linked to mental health.^54,55 Hence, we examined if the gut microbiota or any related substances are involved in the connection between PH and schizophrenia. We analyzed several dataset of gut macrobiotics and found that 1,4-DHNA served as a potential link between PH and schizophrenia. 1,4-DHNA is an intermediate in the biosynthesis of menaquinone (vitamin K2) in Escherichia coli⁵⁶ which was reported to prevent and reverse anhedonia-like behavior in mice through arising from its activities as AhR ligands.²⁹ Nevertheless, few studies focused on its correspondence with schizophrenia to date. Our study demonstrated that 1,4-DHNA has been found to have a suppressive effect on the impact of PH on schizophrenia. Furthermore, this result raises the attention of whether altered gut microbiota contributes to the development of psychiatric disorders, particularly when PH is a prerequisite. The gut microbiota dysbiosis has been observed in patients with PH.⁵⁷ An example of this is a study in which it was discovered that there was a significant decrease in the relative abundance of beneficial flora, such as bifidobacterium, in patients with PH compared to the control group.⁵⁸ In other hand, the changes of gut microbiota or their metabolites are different in various psychiatric diseases.^59,60 Thus, we hypothesize that alterations in the beneficial microbiota of patients with PH may contribute to the development of schizophrenia in their offspring. Further research is needed to elucidate the precise mechanism involved.

5. Strengths and limitations

The current research exhibits certain advantages. First of all, there was few researches to explore the casual relationship between PH and schizophrenia. In addition to demonstrating the genetic link between the two illnesses, our investigation also uncovered a new mediation mechanism that may lessen the negative effects of pregnancy on schizophrenia. Secondly, our study research utilized the extensive and latest GWAS datasets, as well as conducting multiple sensitivity analyses, all of which contributed to the strength of our findings. Thirdly, the MR design reduced the likelihood that confounding variables and reverse causation, which would affect the causal inference.

Our analysis is still subject to several limitations. Firstly, we set the significance threshold of exposure IVs as 2 × 10⁻⁵ in the first-step MR analysis due to insufficient IVs under genome-wide significance, which may include the IVs with low correlation to increase weak instruments bias. However, included IVs were corroborated to have strong correlation with F-statistics > 10. Secondly, limited 16S rRNA sequencing resolution prevented MR studies from being performed at a more precise level, only at the bacterial genus level. As a third point, our study failed to offer additional insights into the mechanisms behind the PH, schizophrenia and 1,4-DHNA, which require further functional research for clarification.

6. Conclusion

In summary, our two-sample MR study causally identified that PH significantly increase the occurrence of offspring schizophrenia, establishing a strong theoretical foundation for the prevention and management of schizophrenia. 1,4-DHNA was demonstrated to have a genetical association with PH and schizophrenia. Further researches to elucidate the underlying mechanisms of PH on schizophrenia are urgently needed to make up for the limitations of this research.

Footnotes

Acknowledgment

This study was supported by the Natural Science Foundation of Chongqing (cstc2021jcyj-msxmX0862).

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the the Natural Science Foundation of Chongqing, (grant number cstc2021jcyj-msxmX0862).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Zheng

Zhang

, et al. The effects of group-based cognitive behavioral therapy in the rehabilitation of patients with chronic schizophrenia with more than two years of community-based mental health group rehabilitation. Technol Health Care 2023; 31: 1911–1922.

Millan

Andrieux

Bartzokis

, et al. Altering the course of schizophrenia: progress and perspectives. Nat Rev Drug Discov 2016; 15: 485–515.

Correll

Solmi

Croatto

, et al. Mortality in people with schizophrenia: a systematic review and meta-analysis of relative risk and aggravating or attenuating factors. World Psychiatry 2022; 21: 248–271.

Weinberger

. Future of days past: neurodevelopment and schizophrenia. Schizophrenia Bull 2017; 43: 1164–1168.

Owen

O'Donovan

. Schizophrenia and the neurodevelopmental continuum:evidence from genomics. World Psychiatry 2017; 16: 227–235.

Davis

Eyre

Jacka

, et al. A review of vulnerability and risks for schizophrenia: beyond the two hit hypothesis. Neurosci Biobehav Rev 2016; 65: 185–194.

Clifton

Hannon

Harwood

, et al. Dynamic expression of genes associated with schizophrenia and bipolar disorder across development. Transl Psychiat 2019; 9: 74.

Powell

. Models of neurodevelopmental abnormalities in schizophrenia. Curr Top Behav Neurosci 2010; 4: 435–481.

Cattane

Richetto

Cattaneo

. Prenatal exposure to environmental insults and enhanced risk of developing Schizophrenia and Autism Spectrum Disorder: focus on biological pathways and epigenetic mechanisms. Neurosci Biobehav Rev 2020; 117: 253–278.

10.

Iannitelli

Quartini

Tirassa

, et al. Schizophrenia and neurogenesis: a stem cell approach. Neurosci Biobehav Rev 2017; 80: 414–442.

11.

Brown

. The environment and susceptibility to schizophrenia. Prog Neurobiol 2011; 93: 23–58.

12.

Robinson

Lahdepuro

Tuovinen

, et al. Maternal hypertensive pregnancy disorders and mental and behavioral disorders in the offspring: a review. Curr Hypertens Rep 2021; 23: 30.

13.

Trubetskoy

Pardinas

, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nature 2022; 604: 502–508.

14.

Dickinson

Zaidman

Giangrande

, et al. Distinct polygenic score profiles in schizophrenia subgroups with different trajectories of cognitive development. Am J Psychiat 2020; 177: 298–307.

15.

Vita

Barlati

Bellomo

, et al. Patterns of care for adolescent with schizophrenia: a delphi-based consensus study. Front Psychiatry 2022; 13: 844098.

16.

Umesawa

Kobashi

. Epidemiology of hypertensive disorders in pregnancy: prevalence, risk factors, predictors and prognosis. Hypertens Res 2017; 40: 213–220.

17.

Vogel

Souza

Mori

, et al. Maternal complications and perinatal mortality: findings of the World Health Organization Multicountry Survey on Maternal and Newborn Health. Bjog-Int J Obstet Gy 2014; 121: 76–88.

18.

Wang

, et al. Hypertensive disorders of pregnancy and risk of cardiovascular disease-related morbidity and mortality: a systematic review and meta-analysis. Cardiology 2020; 145: 633–647.

19.

Mengistu

Turner

Flatley

, et al. Impact of severe maternal morbidity on adverse perinatal outcomes in high-income countries: systematic review and meta-analysis protocol. Bmj Open 2019; 9: e027100.

20.

Webster

Conti-Ramsden

Seed

, et al. Impact of antihypertensive treatment on maternal and perinatal outcomes in pregnancy complicated by chronic hypertension: a systematic review and meta-analysis. J Am Heart Assoc 2017; 6: e005526.

21.

Lahti-Pulkkinen

Girchenko

Tuovinen

, et al. Maternal hypertensive pregnancy disorders and mental disorders in children. Hypertension 2020; 75: 1429–1438.

22.

Dachew

Scott

Betts

, et al. Hypertensive disorders of pregnancy and the risk of offspring depression in childhood: findings from the Avon Longitudinal Study of Parents and Children. Dev Psychopathol 2020; 32: 845–851.

23.

Paquin

Lapierre

Veru

, et al. Early environmental upheaval and the risk for schizophrenia. Annu Rev Clin Psycho 2021; 17: 285–311.

24.

Maher

Dalman

O'Keeffe

, et al. Association between preeclampsia and attention-deficit hyperactivity disorder: a population-based and sibling-matched cohort study. Acta Psychiat Scand 2020; 142: 275–283.

25.

Maher

O'Keeffe

Dalman

, et al. Association between preeclampsia and autism spectrum disorder: a population-based study. J Child Psychol Psyc 2020; 61: 131–139.

26.

Dachew

Scott

Mamun

, et al. Pre-eclampsia and the risk of attention-deficit/hyperactivity disorder in offspring: findings from the ALSPAC birth cohort study. Psychiat Res 2019; 272: 392–397.

27.

Gou

, et al. Landscape of the gut mycobiome dynamics during pregnancy and its relationship with host metabolism and pregnancy health. Gut 2024; 73: 1302–1312.

28.

Zhang

Zhao

, et al. Gut microbiota dysbiosis and increased NLRP3 levels in patients with pregnancy-induced hypertension. Curr Microbiol 2023; 80: 168.

29.

Madison

Kuempel

Albrecht

, et al. 3,3'-Diindolylmethane And 1,4-dihydroxy-2-naphthoic acid prevent chronic mild stress induced depressive-like behaviors in female mice. J Affect Disord 2022; 309: 201–210.

30.

Davies

Holmes

Davey

. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. Bmj-Brit Med J 2018; 362: k601.

31.

Burgess

Scott

Timpson

, et al. Using published data in Mendelian randomization: a blueprint for efficient identification of causal risk factors. Eur J Epidemiol 2015; 30: 543–552.

32.

Burgess

Thompson

. Avoiding bias from weak instruments in Mendelian randomization studies. Int J Epidemiol 2011; 40: 755–764.

33.

Larsson

Butterworth

Burgess

. Mendelian randomization for cardiovascular diseases: principles and applications. Eur Heart J 2023; 44: 4913–4924.

34.

Davey

Holmes

Davies

, et al. Mendel's laws, Mendelian randomization and causal inference in observational data: substantive and nomenclatural issues. Eur J Epidemiol 2020; 35: 99–111.

35.

Carter

Sanderson

Hammerton

, et al. Mendelian randomisation for mediation analysis: current methods and challenges for implementation. Eur J Epidemiol 2021; 36: 465–478.

36.

Park

Lee

Kim

, et al. Atrial fibrillation and kidney function: a bidirectional Mendelian randomization study. Eur Heart J 2021; 42: 2816–2823.

37.

Bottigliengo

Foco

Seibler

, et al. A Mendelian randomization study investigating the causal role of inflammation on Parkinson's disease. Brain 2022; 145: 3444–3453.

38.

Zhao

Holmes

Zheng

, et al. The impact of education inequality on rheumatoid arthritis risk is mediated by smoking and body mass index: Mendelian randomization study. Rheumatology 2022; 61: 2167–2175.

39.

Varbo

Benn

Smith

, et al. Remnant cholesterol, low-density lipoprotein cholesterol, and blood pressure as mediators from obesity to ischemic heart disease. Circ Res 2015; 116: 665–673.

40.

Bowden

Davey

Burgess

. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol 2015; 44: 512–525.

41.

Palmer

Lawlor

Harbord

, et al. Using multiple genetic variants as instrumental variables for modifiable risk factors. Stat Methods Med Res 2012; 21: 223–242.

42.

Davey

Hemani

. Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Hum Mol Genet 2014; 23: R89–R98.

43.

Namazi

Aghasian

Ala

. Fractal-based classification of electroencephalography (EEG) signals in healthy adolescents and adolescents with symptoms of schizophrenia. Technol Health Care 2019; 27: 233–241.

44.

Gudmundsson

Walters

Ingason

, et al. Attention-deficit hyperactivity disorder shares copy number variant risk with schizophrenia and autism spectrum disorder. Transl Psychiat 2019; 9: 258.

45.

Beukers

Aarnoudse-Moens

van Weissenbruch

, et al. Fetal growth restriction with brain sparing: neurocognitive and behavioral outcomes at 12 years of age. J Pediatr-Us 2017; 188: 103–109.e2.

46.

Stampanoni

Iezzi

Gilio

, et al. Synaptic plasticity shapes brain connectivity: implications for network topology. Int J Mol Sci 2019; 20: 6193.

47.

McCutcheon

Reis

Howes

. Schizophrenia-An overview. Jama Psychiat 2020; 77: 201–210.

48.

Turbeville

Sasser

. Preeclampsia beyond pregnancy: long-term consequences for mother and child. Am J Physiol-Renal 2020; 318: F1315–F1326.

49.

Dachew

Mamun

Maravilla

, et al. Association between hypertensive disorders of pregnancy and the development of offspring mental and behavioural problems: a systematic review and meta-analysis. Psychiat Res 2018; 260: 458–467.

50.

Davies

Segre

Estrade

, et al. Prenatal and perinatal risk and protective factors for psychosis: a systematic review and meta-analysis. Lancet Psychiat 2020; 7: 399–410.

51.

Raina

El-Messidi

Badeghiesh

, et al. Pregnancy hypertension and its association with maternal anxiety and mood disorders: a population-based study of 9 million pregnancies. J Affect Disorders 2021; 281: 533–538.

52.

Nikolova

Smith

Hall

, et al. Perturbations in gut microbiota composition in psychiatric disorders: a review and meta-analysis. Jama Psychiat 2021; 78: 1343–1354.

53.

Asif

Dai

, et al. Alteration of the gut microbiome in first-episode drug-naive and chronic medicated schizophrenia correlate with regional brain volumes. J Psychiatr Res 2020; 123: 136–144.

54.

Penninx

Lange

. Metabolic syndrome in psychiatric patients: overview, mechanisms, and implications. Dialogues Clin Neuro 2018; 20: 63–73.

55.

Goltsman

Sun

Proctor

, et al. Metagenomic analysis with strain-level resolution reveals fine-scale variation in the human pregnancy microbiome. Genome Res 2018; 28: 1467–1480.

56.

Doukyu

Fujisawa

Saito

. Improving E. coli organic solvent tolerance by 1,4-dihydroxy-2-naphthoic acid. Biosci Biotech Bioch 2022; 86: 1128–1135.

57.

Chen

Liu

, et al. Gut dysbiosis induces the development of pre-eclampsia through bacterial translocation. Gut 2020; 69: 513–522.

58.

Zhang

Miao

, et al. Dietary nutrition and gut microbiota composition in patients with hypertensive disorders of pregnancy. Front Nutr 2022; 9: 862892.

59.

Hao

Zhang

, et al. A review of antibiotics, depression, and the gut microbiome. Psychiat Res 2020; 284: 112691.

60.

Klein-Petersen

Kohler-Forsberg

Benros

. Infections, antibiotic treatment and the Microbiome in relation to schizophrenia. Schizophr Res 2021; 234: 71–77.

Exposure	Outcome	Number of SNPs	SNP	F Value
PH	1,4-DHNA	4	rs13306561	44.24
			rs167479	46.19
			rs16998073	55.18
			rs236721	48.98
1,4-DHNA	Schizophrenia	26	rs10192555	41.82
			rs10793056	41.72
			rs10950493	41.57
			rs11030125	45.10
			rs11761901	48.89
			rs12156434	38.46
			rs12432473	36.25
			rs139901010	41.07
			rs207254	46.48
			rs2834173	43.02
			rs28433015	37.86
			rs28634397	43.67
			rs34011530	41.37
			rs34278344	39.24
			rs395671	39.05
			rs4585628	41.83
			rs61861701	55.10
			rs62079926	47.64
			rs62465767	41.47
			rs63721662	41.55
			rs6925364	38.80
			rs73227800	38.36
			rs7504312	38.40
			rs76292847	39.24
			rs7691920	43.41
			rs9886448	41.67
			rs9941912	50.53
			rs996888	44.88
PH	Schizophrenia	4	rs12901406	884.80
			rs13306561	1012.22
			rs16998073	1262.57
			rs236721	1120.64

		Heterogeneity			Pleiotropy
Exposure	Outcome	Method	Cochran's Q	P Value	Egger-Intercept	P Value
PH	1,4-DHNA	MRE	1.31	0.51	−0.10	0.45
		IVW	2.17	0.53
1,4-DHNA	Schizophrenia	MRE	25.50	0.37	−0.004	0.59
		IVW	25.80	0.41
PH	Schizophrenia	MRE	1.09	0.57	0.001	0.95
		IVW	1.10	0.77
Schizophrenia	PH	MRE	223.75	0.17	−0.01	0.07
		IVW	227.26	0.14

A Mendelian randomization approach for advancing healthcare innovation: From pregnancy hypertension to schizophrenia

Abstract

Background:

Objective:

Methods:

Results:

Conclusion:

Keywords

1. Introduction

2. Materials and methods

2.1. Study design

Table 1. Details of GWAS data included in MR analyses. Trait Consortium Ethnicity Sample size PH FinnGen Biobank European 123,579 Gut bacteria pathway abundance (1,4-DHNA) UK biobank European 7738 Schizophrenia UK biobank Mixed 320,404

2.5 Sensitivity analysis

3. Results

3.1. Selection of IVs

3.2. Bidirectional two-sample MR analysis

3.2.1 Causal effect of PH on schizophrenia

3.3. Two step MR analysis

5. Strengths and limitations

6. Conclusion

Footnotes

Acknowledgment

Funding

Declaration of conflicting interests

References

Table 1.
Details of GWAS data included in MR analyses.

Trait Consortium Ethnicity Sample size

PH FinnGen Biobank European 123,579

Gut bacteria pathway abundance (1,4-DHNA) UK biobank European 7738

Schizophrenia UK biobank Mixed 320,404