The Preconception Period Analysis of Risks and Exposures Influencing Health and Development (PrePARED) Consortium

Abstract

Background:

Preconception health may have intergenerational influences. We have formed the PrePARED (Preconception Period Analysis of Risks and Exposures influencing health and Development) research consortium to address methodological, conceptual, and generalisability gaps in the literature.

Objectives:

The consortium will investigate the effects of preconception exposures on four sets of outcomes: 1) fertility and miscarriage; 2) pregnancy-related conditions; and 3) perinatal and child health, and 4) adult health outcomes.

Population:

A study is eligible if it has data measured for at least one preconception time point, has a minimum of selected core data, and is open to collaboration and data harmonization.

Design:

The included studies are a mix of studies following women or couples intending to conceive, general-health cohorts that cover the reproductive years, and pregnancy/child cohort studies that have been linked with preconception data. The majority of the participating studies are prospective cohorts, but a few are clinical trials or record linkages.

Methods:

Data analysis will begin with harmonization of data collected across cohorts. Initial areas of interest include nutrition and obesity; tobacco, marijuana, and other substance use; and cardiovascular risk factors.

Preliminary results:

23 cohorts with data on almost 200,000 women have combined to form this consortium, begun in 2018. Twelve studies are of women or couples actively planning pregnancy, 6 are general-population cohorts that cover the reproductive years; the remainder have some other design. The primary focus for four was cardiovascular health, eight was fertility, one was environmental exposures, three was child health, and the remainder general women’s health. Among other cohorts assessed for inclusion, the most common reason for ineligibility was lack of prospectively collected preconception data.

Conclusions:

The consortium will serve as a resource for research in many subject areas related to preconception health, with implications for science, practice, and policy.

Social media quote

The PrePARED consortium combines data on almost 200,000 women from over 20 studies of perinatal health to study the preconception period.

Background

In recent years, there has been growing interest in the investigation of preconceptional exposures as determinants of long-term health outcomes for mothers and their children.^1–4 Preconception health may have intergenerational influences on human health.^5–7 According to the U.S. Centers for Disease Control and Prevention, “Eliminating disparities in preconception health can potentially reduce disparities in two of the leading causes of death [for women] in early and middle adulthood (i.e., heart disease and diabetes).”⁸ In the U.S., Black and Hispanic women are at higher risk for adverse preconception exposures (e.g., poor nutrition, cigarette smoking, racial discrimination, depression/anxiety/stress, environmental chemicals) adverse health conditions.⁹ and limited access to care.^10–12

The preconception period is also important for understanding and improving infant and child health outcomes. A conference on the Louisiana Birth Outcomes Initiative addressed preconception health, emphasizing that “[P]renatal care may be too little too late. … If we want to do something about preventing [adverse birth outcomes] in this country, we really need to start taking care of women’s health long before they get pregnant.”¹³

“Preconception” can be construed narrowly, for instance, as the time period when pregnancy is planned, or broadly, as any point in the life span prior to pregnancy.^{4, 14, 15} Exposures from both time periods have been found to affect pregnancy and infant health (e.g., folic acid supplementation¹⁶ or adverse childhood experiences¹⁷). Many adverse health behaviors, including smoking, sedentary lifestyle, and poor diet, are apparent preconceptionally, meaning that pregnancy begins under less than optimal conditions and the stage is set for long-term chronic diseases.⁴ In addition, socioeconomic status and exposure to stress and trauma over the life course influence both biology and behavior, and may be crucial to understanding women’s experience of pregnancy and fertility decisions. Thus, the preconception period is an important time window for health interventions that could positively shape women’s and offspring health. Although the traditional focus for preconception research has been women, such interventions should also target men, given not only their direct biological influences on fertility¹⁴ but also their role on child health and development¹⁸ and family well-being.¹⁹

To examine the impact or preconception exposures in women, men, and their children, large-scale analysis may be required. While the importance of some preconception factors has been established, the effect of timing and duration of many exposures (e.g., environmental toxicants) have yet to be investigated in humans, and may have small effect sizes. Rare exposures also require a large sample size to identify associations, to assess potential for etiologic heterogeneity and disease subtypes, and to allow for meaningful subgroup analysis (e.g., by race and ethnicity). Single cohorts often lack the precision to address less common complications and effect modification by factors such as age and parity. Finally, large datasets allow for more fine-grained stratification which is useful to adequately adjust for confounders.

Assessing mechanisms of effect is another important goal. Preconception studies can not only provide insight into child health and developmental programming but may also help to predict risk of pregnancy complications and chronic conditions. Women with chronic hypertension who become pregnant are at higher risk of preterm birth, small-for-gestational-age, stillbirth, in-hospital mortality, and severe forms of hypertensive diseases of pregnancy,^20–22 and some studies indicate that preconception cardiovascular health outside clinical disease also has an impact on pregnancy health.^23–25 Women with low HDL-cholesterol, impaired fasting glucose, or high fasting insulin prior to conception have a two to four-fold higher odds of developing gestational diabetes independent of pre-pregnancy BMI.^26–28 Similarly, preconception depression often carries through pregnancy and postpartum.²⁹ Preconception risk factors could lead to poorer birth outcomes by affecting placentation in the first trimester,^{30, 31} increasing inflammation,³² or producing epigenetic changes that carry into pregnancy.^{33, 34} Often studies lack information about the time period prior to pregnancy that is necessary to clearly place pregnancy health within the life course, or lack sufficient follow-up time post-pregnancy to estimate associations with later-life health (Figure 1). Few studies have information on both preconception and during-pregnancy exposures that allow for distinguishing risks between these time periods.^{35, 36} Pregnancy cohorts usually have limited, retrospectively-collected information on a few exposures and diagnosed preconception conditions, if any information at all.

Figure 1.

Various relationships between preconception exposures, pregnancy health, and post-pregnancy health.

Studies of preconception health may be prone to selection and recall biases. Those that recruit pregnant women and measure data retrospectively produce an achieved-pregnancy bias, if there is a common cause for both the outcome under study and fertility or miscarriage.^{37, 38} These biases may be exacerbated in birth cohort studies that recruit mothers and infants after delivery. Studies including only the most recent pregnancy may be confounded by factors that predict both miscarriage and use of effective contraception,³⁹ while pregnancy/birth cohort studies that rely on retrospectively obtained self-reported data (e.g., preconception nutrition) may be susceptible to recall bias and left truncation.⁴⁰ For these reasons, prospective studies of preconception health are of particular value. However, studies of couples attempting to conceive may be susceptible to selection bias, as they are likely to over- or under-recruit fertile couples, if couples are excluded because they seek fertility treatment,⁴¹ or are included after a period of failing to conceive (even if this is not an explicit criterion for study entry).⁴² Moreover, targeted preconception recruitment usually excludes unintended pregnancies, which are a non-negligible proportion of pregnancies in the United States.⁴³ To capture preconception health of unintended pregnancies, ongoing cohort studies aimed at chronic health conditions can be used or a population of women of reproductive age and potential can be followed prospectively for changes in risk of pregnancy and actual occurrence of pregnancy (if resources allow).⁴⁴ However, cohort studies of cardiovascular, metabolic, or other general health tend to focus on older populations (at higher risk for most conditions) and frequently either exclude pregnant women, or, in the case of younger populations, assess health status outside pregnancy only. Studies assessing preconception health care are often geared towards women with clinical conditions or who are at otherwise high risk.^45–47

For these reasons, we have formed a research consortium of studies (PrePARED, Preconception Period Analysis of Risks and Exposures influencing health and Development) addressing preconception health. The goal of the consortium is to use the large number of studies with preconception data to: 1) identify meaningful (at a clinical or population level) effect sizes of preconception exposures; 2) determine the relative importance of pre-, during-, inter-, and post-pregnancy effects; 3) distinguish selection effects due to differential fertility and pregnancy loss; 4) identify plausible mechanisms of effect; 5) examine rare-prevalence exposures and outcomes; and 6) examine effect modification in associations between preconception exposures and health outcomes across subgroups of interest (e.g., age, race/ethnicity, socioeconomic status).

Methods

Overview, structure, and operations

PrePARED is a consortium of studies incorporating information on preconception and pregnancy health, both studies of couples actively planning pregnancy and more general studies that cover the reproductive years. Participating studies are ongoing and completed studies from around the world. The organizing committee is formed of investigators contributing data.

Eligibility criteria

A study is eligible for inclusion in the consortium if it meets three criteria: 1) The study has data measured for at least one preconception time point (i.e., preconception exposures should not be entirely retrospectively reported.) In other words, studies where the only information on the preconception period is what the woman reports during pregnancy are not eligible. 2) A minimum set of data (listed in table 1) is available. This set of data was chosen as necessary for analysis of almost any pregnancy-related outcome (fertility, miscarriage, pregnancy complications, birth outcomes), for comparing across cohorts, and for assessing the definition of the preconception period for a given analysis. 3) The study is open to collaboration and data harmonization. All studies have information beyond the minimum, and can opt in or out of a given analysis. Preconception subcohorts of larger pregnancy or general-population studies are eligible.

Table 1.

Minimum data required for consortium participation

parity

gravidity

pregnancy outcome(s)

age at pregnanc(ies)

measure of socioeconomic status

pre-pregnancy BMI

time between measure and pregnancy

calendar year of measure

Procedure for adding new cohorts

To date, cohorts have been selected through word of mouth, literature searches, and web searches. Additional cohorts may join this open consortium, which will serve as a rich resource for identifying preconception factors with perinatal and potentially intergenerational long-term influences. Interested cohorts interested should contact the corresponding authors for more information.

Ethical issues, data sharing, and security

No analysis will be conducted that is not permissible under the studies’ and institutions’ informed consent or data use policies. Data harmonization consists of two phases: 1) Developing and maintaining a database at each project site (largely already done by the individual studies); and 2) Establishing a data repository to integrate and share data sources across study sites. Data capture and management will be centralized at Tulane’s Global Research Data Center. While the preference is for (de-identified, limited) data to be centrally stored and analyzed, this may not be possible for all topics or all studies, so procedures will be flexible to the exigencies of a given case. Data sharing agreements will be required for initial participation, and the principal and lead investigators from each study will approve all other requests for reports or datasets before dissemination. When consistent with the studies’ data use agreements, data will be released after publication of the primary results. Metadata templates will be designed and coding guides will be developed for all study variables.

Human subject data will be uploaded onto secured computers only, with no unique identifiers attached. All study personnel will be required to use password-protected computers to access the project data, and permission will be required for any data reuse. Data will be regularly backed up to a secure server and made available only to key project personnel, and analysis will generally be performed on this central server. Digital signatures and encryption will prevent unauthorized changes to datasets or documents as well as identifying who created the document and verifying time and date of any change.

This description of study design and measures is not human subjects research and does not require Institutional Review Board review.

Results

Study populations

A description of each cohort is provided in Tables 2 and 3. The consortium brings together 23 studies with data on almost 200,000 women. Eighteen of the participating studies are based in the U.S., with 2 in China, and 1 each in Canada, Australia, and Denmark. All include data collected during the reproductive years (approximately ages 15–44), but some have followed the participants before or after. Studies have also enrolled over 6500 male partners. 26,000 children were enrolled as part of a mother-child dyad; two studies (Bogalusa Heart Study and GUTS) also include children (n=28603) of the original study participants. Twelve studies are of women or couples actively planning pregnancy, 6 are general-population cohorts that cover the reproductive years. and the remainder have some other design.

Table 2.

Participant life stages represented in PrePARED consortium studies

Name	Reproductive-aged women (women enrolled regardless of pregnancy planning status)	Fecundability/fertility (follows women intending to get pregnant)	Paternal (couples/fathers enrolled, not just reported on)	Mother-child (information collected on preconception and pregnancy of mother, as well as on child)
Australian Longitudinal Study on Women’s Health (ALSWH)50, 51	X	X		X
Bogalusa Heart Study	X			subset
Coronary Artery Risk Development in Young Adults (CARDIA)52	X
Canadian Assisted Reproductive Technology Cohort Study (CARTS)53		X	X	X
Chinese Preconception Cohort Study (CPCS)		X	X	X
CoLab Prepregnancy and Early Pregnancy Consortium54	X	X	X	X
Effects of Aspirin in Gestation and Reproduction (EAGeR)55		X
Early Pregnancy Study (EPS)56		X
Hispanic Community Health Study/Study of Latinos (HCHS/SOL)57	X			subset
Home Observation of Periconceptional Exposures (HOPE) Study58		X	X
Longitudinal Investigation of Fertility and the Environment (LIFE)59		X	X
NHLBI Growth and Health Study (NGHS)60	X
Nurses Health Study 3/Growing Up and Today Study (GUTS)61	X	subset	subset	X
Preconceptional Offspring Trajectory Study (PLOTS)		X	X	X
Pregnancy Study Online (PRESTO)62		X	X	X
Retrospective cohort study of assisted reproductive technology53		X
Shanghai Birth Cohort63		X	X	X
Snart Gravid/Snart Foraeldre64		X	X	X
Strong Healthy Women and SMART Strong Healthy Women65	X	X
Time to Conceive		X		Tentative
Time to Pregnancy in Couples of Proven Fecundity66		X
University of Iowa Women’s Health Tissue Repository67	X	X	X	X
Upstate KIDS59				X

Table 3.

Study populations, the PrePARED Consortium

STUDY	AGE RANGE	STUDY FOCUS	STUDY POPULATION DESCRIPTION	N
Reproductive-aged women
Australian Longitudinal Study on Women’s Health	18–27 for the 1989–95 cohort (baseline in 2012/13) 18–42 for the 1973–78 cohort (baseline in 1996)	General women’s health	Representative sample of Australian women	21,032
Bogalusa Heart Study	3–80 (when studied) 1st gen 30–70, 2nd gen 12–45 (now)	Cardiovascular	Black and white, semi-rural men and women	4049 women, 3500 with pregnancy data, 917 generation 2
CARDIA	Age 18–30 years at baseline (1985–1986), 48–60 years in 2015–2016 Multi-center USA	Cardiovascular	50% Black and 50% White men and women, multiple exams every 2 to 5 years during the 30 years of follow up;	2787 women enrolled; gravidity and parity at exams; (1,392 women who had 2,510 births after baseline during 30-y follow up; 1,951 women gave birth before and/or after baseline)
CoLab Prepregnancy and Early Pregnancy Consortium	> 18 y/o (Adults) and associated children	General Women’s Health with Pregnancy Focus	National and international collaboration to share biosamples and clinical data in the peripregnancy time frame.	13478 to date (enrollment ongoing)
Growing Up Today Study	7–16 (2 enrollment waves 1996 and 2004) now 20–36	Child health	Children of women in Nurses’ Health Study II	27, 706 (13,000 in active follow-up)
Hispanic Community Health Study/Study of Latinos	18–74 baseline (2008–2011), 24–80 (2014–2017), 3rd 2020–2023	Cardiovascular	Largest Hispanic health study in the US	16,415; 9835 women, 3801 aged 18–41; 440 mom-child dyads (ancillary; 2019–2021)
NHLBI Growth and Health Study	9–28 (when studied) 40–42 (now)	Cardiovascular	Girls enrolled in 1987 at age 9–10, 50% black 50% white, followed annually or biannually first 20 years	871 enrolled, approx. 500 in follow-up
Nurses’ Health Study 3	18–52 (mothers)	General women’s health	Nurses in N. America	45,066 enrolled, 6,623 pregnancies (at least 20 wks) to date
Women or couples attempting pregnancy
Canadian Assisted Reproductive Technology Cohort Study (CARTS)	Couples (30–45 years) who underwent IVF/ICSI treatment	Long-term outcomes in mothers and their offspring child	18000 couples who under infertility treatment in Canadian province of Ontario with success pregnancy; follow-up through provincial health record linkage is possible	18000 mother-baby pairs (up to December 2018)
Chinese Preconception Cohort Study (CPCS)	Couples (18–35 years) plan to have baby soon (< 6 months)	Long-term outcomes in mothers and their offspring child	5000 couples who planned to have a baby soon in Liuyang, China; follow-up at 8–12 years after childbirth was planned	2000 mother-baby pairs
CoLab Prepregnancy and Early Pregnancy Consortium	> 18 y/o	General Women’s Health with Fertility/Pregnancy Focus	National and international collaboration to share biosamples and clinical data in the peripregnancy time frame.	516 to date (enrollment ongoing)
EAGeR	18–40 (when studied) 24–46 (now)	Fertility	Women with 1–2 prior pregnancy losses attempting pregnancy	1228
Early Pregnancy Study	Median age is 29, range ~ 22–40	Early loss	Women from the Triangle area of North Carolina, enrolled in 1982–86,	221
HOPE Study	Preconception (women ages 18–35, men ages 18–40 upon enrollment)	Environment	Heterosexual couples trying to conceive, greater Salt Lake City area, primarily white	183 couples, 170 with follow-up information
LIFE	18–40 (when studied), 27–53 (now)	Fertility	Couples attempting pregnancy	501
Time to Pregnancy in Couples of Proven Fecundity	Women age 18–35	Fertility awareness instruction	143 women attempting to conceive, prior pregnancy, greater Salt Lake City area, primarily white	140
Preconceptional Offspring Trajectory Study (PLOTS)	20s (two years follow up now)	Environmental, social determinants and life style etc. on Fertility, pregnancy outcomes, child health	Preconception couples followed up to conception and childbirth, continue to follow offspring	at least 3000 couples
Pregnancy Study Online (PRESTO)	Men ≥21 years; women aged 21–45 years	Fertility, pregnancy, child’s health, environment	Preconception cohort of North American (50 US states and 10 Canadian provinces) couples attempting pregnancy	11,000 North American women; 2,500 North American men
Retrospective cohort study of assisted reproductive technology	Mothers 20–40 (at time of birth), all stillbirths and live births (with follow up to one year of age)	Infertility treatment and child health	All births between April 1, 2012 and October 31, 2015 in Ontario captured by BORN, the provincial perinatal registry	17,907 assisted pregnancies
Shanghai Birth Cohort	≥ 20 years	Pregnancy	Women in two preconception care clinics	1180
Snart Gravid (“Soon Pregnant”), Snart Foraeldre (“Soon Parents”)	Men ≥21 years; women 21–49 years	Fertility, pregnancy, child’s health, environment	Preconception cohort of Danish couples attempting pregnancy	13,500 women (mothers); 1,200 Danish men (fathers)
Strong Healthy Women and SMART Strong Healthy Women (adapted for mHealth delivery)	18–35 (when studied) 28–50 (now)	General women’s health and pregnancy	Preconception women residing in rural and urban communities in Central PA	692* (*includes N=45 women who became pregnant during 12-month follow-up)
Time to Conceive	30–44 at enrollment in 2008–2015	Biomarkers of fertility	Women with no history of infertility, over age 30, from the Triangle area of North Carolina	750
University of Iowa Women’s Health Tissue Repository	18–50	General Women’s Health, Fertility, Reproductive Endocrinology, pregnancy and long-term outcomes	Individuals and couples seeking care in the Reproductive Endocrinology & Infertility clinic at the University of Iowa	377 to date (enrollment ongoing)
Upstate KIDS	0–3 years (now 8 years ongoing)	Child health	Population based cohort oversampled on conceptions by infertility treatment in upstate NY	5034 mothers, 6171 children (twins and higher order multiples included)

CARDIA, Coronary Artery Risk Development in Young Adults; EAGeR, Effects of Aspirin in Gestation and Reproduction; HCHS/SOL, Hispanic Community Health Study/Study of Latinos; HOPE, Home Observation of Periconceptional Exposures; LIFE, Longitudinal Investigation of Fertility and the Environment; NCS, National Children’s Study; NGHS, National Growth and Health Study; GUTS, Growing Up and Today Study; PLOTS, Preconceptional Offspring Trajectory Study; PRESTO, Pregnancy Study Online; SMART, KIDS

What does it cover?

The studies in the consortium are designed to address questions of health behaviors, environmental risk factors, fertility, or cardiometabolic health status (Tables 2, 4). The primary focus for four was cardiovascular health, eight was fertility, one was environmental exposures, three was child health, and the remainder general women’s health or no single focus. All have some information on pregnancy health and outcomes. The consortium will investigate the effects of preconception exposures on four broad sets of outcomes: 1) fertility, fecundability, and miscarriage; 2) pregnancy-related conditions (e.g., gestational diabetes, gestational hypertension, preeclampsia, and placental disorders, such as intrauterine growth restriction, placenta previa and placental abruption), 3) perinatal and child health outcomes (e.g., fetal growth, gestational age, child development); and 4) effects of pregnancy course, outcomes, and lactation on women’s health and disease over the life course. Individual studies generally have a strength in one of these areas.

Table 4.

Data included in cohorts for major exposures and outcomes of interest, the PrePARED Consortium

	preconception
	BMI	tobacco/substance use	mood/depression	endocrine conditions	diet	medication	physical activity	stress	cardiovascular
Australian Longitudinal Study on Women’s Health	x	x	x	x	x	x	x	x
Bogalusa Heart Study	x	x	xsubset	subset	x	x	x	subset	x
CARDIA	x	x	x	x	x	x	x	x	x
Canadian Assisted Reproductive Technology Cohort Study (CARTS)	x	x	x	x		x		x	x
Chinese Preconception Cohort Study (CPCS)	x	x		x		x			x
CoLab Prepregnancy and Early Pregnancy Consortium	x	x	x	x	subset	x	subset	subset	x
Growing Up Today Study (GUTS)	x	x	x	x	x	x	x	x	x
Hispanic Community Health Study/Study of Latinos (HCHS/SOL)	x	x	x		x	x	x	x	x
NHLBI Growth and Health Study	x		x	x	x	x	x	x	x
Nurses Health Study 3	x	x	x	x	x	x	x	x	x
EAGeR	x	x			x	x	x	x
Early Pregnancy Study (EPS)		x				x
HOPE Study	x		x	x	x	x	x	x
LIFE	x	x	x	x		x	x	x
Preconceptional Offspring Trajectory Study (PLOTS)	x	x	x		x	x	x	x	x
PRESTO	x	x	x	x	x	x	x	x	x
Retrospective cohort study of assisted reproductive technology	x	x	x
Shanghai Birth Cohort	x	x	x	x	x	x	x	x	x
Snart Gravid, Snart Foraeldre	x	x	x	x	x	x	x	x	x
Strong Healthy Women and SMART Strong Healthy Women	x		x	x	x	x	x	x
Time to Conceive	x	x		x	x	x	x
Time to Pregnancy in Couples of Proven Fecundity	x	x				x		x
University of Iowa Women’s Health Tissue Repository	x	x	subset	x	subset	x	subset	subset	x
Upstate KIDS	x		x	x

	pregnancy				post-pregnancy			child			male partner
	fertility	birth outcomes	pregnancy complications		depression	cardiovas-cular health		asthma	growth/adiposity
Australian Longitudinal Study on Women’s Health	x	x	x		x	x		x	x
Bogalusa Heart Study	x	x	x		subset	subset		x	x		subset enrolled
CARDIA	x	x	x		x	x
Canadian Assisted Reproductive Technology Cohort Study (CARTS)	x	x	x		x	x		x	x		x
Chinese Preconception Cohort Study (CPCS)		x	x		x	x		x	x		x
EAGeR	x	x	x
Early Pregnancy Study	x	x
Growing Up Today Study	x	x	x			x		x	x		reported
Hispanic Community Health Study/Study of Latinos	x	x	x		x	x			x
HOPE Study	x	x	x								enrolled
LIFE	x	x	x								enrolled
National Children’s Study	x	x	x		x				x		reported
NHLBI Growth and Health Study	x	x	subset			x
Nurses Health Study 3	x	x	x		x	x					subset enrolled, others reported
Preconceptional Offspring Trajectory Study (PLOTS)	x	x	x		x			x	x		reported
Preconceptional cohort study of pre-eclampsia and gestational diabetes
PRESTO, Snart Gravid, Snart Foraeldre	x	x	x		x			x	x		enrolled
Retrospective cohort study of assisted reproductive technology	x	x
Shanghai Birth Cohort	x	x	x		x	x		x	x
Snart Gravid, Snart Foraeldre	x	x	x		x			x	x		enrolled
Strong Healthy Women and SMART Strong Healthy Women
Time to Conceive	x	x	x
Time to Pregnancy in Couples of Proven Fecundity	x	x	x								reported
Upstate KIDS	x	x	x		x	x		x	x		reported

x=routinely measured; subset=specifically measured in a subset (not just missing data)

BHS, Bogalusa Heart Study; CARDIA, Coronary Artery Risk Development in Young Adults; EAGeR, Effects of Aspirin in Gestation and Reproduction; HCHS/SOL, Hispanic Community Health Study/Study of Latinos; HOPE, Home Observation of Periconceptional Exposures; LIFE, Longitudinal Investigation of Fertility and the Environment; NCS, National Children’s Study; NGHS, National Growth and Health Study; GUTS, Growing Up and Today Study; PLOTS, Preconceptional Offspring Trajectory Study; PRESTO, Pregnancy Study Online; KIDS

CARDIA, Coronary Artery Risk Development in Young Adults; EAGeR, Effects of Aspirin in Gestation and Reproduction; HOPE, Home Observation of Periconceptional Exposures; LIFE, Longitudinal Investigation of Fertility and the Environment; NCS, NGHS, National Growth and Health Study; GUTS, Growing Up and Today Study; PRESTO, Pregnancy Study Online; KIDS

How often have they been followed up and what is the anticipated attrition?

Among other cohorts approached for inclusion, the most common reasons for declining or ineligibility were a) lack of time by the investigators, and b) lack of prospectively collected preconception data. Attrition within the cohorts is quite variable and tends to track with the length of time the cohort has been followed; however, many of the long-term studies have impressive rates of follow-up.⁴⁸ An eligibility criterion is that all studies should include at least some follow-up for pregnancy outcomes. Studies also vary in their ability to recontact participants. Some studies also include participants who were never pregnant or delivered a live birth, allowing for comparisons that address selective fertility and pregnancy loss.

What are the major areas of research?

Data analysis will consist of determining how existing data corresponds across cohorts, and harmonizing key exposures, outcomes and covariates. Initial preconception exposures of interest include nutrition, diet, and obesity; marijuana, tobacco and other substance use; cardiometabolic risk factors; and male partners’ exposures. These topics were chosen because of their relevance to current public health concerns and the fact that they are measured across many of the cohorts, allowing for establishing data harmonization protocols. In addition, the harmonized variables developed in the course of these projects are likely to be important confounders and mediators of analyses on other topics of interest.

Comment

The large numbers will allow for subgroup analysis. Most of the studies are quite detailed in at least one area, and often have extensive measures of cardiovascular risk, diet, fertility, and/or environmental factors. The included studies are diverse in their populations and in study design, which will increase generalisability, as well as allowing the description of the interrelationships between the natural history of pregnancy and women’s health over the life course. Findings that indicate systematic similarities or differences across cohort studies can provide insights into the behavioural, social, or biological factors at play.

Weaknesses include the fact that many of the studies are complete, with wide variability in the types of measures and biological samples available. Some of the studies are limited to pregnancy planners, who tend to be healthier and older than the average pregnant woman, while general population studies do not have a consistent time frame between the measures and the pregnancy, and may lack data collected during pregnancy itself. Some studies of reproductive-aged populations, such as the Coronary Artery Risk Development in Young Adults (CARDIA) Study, use a mix of data collection methods, conducting assessments in women both before and after pregnancy, with pregnancy data validated with medical records. Other studies were conducted among women at risk of pregnancy loss or using of infertility treatment, as such groups are more prone to be followed prospectively. The combination of study designs means that in some cases, subsamples from some cohorts may be needed to establish a consistent time frame of measurement (either age or relative to pregnancy) or degree of validation (e.g., medical record confirmation). A range of calendar time is covered by the studies, both a potential strength and limitation, as it allows for examining whether an effect is constant over time, and thus more likely to be causal, but also increases the possibility of inconclusive findings due to heterogeneity. Cohort effects will most commonly be addressed, by adjustment or interaction analysis, where possible. If there is reason to believe that results will be relevant to a particular time period, analyses will be limited to studies with data from that period only.

Preconception health can be defined as including men’s as well as women’s health.⁴⁹ Several of the cohorts include male participants, either as partners in couples planning pregnancy, or as participants in general-population studies. Pregnancy health is a major focus of the consortium and so participating cohorts are required to have at least some information on female health and pregnancy. Cohorts that focus on male fertility are eligible to participate as long as they also include this information.

Conclusions

The consortium welcomes proposals to collaborate from qualified researchers. The corresponding authors should be contacted for more information. Persons interested in requesting data should contact the corresponding author for more details on procedure. At a minimum, a data use agreement will be required, and costs may be required to be shared. Updated information on the consortium can be found at the consortium website (https://sph.tulane.edu/prepared-consortium). Beyond the topics identified, the consortium will serve as a resource for work in many subject areas related to the preconception period, with implications for science, practice, and policy.

Synopsis.

• Study question: Can studies covering the preconception period be combined to allow for better understanding of this life period?

• What’s already known: Several risk factors for adverse pregnancy outcomes begin prior to pregnancy.

• What this study adds: By combining efforts from several studies and study designs, gaps in understanding such as the effect of timing and duration of exposures, rare exposures, small effect sizes, etiologic heterogeneity, disease subtypes, selection issues, and meaningful subgroup analysis (e.g., by race and ethnicity) can be addressed.