The Relation Between Maternal Schizophrenia and Low Birth Weight Is Modified by Paternal Age

Objective: Paternal characteristics have never been considered in the relation between maternal schizophrenia and adverse pregnancy outcomes. The aim of our study was to consider different paternal ages while investigating the relation between maternal schizophrenia and low birth weight (LBW), using a nationwide population-based dataset.

Method: Our study used data from the 2001 to 2003 Taiwan National Health Insurance Research Dataset and birth certificate registry. A total of 543 394 singleton live births were included. We performed multivariate logistic regression analyses to explore the relation between maternal schizophrenia and the risk of LBW, taking different paternal age groups into account (aged 29 years or younger, 30 to 39 years, and 40 years and older), and after adjusting for other characteristics of infant, mother, and father as well as the difference between the parent's ages.

Results: Mothers with schizophrenia had a higher percentage of LBW infants than mothers who did not (11.8%, compared with 6.8%). For infants whose mothers had schizophrenia, the adjusted odds ratios of LBW were 1.47 (95% CI 1.02 to 2.27, P < 0.05) and 2.80 (95% CI 1.42 to 5.51, P < 0.01) times greater than for infants whose mothers did not have schizophrenia, for paternal age groups of 30 to 39 years and 40 years or older, respectively. However, maternal schizophrenia was not a significant predictor of LBW for infants whose fathers were aged 29 years and younger.

Conclusions: The relation between LBW and maternal schizophrenia is modified by paternal age. More attention should be paid to the interaction of paternal characteristics and maternal psychiatric disorders in producing adverse pregnancy outcomes.

Can J Psychiatry. 2010;55(6):377-385.

Clinical Implications

* Maternal schizophrenia was not a significant predictor for LBW for children born to fathers aged 29 years and younger.

* We found that mothers with schizophrenia were at an increased risk of having LBW infants only if fathers were aged 30 years or older.

* For infants whose mothers did not have schizophrenia, the adjusted odds ratios of LBW was 1.47 (95% CI 1.02 to 2.27, P < 0.05) and 2.80 (95% CI 1.42 to 5.51 , P < 0.01) times greater than infants whose mothers did not have schizophrenia for paternal age groups of 30 to 39 years and 40 years or older, respectively.

Limitations

* Information such as parents' smoking history, substance abuse, alcohol consumption, nutrition, and body mass index are not available through ourdatasets.

* The dataset does not allow us to account for differences in schizophrenia severity among patients.

Key Words: schizophrenia, parental schizophrenia, paternal schizophrenia

Abbreviations used in this article

LBW low birth weight

NHI National Health Insurance

NHIRD NHI Research Dataset

Although women with schizophrenia have long been reported to have an increased risk of adverse pregnancy outcomes,1 no consensus has been reached. McNeil2 observed no difference in adverse birth outcomes for mothers with schizophrenia, compared with unaffected mothers. However, in a meta-analysis of 14 studies conducted before 1990, Sacker et al3 concluded the incidence of excessive LBW risk among neonates to mothers with schizophrenia, compared with mothers in the general population. Recently, a large register-based Danish study4 reported an increased risk of LBW among women with schizophrenia. Nilsson et al5 also confirmed this finding using different inclusion criteria (that is, diagnosis before childbirth and hospitalization during pregnancy) for mothers with schizophrenia and adjusting for possible risk factors (for example, socioeconomic status or smoking during pregnancy). The sources of discrepancy in these studies2'3'6 are likely to include different definitions and severities of schizophrenia and varieties in study designs and sample sizes. In fact, causes contributing to LBW are multifactorial, incorporating factors such as racial or ethnic origin, cigarette smoking, gestational nutrition, and paternal weight and height.7 Previous studies might also be limited by a lack of sufficient consideration of paternal traits.

Indeed, the important role of paternal characteristics (for example, paternal age, height, and birth weight) in pregnancy outcomes has recently been documented.8"10 Among these characteristics, paternal age was observed as a significant predictor of LBW in offspring, irrespective of other maternal or pregnancy-specific factors.8 Biological plausibility was proposed, as the placenta is mainly dependent on the expression of genes of paternal origins." Occurring more often among men of young and advanced age, these potentially detrimental mutations in genes on placentation may contribute to adverse pregnancy outcomes.12 Thus it is possible that paternal age might be involved in the association between maternal schizophrenia and birth outcomes. However, in studies on the relation between maternal schizophrenia and LBW, paternal characteristics have seldom been taken into consideration.

Based on a nationwide, population-based dataset, our study thus aimed to investigate the relation between maternal schizophrenia and LBW, as affected by the father's age. Multivariate logistic regression analyses were performed, with the simultaneous consideration on other characteristics of the infant (for example, sex and parity), the mother (for example, age, education level, and marital status), the father (for example, education), and family monthly income.

Methods

Database

Our study used 2 large-scale national databases. First, we used data from the 2001 to 2003 NHIRD, published by the National Health Research Institute in Taiwan. Taiwan initiated the NHI in 1995. The NHIRD includes all inpatient and ambulatory care claims for over 21 million enrollees, representing around 96% of the island's population. The NHIRD is plausibly one of the most comprehensive nationwide population-based datasets in the world.

The second database used in our study is the birth certificate registry published by Taiwan's Ministry of the Interior. Birth certificates include data on birthdates (for both infants and their parents), gestational week at birth, infant birth weight, sex, parity, place of birth, parental educational level, and maternal marital status.

With assistance from the Department of Health in Taiwan, the parents' and infants' unique personal identification numbers were used to link the NHIRD and birth certificate data. All personal identifiers were encrypted by the Bureau of the NHI before release to the researchers. Confidentiality assurances were addressed by abiding by the data regulations of the Bureau of the NHI. As our study uses secondary data released for public access for research purposes, with all identities removed, our study was exempt from full review by the Internal Review Board.

Study Sample

The study sample was initially comprised of all singleton live births in Taiwan between January 1 , 200 1 , and December 3 1 , 2003 (n = 593 205). We excluded infants whose mothers had any serious mental disorder, other than schizophrenia, identified in the registry of catastrophic illness (n = 835). In Taiwan, patients with serious mental disorders (senile and presenile organic psychotic conditions, transient organic psychotic conditions, other organic psychotic conditions, schizophrenia, affective psychoses, and paranoid states) are issued a catastrophic illness card once their diagnoses have been verified. As copayments for out- and inpatient psychiatric care are waived for catastrophic illness cardholders, most patients with serious mental disorders are likely to have applied for the catastrophic illness card and to be recorded in the registry of catastrophic illness.

We also excluded infants whose fathers had any serious mental disorder identified in the registry of catastrophic illness (� = 1050). In addition, for mothers who had more than one delivery during the study period, we only selected their first delivery to avoid the possible nonindependent observations contributed by the same mother (n = 47 926). Ultimately, 543 394 singleton live births fulfilled our criteria and were included in our study.

Variables of Interest

The key independent variable was dichotomous - whether or not an infant's mother had schizophrenia - while the dependent variable of interest was LBW, defined as a birth weight of less than 2500 g.

Other possible factors contributing to LBW in infants were adjusted for in the regression model. These factors include characteristics of the infant (sex and parity), the mother (age, gestational age, educational level, and marital status), the father (educational level), and family monthly income (including both the mother's and the father's monthly income). Parental ages were defined as each parent's age, in years, at the time of the infant's birth. Parity was grouped into the following categories: 1, 2, 3, or more. Maternal and paternal education levels were categorized into 4 levels: elementary school or lower, junior high school, senior high school, and college or above. Gestational age was selected to capture the effect of preterm birth (less than 37 weeks). Family monthly income was categorized into 4 groups: less thanNT$15 000, NTS 15 000 to NT$30 000, NT$30 001 to NTS50 000, and NT$50 001 or more.

Statistical Analysis

The SAS System statistical package version 8.2 (SAS Institute Inc, Cary, NC) was used to perform the analyses in our study. Pearson chi-square tests were used to examine differences between mothers with and without schizophrenia as they related to characteristics of the infant, mother, and father. Multivariate logistic regression analysis was also performed to explore the relation between maternal schizophrenia and the risk of LBW according to different paternal age groups (29 years and younger, 30 to 39 years, and 40 years and older), after adjusting for the other characteristics of the infant, mother, and father and differences between the parents' ages (that is, the difference between the people within the parental couple). Although maternal and paternal educational level are typically highly correlated with each other in other studies, the collinearity (coefficient = 0.466) between these 2 variables was tolerable in our study. Therefore, we kept both variables in the regression models. In addition, as prior studies have reported, parental age difference is associated with adverse pregnancy outcomes.13 We also added this variable into the regression model. A 2-sided P value of less than 0.05 was considered statistically significant for our study.

Results

Among the total sample of 543 394 singleton live births, 475 (0.087%) infants had mothers diagnosed with schizophrenia. Table 1 describes the distributions of sociodemographic characteristics of infant and parents relating to mothers with and without schizophrenia. It shows that mothers with schizophrenia had a higher percentage of LBW infants than mothers who did not (11.8%, compared with 6.8%). Pearson chi-square tests also show that there were significant differences in infant parity (P = 0.009), paternal age (P < 0.001), paternal educational level (P < 0.001), maternal age (P < 0.00 1 ), maternal educational level (P < 0.00 1 ), gestational age (P = 0.003), marital status (P < 0.001), and family monthly income (P < 0.00 1 ) between mothers with and without schizophrenia. Table 2 describes the distribution characteristics of infant, mother, and father by paternal age group. The relation between LBW and maternal schizophrenia was not significant for the paternal age group aged 29 years and younger. However, this relation became significant for the 30 to 39 years and 40 years and older paternal age groups. Pearson chi-square tests show that there were consistently significant differences in maternal highest educational level, paternal highest educational level, and family income across 3 paternal age groups, when comparing mothers with and without schizophrenia. The adjusted odds ratios of LBW for mothers with and without schizophrenia relating to paternal age are presented in Table 3. Interestingly, the relation between LWB and maternal schizophrenia was increasingly significant with advancing paternal age, after adjusting for the characteristics of infants and parents and the difference between parental ages. For infants whose mothers had schizophrenia, the adjusted odds ratios of LBW was 1.47 (95% CI 1.02 to 2.27, P < 0.05) and 2.80 (95% CI 1.42 to 5.51,P<0.01) times greater than infants whose mothers did not have schizophrenia for paternal age groups of 30 to 39 years and 40 years and older, respectively. Maternal schizophrenia was not a significant predictor for LBW for children born to fathers aged 29 years and younger. Finally, to further examine the effects of parental age among mothers with and without schizophrenia, we found that there was no relation between maternal age (aged 29 years and younger, 30 to 34 years, and 35 years and older) and LBW among mothers with schizophrenia. A null relation between paternal age (aged 29 years and younger, 30 to 39 years, and 40 years and older) and LBW was also observed among mothers without schizophrenia (data not shown in table).

Discussion

To our knowledge, this is the first attempt to compare the risk of LBW among women with and without schizophrenia that takes paternal characteristics into consideration. Briefly, we found that mothers with schizophrenia would be at an increased risk of having LBW infants only if fathers were aged 30 years or older. Thus it is reasonable to assume the relation between LBW and maternal schizophrenia is modified by paternal age. Consequently, the neglect of paternal characteristics in prior studies might be a cause of inconsistent findings.

Certain risk factors for LBW might be more common among mothers with schizophrenia, such as smoking, substance abuse, alcohol consumption, illicit drugs and caffeine, socioeconomic factors, parity and maternal age, nutritional factors, maternal physical illness, and antenatal care.6 As schizophrenia is regarded as involving a high degree of genetic influence and has a strong tendency to be transmitted genetically to offspring,1415 a genetic contribution is also plausibly a component of the relation between maternal schizophrenia and LBW.

In addition, advanced maternal and paternal ages together with other maternal risk behaviours and nonoptimal social circumstances might increase the risks of LBW.7 Indeed, existent literature consistently indicated that advanced maternal age was associated with an excess risk of adverse birth outcomes of LBW,1617 fetal death,18 and preterm delivery. Lower rates of ovulatory cycles among mothers with advanced age might be considered an adaptation that increases risk of an adverse pregnancy outcome.20 In contrast to the persistent attention on maternal factors, more evidence has accumulated regarding direct paternal influence on birth outcomes. 8-10 In recent years, numerous studies have found that advanced paternal age was related to a higher risk of sperm abnormalities and gene mutation,21-22 possibly resulting in adverse pregnancy outcomes such as preeclampsia,23 stillbirth,24 congenital malformations, and schizophrenia in offspring.25-27 One recent study by Reichman and Teitler28 suggested advanced paternal age as an independent risk factor of LWB. It is possible that owing to the sustained nature of spermatogenesis, it is affected by an age-related vulnerability to compromised DNA-protective mechanisms, with results that impede fetal growth.28,29 Potentially detrimental mutations in genes on placentation, more frequently observed among advancing age in men, may also be involved. I2 Taking the above findings into consideration, it is reasonable to assume that the relation between maternal schizophrenia and LBW is modified by paternal age. Indeed, we found that the risk of LBW was dependent on paternal age; specifically, the grathent of LBW risk increased with paternal age, after adjusting for characteristics of the infant, mother, and father.

Kinzler et al13 reported that adverse pregnancy outcomes were associated with parental age gaps. In our study, we found that parental age differences increased for the advanced paternal age group (2.4, 3.8, and 7.8 years for paternal age of 29 years and younger, 30 to 39 years, and 40 years and older groups, respectively). As we have adjusted for age difference in the regression model, the relation between maternal schizophrenia and LBW in the different paternal age groups found in our study were independent of the effects of a parental age gap.

Our study has 3 strengths. First, we used nationwide, 3-year, population-based datasets linking the NHIRD with birth certificates. Therefore, its robust findings can be generalized to the population as a whole. Second, the large size of our study allowed our analyses to be stratified by paternal age groups while maintaining a sample size that provided ample statistical power to detect differences between mothers with and without schizophrenia, after adjusting for confounding variables. Third, in contrast to other national data, in which a substantial portion of infants are lacking registered data about their fathers,30 only 0.3% of infants in our sample were missing fatherhood data, because registration of all birth information is mandatory and there is a very low rate (about 3% in 2001) of single motherhood in Taiwan.31

There are a couple of limitations to our study. First of all, information such as parents' smoking history, substance abuse, alcohol consumption, nutrition, and body mass index are not available through our datasets. These factors may increase the risk of LBW. Second, we identify mothers diagnosed with schizophrenia by the International Classification of Diseases, Ninth Revision, Clinical Modifications, codes from the registry of catastrophic illness released by the Bureau of the NHI, and this dataset does not allow us to account for differences in schizophrenia severity among patients. In addition, some people might suffer from schizophrenia without possessing a catastrophic illness card. Probably owing to the stigma associated with psychiatric diseases or an unawareness of their illness, they were neither diagnosed nor given treatment for this mental illness.

As far as we know, this is the first study to document that the relation between maternal schizophrenia and LBW is modified by paternal age. Although the mechanism underlying the relation and the role of paternal age as a modifier remain unclear, they should be topics for future study. At a minimum, efforts should be made to protect pregnant women with schizophrenia whose partners are aged 30 years or older from the increased risk of LBW through more active monitoring and early intervention. Our findings also suggest that more attention needs to be paid to the interactive effect of paternal characteristics and maternal psychiatric disorders on adverse pregnancy outcomes.

Acknowledgements

This study is based, in part, on data from the National Health Insurance Research Dataset provided by the Bureau of NHI, Department of Health, Taiwan, and managed by the National Health Research Institutes. The interpretations and conclusions contained herein do not represent those of the Bureau of NHI, Department of Health, or the National Health Research Institutes.

No funding or support was received for this study.

[Sidebar]

R�sum� : La relation entre la schizophr�nie maternelle et l'insuffisance de poids � la naissance est modifi�e par l'�ge paternel

Objectif : Les caract�ristiques paternelles n'ont jamais �t� prises en compte dans la relation entre la schizophr�nie maternelle et les issues de grossesse d�favorables. Notre �tude visait � consid�rer diff�rents �ges paternels tout en explorant la relation entre la schizophr�nie maternelle et l'insuffisance de poids � la naissance (IPN), � l'aide d'un ensemble de donn�es bas�es dans la population nationale de Taiwan.

M�thode : Notre �tude a utilis� des donn�es de 2001 � 2003 du Taiwan National Health Insurance Research Dataset ainsi que du registre des certificats de naissance. Au total, 543 394 naissances vivantes simples ont �t� incluses. Nous avons effectu� des analyses de r�gression logistique multivari�e pour explorer la relation entre la schizophr�nie maternelle et le risque d'IPN, en prenant en compte diff�rents groupes d'�ge paternel (29 ans ou moins, 30 � 39 ans, et 40 ans et plus), et apr�s ajustement pour d'autres caract�ristiques du b�b�, de la m�re, et du p�re ainsi que pour la diff�rence d'�ge entre les parents.

R�sultats : Les m�res souffrant de schizophr�nie avaient un pourcentage plus �lev� de b�b�s ayant une IPN que les m�res qui n'en souffraient pas (1 1 ,8 %, compar� � 6,8 %). Pour les b�b�s dont les m�res souffraient de schizophr�nie, les rapports de cotes ajust�s d'IPN �taient 1 ,47 (IC � 95 % 1 ,02 � 2,27; P < 0,05) et 2,80 (IC � 95 % 1 ,42 � 5,51 ; P < 0,01) fois plus grands que pour les b�b�s dont les m�res ne souffraient pas de schizophr�nie, pour les groupes d'�ge paternel de 30 � 39 ans et de 40 ans et plus, respectivement. Cependant, la schizophr�nie maternelle n'�tait pas un pr�dicteur significatif d'IPN pour les b�b�s dont les p�res �taient �g�s de 29 ans et moins.

Conclusions : La relation entre l'IPN et la schizophr�nie maternelle est modifi�e par l'�ge paternel. Il faut se pencher davantage sur l'interaction des caract�ristiques paternelles et des troubles psychiatriques maternels face au risque de grossesses d�favorables.

[Reference]

References

1. Rieder RO, Rosenthal D, Wender P, et al. The offspring of schizophrenics. Fetal and neonatal deaths. Arch Gen psychiatry. 1975;32(2):200-211.

2. McNeil TF. Obstetric complications in schizophrenic parents. Schizophr Res. 1991;5(2):89-101.

3. Sacker A, Done DJ, Crow TJ. Obstetric complications in children born to parents with schizophrenia: a meta-analysis of case-control studies. Psychol Med. 1996;26(2):279-287.

4. Bennedsen BE, Mortensen PB, Olesen AV, et al. Preterm birth and intra-uterine growth retardation among children of women with schizophrenia. Br J Psychiatry. 1999;175:239-245.

5. Nilsson E, Lichtenstein P, Cnattingius S, et al. Women with schizophrenia: pregnancy outcome and infant death among their offspring. Schizophr Res. 2002;58(2-3):221-229.

6. Bennedsen BE. Adverse pregnancy outcome in schizophrenic women: occurrence and risk factors. Schizophr Res. 1998;33(1-2):1-26.

7. Kramer MS. Determinants of low birth weight: methodological assessment and meta-analysis. Bull World Health Organ. 1987;65(5):663-737.

8. Chen XK, Wen SW, Krewski D, et al. Paternal age and adverse birth outcomes: teenager or 40+, who is at risk? Hum Reprod. 2008;23(6):1290-1296.

9. Magnus P, Gjessing HK, Skrondal A, et al. Paternal contribution to birth weight. J Epidemiol Community Health. 2001;55(12):873-877.

10. Nahum GG, Stanislaw H. Relationship of paternal factors to birth weight. J Reprod Med. 2003;48(12):963-968.

11. Miozzo M, Simoni G. The role of imprinted genes in fetal growth. Biol Neonate. 2002;81(4):217-228.

12. Slama R, Bouyer J, Windham G, et al. Influence of paternal age on the risk of spontaneous abortion. Am J Epidemiol. 2005;161(9):816-823.

13. Kinzler WL, Ananth CV, Smulian JC, et al. Parental age difference and adverse perinatal outcomes in the United States. Paediatr Perinat Epidemiol. 2002;16(4):320-327.

14. Avila M, Thaker G, Adami H. Genetic epidemiology and schizophrenia: a study of reproductive fitness. Schizophr Res. 2001;47(2-3):233-241.

15. Cannon TD, Kaprio J, Lonnqvist J, et al. The genetic epidemiology of schizophrenia in a Finnish twin cohort. A population-based modeling study. Arch Gen Psychiatry. 1998;55(1):67-74.

16. Nahum GG. Stanislaw H. Validation of a birth weight prediction equation based on maternal characteristics. J Reprod Med. 2002;47(9):752-760.

17. Khoshnood B, Wall S, Lee KS. Risk of low birth weight associated with advanced maternal age among four ethnic groups in the United States. Matern Child Health J. 2005;9(1):3-9.

18. Fretts RC, Usher RH. Causes of fetal death in women of advanced maternal age. Obstet Gynecol. 1997;89(1):40-45.

19. Astolfi P, Zonta LA. Risks of preterm delivery and association with maternal age, birth order, and fetal gender. Hum Reprod. 1999; 14(11):2891-2894.

20. Kirchengast S, Hartmann B. Advanced maternal age is not only associated with newborn somatometrics but also with the mode of delivery. Ann Hum Biol. 2003;30(1):1-12.

21. Glaser RL, Broman KW, Schulman RL, et al. The paternal-age effect in Apert syndrome is due, in part, to the increased frequency of mutations in sperm. Am J Hum Genet. 2003;73(4):939-947.

22. Goriely A, McVean GA, Rojmyr M, et al. Evidence for selective advantage of pathogenic FGFR2 mutations in the male germ line. Science. 2003;301(5633):64 3-646.

23. Harlap S, Paltiel O, Deutsch L, et al. Paternal age and preeclampsia. Epidemiology. 2002;13(6):660-667.

24. Astolfi P, De Pasquale A, Zonta LA. Late paternity and stillbirth risk. Hum Reprod. 2004:19(11):2497-2501.

25. Archer NP, Langlois PH, Suarez L, et al. Association of paternal age with prevalence of selected birth defects. Birth Defects Res. 2007;79(1):27-34.

26. Sipos A, Rasmussen F, Harrison G, et al. Paternal age and schizophrenia: a population based cohort study. BMJ. 2004;329(7474):1070.

27. Yang Q, Wen SW, Leader A, et al. Paternal age and birth defects: how strong is the association? Hum Reprod. 2007;22(3):696-701.

28. Reichman NE, Teitler JO. Paternal age as a risk factor for low birthweight. Am J Public Health. 2006;96(5):862-866.

29. Walter CA, Walter RB, McCarrey JR. Gennline genomes - a biological fountain of youth? Sci Aging Knowledge Environ. 2003;2003(8):PE4.

30. Webb RT, Pickles AR. King-Hele SA, et al. Parental mental illness and fatal birth defects in a national birth cohort. Psychol Med. 2008;38(10):1495-1503.

31. Lin HC, Tang CH, Lee HC. Association between paternal schizophrenia and low birthweight: a nationwide population-based study. Schizophr Bull. 2009;35(3):624-630. Epub 2008 Jul 15.

[Author Affiliation]

Herng-Ching Lin, PhD1; Hsin-Chien Lee, MD, MPH2; Chao-Hsiun Tang, PhD1; Yi-Hua Chen, PhD3

[Author Affiliation]

Manuscript received April 2009, revised, and accepted October 2009.

1 Professor, School of Health Care Administration, Taipei Medical

University, Taipei. Taiwan.

2 Director, Department of Psychiatry, Taipei Medical University-Shuang

Ho Hospital, Taipei, Taiwan; Professor, Department of Psychiatry,

Taipei Medical University, School of Medicine, Taipei, Taiwan.

3 Associate Professor, Taipei Medical University, School of Public

Health, Taipei, Taiwan.

Address for correspondence: Dr H-C Lin, School of Health Care

Administration, Taipei Medical University, 250 Wu-Hsing St, Taipei 110,

Taiwan; henry11111@tmu.edu.fw