Schizophrenia: Secondary analysis of placental growth measures: modeling placental measures with intermediates for poor neurodevelopmental outcome

Contact name: Carolyn M. Salafia MD MS

Phone: (914)-834-2598

Fax: (914)-833-1699

Email: salafiacm@aol.com,cs536@columbia.edu

Disorder: Schizophrenia

Geographic location: New York, New York

Other keywords: schizophrenia, placental, fetal origins, birthweight

Participant compensation: None

Time commitment required: 2 years

Drug trial: No

Recruitment end date (approximate): 2006

Study criteria (i.e., age or gender requirements):
We will use the Prenatal Determinants of Schizophrenia study cohort to explore hypotheses relating to placental growth and development to schizophrenia. The Prenatal Determinants of Schizophrenia Cohort derives from the Child Health and Development Study (CHDS) (1). The goal of the CHDS was to examine influences on prenatal and childhood health and development. Initiated in 1959, the CHDS captured approximately 20,000 pregnancies among Kaiser Foundation Health Plan members in Oakland California, a racially and ethnically diverse group of women. The PDS Study sample consists of the 12,094 CHDS offspring who were Health Plan members at any time during the follow-up study period (December 31, 1980 to December 31, 1997). Cases of schizophrenia and schizophrenia spectrum disorders (SSD) were ascertained among the 12,094 members present in the follow-up study period. Potential cases were identified among the PDS cohort members through multi-stage screening. of several health plan registries (the hospitalizations registry, the outpatient treatment registry, the pharmacy registry). All potential cases (N=183) were pursued for diagnostic interview. Consensus diagnoses were rendered by three psychiatrist researchers after discussion with interviewer and review of DIGS and medical records. Those who could not be interviewed, or refused to be interviewed were diagnosed based on detailed chart review. DSM-IV criteria were applied. 71 cases of schizophrenia spectrum disorders were ultimately diagnosed: 43 schizophrenia, 17 schizoaffective, 1 delusional disorder, 5 schizotypal personality disorder, and 5 “other schizophrenia spectrum psychosis”. Case Definition. In this study schizophrenia is defined as DSM-IV schizophrenia (295.1, .2, .3, .6, .9),and schizophrenia spectrum disorders including DSM-IV diagnoses of schizophrenia (295.1, .2, .3, .6, .9), schizoaffective disorder (295.7), other non-affective psychoses (schizophreniform disorder (295.4), delusional disorder(297.1), and psychotic disorder NOS (298.9)), and schizotypal personality disorder (301.22)). Placental Measures. Placentas were assessed in the cohort according to the Benirschke protocol in 78% of the 12094 followed in the PDS, and 79% of the cases.

Exclusion criteria:

Brief description of study:
Evidence suggests that the risk of developing schizophrenia is influenced by intrauterine life events. The placenta provides a rich, and complex view of prenatal experience: the placenta is both a key determinant of the intrauterine environment, and a reflection of intrauterine life events. Here we plan to investigate the relationship between placental growth and development and risk for schizophrenia examining two major hypotheses:
1. Abnormal placental growth and development, a key determiner of fetal oxygen and nutrient availability, will be directly related to risk of schizophrenia.
2. Greater degrees of abnormal placental growth and development (the more chronic the poor maternal supply of oxygen and nutrients) will have progressively stronger associations with schizophrenia.

Additional information:
Significance. In studies of schizophrenia, the examination of intrauterine life has not fully utilized available information on the placenta. Because of its central role in fetal experience, a more complete articulation of placenta-mediated processes will significantly advance our understanding of the association of prenatal experience and schizophrenia outcomes.
To date, we have explored the relationships of standard placental measures to fetal growth, as follows:
1. Does the fetoplacental weight ratio (FPR, the ratio of the grams of birthweight supported by each gram of placental weight) vary by pattern of placental growth? Answer: Yes, both in the CHDS and the Collaborative Perinatal Project (CPP).
Significance: If the pattern of placental growth is associated with FPR differences, this has both physiologic and functional implications:
a. Placental growth measures capture aspects of placental function. The larger and smaller diameters of the chorionic disk, the disk shape, and the distance from the cord insertion to the nearest disk edge each relate to the growth of the chorionic plate. The laterally expanding growth of the chorionic disk serves 2 functions. First, the umbilical- chorionic vessels bear the burden of rapid transfer of large volumes of fetal blood to and from the villus capillary bed. Disk shape, distance from cord insertion to margin and the larger and smaller chorionic disk diameters are each measures that capture aspects of this high capacitance/low resistance flow system. Second, the chorionic disk area measures the area of the uterine lining covered by the placenta and in effect, how many maternal spiral arteries are potential suppliers of the placenta. The disk thickness, by contrast, marks the extent of arborization of the villus capillary bed, the actual locus of maternal fetal exchange. Fetal stem arterioles are also principal sites of placental vascular resistance, thus they contribute to total fetal peripheral resistance and fetal heart work.
b. Placental growth measures may have different “critical periods of determination”. Chorionic disk expansion and increasing disk thickness are also conventionally thought to have their peak growth or determination at different broad times in gestation.. For example, placental shape is considered to be determined early in pregnancy. Up to eight weeks gestation, chorionic villi cover the entire chorionic sac. Villous atrophy forms the future extraplacental membranes; this is effectively complete by 13 weeks gestation. Failure of compete atrophy over the extraplacental membranes results in accessory placental lobes, and abnormal placental shape. However, abnormal shape can be caused by later disruption of a normally growing (and roud/oval) placenta. Benirschke and Kaufmann acknowledge that placental infarct may also secondarily convert a round placenta into one with an irregular outline or even one with an extra lobe, if death of a portion of placenta isolates a viable piece from the main body of the placenta.
2. Does the relative eccentricity of the umbilical cord insertion on the chorionic plate (a measure of asymmetry or non-uniformity of placental expansion about the umbilical cord and thus a marker of a problematic intrauterine environment) affect BW after adjustment for placental weight?
Answer: Yes, both in the CHDS and the CPP.
Significance: It is clinically well appreciated that extreme abnormal cord insertions (marginal and velamentous) have correlations with abnormal outcomes. Many times this is explained on the basis of increased mechanical vulnerability of the marginal or velamentous cord. However, our findings point to a more profound biological truth; even cord insertions closer to the margin—relative to their chorionic plate area—are related to reduced BW. The cord insertion site reflects the confluence of the large chorionic vessels. Their radial organization would reasonably, by basic hemodynamic principles, be expected to alter the placental vascular resistance “perceived” by the fetal heart and, over gestation, alter the efficacy of the translation of placental growth into efficient nutrient and oxygen transport so as to result in fetal growth. Placental size would develop contingent upon the maternal uteroplacental environment. The cord insertion would “translate” this growth into meaningful terms that would permit fetal growth. Placental growth would likely be “reactive”, responsive to a local poor intrauterine environment by limiting growth in one area, and growly lushly in regions in which the maternal environment permitted lush growth. This is the basis of the trophotropism theory of Benirschke. In this cohort the fact that an eccentric insertion did not effect the placental weight suggests the placenta had compensated for this to a large extent, although the fetal weight was still reduced.
Again, we demonstrate subtle functional influences of placental growth patterns on BW. Cord insertion site is generally considered to be relatively stable in the late third trimester. Thus this also marks influences that may modify fetal homeostasis remote to parturition.
3. Does fetal fatness, the ratio of birthweight to birth length (ponderal index or PI) depend more on the lateral expansion of the chorionic plate area or on the arborization of the nutrient exchange surface, as reflected in disk thickness?
Answer: Yes, the ponderal index varies directly with chorionic plate area and is independent of disk thickness.
Significance: Late in the third trimester, fetal metabolism shifts from production of new cells to the laying down of fat. Appropriate deposition of fat requires optimum placental function. In our original linear model, we could not demonstrate an independent effect of thickness on BW after adjustment for the other placental measures; an interaction between placental thickness and placental weight was negative, suggesting that increased thickness for a given placental mass was associated with a lesser BW than expected. However, this interaction, while statistically “significant”, contributed little to the predictive power of the placental growth-BW model. Here we show that disk thickness, the dimension containing the arborizing villous tree, the actual site of nutrient and oxygen exchange, does not affect ponderal index, a measure of fetal weight gain in the third trimester. The expansion of the chorionic plate, beginning early in pregnancy, is the principal determiner of the ability of the term placenta to translate its mass and its function efficiently into BW.
4. In terms of placental growth and its translation to BW, “is bigger always better”?
Answer: No, not in terms of umbilical cord length, larger and smaller placental disk diameters (as surrogates for chorionic plate area), and disk thickness.
We elected to explore the placental growth-BW in the MARS modeling framework because it both allows local (nonparametric) solutions (that may be germane to individualized trajectories of intrauterine growth), and automates model specification(freeing the model from preconceptions). The local solutions, or splines, are generated by piece-wise linear regression by search procedures within the software. Knots mark the end of one region of data with one optimal local solution and the beginning of another region with its own local optimal solution. Knots identify critical values above and below which the predictors-outcome relationship changes. The only constraint is that splines be continuous between regions at the knots (so-called “smoothing splines”). MARS is robust to outliers; outliers are able to be segregated within their own region, restricting their impact on predictor-variable solutions within the remainder of the data. If the best fit is a linear one, there will be no knots identified within the variable range. MARS works by first developing a clearly overfit model, with too many knots, and then, using appropriate statistical criteria, removing knots that contribute least to overall fit. Our results (presented for umbilical cord length, and similar for placental disk diameters and thickness)
Significance: Our goal is to identify measures or aspects of measures that may may infants at risk for long term neurodevelopmental and/or neuropsychiatric poor outcome. We hypothesize (and will test same) that circumstances in which placental growth is “favored” over fetal growth (manifested as birth weight) are “not physiologic”, and will mark infants at risk for long term neurodevelopmental poor outcome.