Refugee Births and the Migrant Health Effect in Syracuse NY

Abstract

Objectives

In Syracuse, NY among 5998 births in a 3-year period (2017–2019), 24% were to foreign-born women, among whom nearly 5% were refugees from the Democratic Republic of the Congo and Somalia. The impetus for the study was to identify potential risk factors and birth outcomes of refugee women, other foreign-born women, and US born women to inform care.

Methods

This study reviewed 3 years of births (2017–2019) in a secondary database of births in Syracuse, New York. Data reviewed included maternal demographics, natality, behavioral risk factors (e.g., drug use, tobacco use), employment, health insurance, and education.

Results

In a logistic regression model controlling for race, education, insurance status, employment status, tobacco use and illicit drug use, compared to US born mothers, refugees (OR 0.45, 95% CI 0.24–0.83) and other foreign born (OR 0.63, 95% CI 0.47–0.85) had significantly fewer low birth weight births.

Conclusion

The results of this study supported the “healthy migrant effect,” a concept that refugees have fewer low birth weight (LBW) births, premature births, and cesarean section deliveries than US born women. This study adds to the literature on refugee births and the healthy migrant effect.

Significance

The United States is a major receiving country for refugees. Because refugees have often experienced serious trauma and limited healthcare prior to resettlement, analyses of their birth experiences and pregnancy outcomes would help to guide care during pregnancy and post-partum.

Introduction

Since 1980 the United States (US) has accepted more than 3.1 million refugees (US Department of State, n.d). Among refugees resettled in the US in 2019, 50.9% were of reproductive ages (15–49 years old) (Statistia, 2022) and 49% were female (Statista, 2022). The fertility of foreign-born US residents (including refugees and other immigrants) is about one third higher than their native-born counterparts. According to the Annie E Casey Foundation 23% of US births in 2019 (n = 843,869) were to foreign-born persons (Kids Count Data Center—The Annie E. Casey Foundation, 2022). Foreign-born persons in the US experience challenges with language, cultural differences, stigma, poverty, and access to healthcare. Refugees have often fled from their homes with little or no preparation, many spending months or years in a resettlement camp, where they may have been exposed to violence, disease, and the deaths of family members including children. They may speak little or no English upon arrival, and some lack literacy in their native language.

The “healthy migrant effect” (HME) with regard to pregnancy is an unexpected finding that foreign-born women have healthier birth outcomes (Fennelly, 2007; Franzini et al., 2001; Juárez et al., 2017). Studies identifying the HME have found that the infants of refugees have fewer low birth weight births, lower premature births, and fewer infant deaths than infants born to US mothers (Hauck, 2019; Wingate & Alexander, 2006). Most of the studies on the HME compare the births of US born and foreign-born persons, without distinguishing between refugees and other immigrants. Wanigaratne et al. (2016) examined the rates of severe neonatal morbidities (SNM) among Canadian infants with regard to mothers’ natality; SNM of infants of refugees to Canada (4.6%), immigrants (4.2%), and Canadian-born mothers (5.0%) differed modestly, with the infants of refugees and immigrants having fewer SNMs. A study that addressed parental natality found that foreign-born mothers and foreign-born fathers were each independently associated with fewer low birth weight (LBW) births, compared with US born parents (Krishnakumar et al., 2011).

Research on the protective factors involved in the HME is less robust than the studies on incidence, prevalence out outcomes cited above. Among the studies that address protective factors resulting in the HME are those that cite migrants’ lower tobacco use (Kuerban, 2016), better oral health of migrants (Sanders, 2010), and lower alcohol use (Hearld et al., 2015). Also, migrants who arrive in the United States are likely to be healthy than similar individuals from their country of origin who did not have the health or stamina to make what is frequently an arduous journey.

Research Question

The study was designed to review the birth outcomes for infants of refugee, foreign-born, and US born women to use in program planning, education of obstetrical residents and other clinicians, and health education to refugees.

Methods

Setting Syracuse, a city in Central New York, has the third highest receipt of refugees of cities in the United States. There are two very active refugee resettlement agencies, who have welcomed new Americans to the city beginning with the mid-1970s arrival of Southeast Asians following the Vietnam war (Lupone et al., 2020). The 2020 US census found 13% of the city population to be foreign born, although the census does not distinguish between voluntary migrants and refugees. Local healthcare providers have extensive experience caring for refugees and immigrants. Local healthcare providers follow the federal Office for Civil Rights mandated under Title VI legislation, to offer Limited English Proficient (LEP) services to patients whose care provider does not speak their primary language (Lane, 2008). But even accurate translation of the patient-provider communication may fall short of full understanding.

Research Design and Participants

The study used secondary data of all births in Syracuse for 3 years (2017–2019), from the New York State Perinatal Data System (SPDS), which compiles and documents prenatal data and birth outcomes from all birth hospitals in New York State. The births included in this study are all births to mothers residing in the city of Syracuse, NY, who gave birth in the city of Syracuse, NY. The SPDS is a de-identified population-based birth registry that captures pregnancy and birth information for quality improvement. The 3 years used for analysis 2017, 2018, 2019 were chosen in order to avoid potential bias resulting from the COVID-19 pandemic.

The SPDS operates under New York State rules and regulations as follows: “All hospitals and freestanding birthing centers that provide perinatal health care services shall participate in the statewide perinatal data system. All live births shall be entered into the state perinatal data system” https://regs.health.ny.gov/content/section-40022-statewide-perinatal-data-system.

We consulted with the staff of a local refugee resettlement agency, who advised that individuals from the DRC and Somalia were the most recently resettled during the 3-year period of the study (1 January 2017–31 December 2019). Mothers’ natality was separated into three groups: US born, mothers, those born in the Democratic Republic of the Congo, and Somalia (coded as “refugee mothers”), and foreign-born women from other countries. As documented in the SPDS, the “other foreign-born mothers” were from 98 different countries. The rank order of their country origins (for countries with more than 25 women each) were: Bahamas, Bhutan, Cuba, Kenya, Yemen, Sudan, China, Thailand, Nepal, Iraq, Vietnam, Syria, India, and Greece. The birth database does not distinguish between refugees and voluntary migrants; the group of “other foreign-born” women may include both refugees and voluntary migrants. Women from the DRC and Somalia may also include some voluntary migrants.

Measures

The variables documented in the SPDS are based on the women’s disclosure during pre-natal care (race, age, insurance, natality, education, substance use (tobacco, alcohol, illicit drugs), depression during pregnancy, intentions of pregnancy, maternal race) coded as Black only, white only, Asian only, and other). Other variables are extracted from the antenatal and hospital delivery records. Self-reported tobacco use, alcohol intake, and illicit substance use were coded as dichotomous (Yes/No) variables, participation in WIC (yes/no), diabetes (yes/no), gestational diabetes (yes/no), employment (yes/no), Medicaid (yes/no). All data, including those about substance use (illicit substances and tobacco) are from the SPDS.

Among births to single mothers in Syracuse-area birth hospitals, the baby’s father is given the opportunity to sign the Declaration of Paternity on the morning after the birth. If the father does not sign, it is an indirect indication of less paternal emotional and financial support at the time of the birth (Keefe et al., 2017; Lane et al., 2004).

Data Analysis

We compared births in three groups categorized by location of the mothers’ births: US born, refugees from the Democratic Republic of the Congo and Somalia, and other foreign-born women from other countries. The main outcome variable in this analysis is low birth weight births, defined as births of less than 2500 g. Variables were compared with ANOVA for continuous variables and chi-square or Fisher’s exact tests for discrete and dichotomous variables, as appropriate.

In order provide sufficient data to understand the differences among the three groups in prenatal risk factors and birth outcomes, we include information about the labor and delivery characteristics, birth weight, gestational age, NICU admissions, and Down’s Syndrome diagnoses.

Multivariable logistic regression analysis was used to estimate risk of having a low birth weight birth (singleton births only) among refugees and other foreign-born mothers relative to US born mothers. We examined maternal age, maternal race, maternal education, insurance status, maternal employment status, tobacco use, alcohol use, illicit drug use, participation in WIC, previous birth, pregnancy intention, and paternal acknowledgement as potential confounding variables. Variables were determined to be confounders if they were independently associated with country of birth, delivery of a low birthweight infant, and changed the odds ratio(s) for country of birth and delivery of a low birthweight infant by at least 10 percent. The final multivariable logistic regression model includes all variables meeting the confounding definition and included maternal race, maternal education, insurance status, maternal employment status, tobacco use, and illicit drug use.

Results

There was a total of 5998 births in Syracuse in the 3 years (2017–2019), which were divided into three categories: births to US born (n = 4635), refugees from the DRC and Somalia (n = 289), and other foreign-born women (1074).

As presented in Table 1, US born mothers had twice the percentage of teen births compared to births of other foreign-born mothers and almost 4 times as many as refugee mothers. Maternal race differed greatly in the three groups, reflecting their countries of origin. The refugees had over twice the proportion of non-high school graduates as the two other groups. All three groups had experienced substantial poverty, as measured by Medicaid insurance status, with nearly all of the refugees having Medicaid coverage. Multiple births were not statistically different among the three groups. None of the refugees used tobacco during the pregnancy, compared with 2.2% of the other foreign-born women, and nearly a quarter of the US born women. Very few women in any of the groups reported using alcohol during the pregnancy. More than one in five US born women reported illicit substance use during their pregnancy compared to 1.4% of other foreign-born women, and zero refugee women. Most women in all three groups had given birth previously. Among US born and other foreign-born women just over 10% were primiparous, compared with just over 5% of the refugees. Among the US born women 45.1% reported that their pregnancy was unintended (the combined score of the two options, “wanted later” or “did not want”), compared with 28.2% of the other foreign-born women, and 6.2% of the refugees. Self-reported depression during pregnancy was documented for 10.2% of US born women, 2.5% of other foreign-born, and 6.7% of refugees. Marital status differed greatly among the three groups: 23% of the US born women were married, compared with 75% of other foreign-born, and 71% of the refugees. In 26.5% of US born, 6.3% of other foreign-born, 9.3% of refugee births the father was not married to the mother and did not sign the Declaration of Paternity. The mean maternal ages ranged from 27.2 to 31 years among the three groups and mean number of prenatal visits ranged from 11 to 11.7. Maternal age and mean number of prenatal visits were statistically different among the groups, but the magnitude of difference was small. Mean maternal weight gain differed significantly among group, ranging from 27.6 Lbs. for US born women, to 23.8 Lbs. for other foreign-born, and 17.4 Lbs. for the refugees.

Table 1.

Maternal demographic, behavioral, and medical history characteristics by maternal natality

Maternal characteristic	US born mothers (N = 4635)		Refugee mothers (N = 289)		Other foreign-born mothers (N = 1074)		P-value
	N	%	N	%	N	%
Age group (years)
< 18	120	2.6	2	0.7	13	1.3	< 0.0001
18–35	4086	88.2	219	75.8	896	83.4
36–40	375	8.1	48	16.6	128	11.9
> 40	54	1.2	20	6.9	37	3.5
Race
White only	1928	41.6	0	0.0	193	18.0	< 0.0001
Black only	1848	39.9	272	94.1	257	23.9
Asian only	18	0.4	0	0.0	438	40.8
Other	841	18.1	17	5.9	186	17.3
Education
Less than HS grad	1186	25.6	207	71.6	422	32.3	< 0.0001
HS grad	1478	31.9	48	16.6	215	20.0
Some college/Assoc	1259	27.2	21	7.3	172	16.0
College grad and beyond	699	15.1	3	1.0	255	23.7
Unknown	13	0.3	10	3.5	10	0.9
Insurance
Medicaid	3387	73.1	281	97.2	821	76.4	< 0.0001
Private	1163	25.1	3	1.0	242	22.5
Other	84	1.8	5	1.7	11	1.0
Employment status during pregnancy
Employed	2586	55.8	88	30.5	371	34.5	< 0.0001
Not employed	2049	44.2	201	69.6	703	65.5
Multiple birth
Yes	153	3.3	11	3.8	31	2.9	0.68
No	4482	96.7	278	96.2	1043	97.1
Smoking
Yes	1142	24.7	0	0.0	24	2.2	< 0.0001
No	3486	75.3	289	100.0	1050	97.8
Alcohol
Yes	30	0.7	0	0.0	2	0.2	< 0.0001
No	4602	99.4	289	100.0	1072	99.8
Illegal drugs
Yes	977	21.1	0	0.0	15	1.4	< 0.0001
No	3657	78.9	289	100.0	1059	98.6
WIC
Yes	3081	66.7	260	90.3	718	67.0	< 0.0001
No	1540	33.3	28	9.7	353	33.0
Previous birth
Yes	4106	88.6	272	94.4	957	89.1	0.01
No	529	11.4	16	5.6	117	10.9
Depression during pregnancy
Not depressed	2953	65.8	174	82.5	777	81.9	< 0.0001
A little depressed	1079	24.0	23	10.9	148	15.6
Moderately depressed	288	6.4	9	4.3	19	2.0
Very depressed	170	3.8	5	2.4	5	0.5
Pregnancy intention
Sooner	462	10.5	50	23.3	192	20.4	< 0.0001
Later	1613	36.6	46	2.5	194	20.6
Then	1957	44.4	111	51.6	484	51.4
Did Not Want	374	8.5	8	3.7	71	7.6
Paternity declaration
Yes (married or father signed PD)	3409	73.6	262	90.7	1006	93.7	< 0.0001
No	1226	26.5	27	9.3	68	6.3

Characteristic (continuous variables)	Mean	SD	Mean	SD	Mean	SD	P-value
Age (years)	27.2	5.8	31.0	6.1	29.1	6.0	< 0.0001
# of prenatal care visits	11.7	4.4	11.0	3.7	11.5	3.9	0.03
Weight gain during pregnancy	27.6	18.9	17.4	15.8	23.8	14.1	< 0.0001

Number missing on counseled about during pregnancy variables varied. Chi-square or Fisher’s exact tests when cells < 5. ANOVA used for continuous variables

Table 2 shows that US born women had slightly over twice the prevalence of pre-pregnancy hypertension, compared with other foreign-born or the refugees. US born women also had twice the prevalence of pregnancy-related hypertension of the refugees, and over twice as much as the other foreign-born. Gestational diabetes, however, was similar in US born women and refugees, but almost twice as high among the other foreign-born.

Table 2.

Maternal medical risk factors by maternal natality

Medical risk present	US (N = 4635)		Refugee mothers (N = 289)		Other foreign-born (N = 1074)		P-value
	N	%	N	%	N	%
Diabetes (Gestational)	252	5.4	16	5.5	95	8.9	0.0001
Hypertension (pre-pregnancy)	165	3.6	5	1.7	19	1.8	0.004
Hypertension (pregnancy-related)	418	9.0	13	4.5	40	3.7	< 0.0001
Pre-labor referral for high risk	517	11.2	54	18.7	161	15.0	< 0.0001

Subjects may have multiple medical risks. Chi-square or Fisher’s exact tests when cells < 5

As presented in Table 3, the refugees had significantly fewer cesarean Sects. (20.9% vs 29.9 for US born and 29% for other foreign-born). Fewer refugees needed medical induction of their labors, maternal analgesia, or epidural/spinal anesthesia.

Table 3.

Characteristics of labor and delivery among singleton births (N = 5803) by maternal natality

Characteristic	US (N = 4482)		Refugee mothers (N = 278)		Other foreign-born (N = 1043)		P-value
	N	%	N	%	N	%
Refused VBAC +	569	12.7	19	6.8	129	12.4	0.02
Delivery route
Spontaneous	3059	68.3	213	76.6	711	68.2	0.006
Vacuum/forceps	83	1.9	7	2.5	30	2.9
C-section	1340	29.9	58	20.9	302	29.0
Fetal presentation
Cephalic	4309	96.1	271	97.5	993	95.2	0.02
Breech combined with other	146	3.3	6	2.2	33	3.2
Other	27	0.6	1	0.4	17	1.6
Labor characteristic
Induction of labor	1210	27.0	55	19.8	230	22.1	0.0003
Maternal analgesia	2458	54.9	86	30.9	533	51.1	< 0.0001
Anesthesia
None	1138	25.4	128	46.0	228	21.9	< 0.0001
Epidural	1884	42.0	50	18.0	381	36.5	< 0.0001
Spinal	1096	24.5	47	16.9	250	24.0	0.02

Subjects may have multiple onset of labor, labor, anesthesia, and reason for C-section present. Chi-square or Fisher’s exact tests when cells < 5

+ Vaginal birth after C-section

As presented in Table 4, the refugees had lower incidence of infants with LBW (4% compared to US born mother (10.3%), and other foreign-born mothers (5.8%), but higher macrosomia (13.3%, compared with 7.6% among US born and 8.5% among other infants born to foreign-born mothers). Among full term infants, refugees and other infants born to foreign-born mothers had nearly 50% lower LBW than infants born to US mothers. Premature births also differed among the three groups, with 3.2% of infants of refugee mothers being premature, compared with 5.7% of infants of other foreign-born mothers and 10% of infants of US born mothers. NICU admissions were lower among infants of refugee mothers (8.6%), compared with infants of US born mothers (12.6%) and infants of other foreign-born mothers (10.4%). An exception to the generally more positive birth outcomes for refugee and foreign-born women is that the full-term infants of refugee mothers had much higher macrosomia (13.8) than infant of US born (8.4%) or infants of other foreign-born mothers (9%). Four infants with Down Syndrome were born to refugees and one was born to other foreign-born women. No US born women gave birth to infants with Down Syndrome.

Table 4.

Infant outcomes among singleton births (N = 5803) by maternal natality

Outcome	US (N = 4482)		Refugee mothers (N = 278)		Other foreign-born (N = 1043)		P-value
	N	%	N	%	N	%
Birth weight (grams)
< 2500	463	10.3	11	4.0	60	5.8	< 0.0001
2500 – 3999	3677	82.0	230	82.7	894	85.7
≥ 4000	342	7.6	37	13.3	89	8.5
Gestation (weeks)
< 37	448	10.0	9	3.2	59	5.7	< 0.0001
≥ 37	4034	90.0	269	96.8	984	94.3
Birth weight (grams) among full-term infants
< 2500	167	4.1	6	2.2	23	2.3	0.002
2500 – 3999	3527	87.4	226	84.0	872	88.6
≥ 4000	340	8.4	37	13.8	89	9.0
NICU Admission	564	12.6	24	8.6	108	10.4	0.03
Down Syndrome	0	0.0	4	1.4	1	0.1	< 0.0001

Outcome (continuous variables)	Mean	SD	Mean	SD	Mean	SD	P-Value
Birth weight (grams)	3214.2	618.3	3449.5	532.3	3295.5	533.9	< 0.0001
Gestation (weeks)	38.6	2.4	39.6	1.5	38.9	1.8	< 0.0001

Subjects may have multiple abnormal conditions of newborn present. Chi-square or Fisher’s exact tests when cells < 5. ANOVA used for continuous variables

Table 5 presents the adjusted odds ratios of LBW among singleton births of the three groups, controlling for maternal race, Medicaid insurance as a proxy for poverty, and tobacco use. We examined age as a potential confounder by comparing logistic regression models with and without age included. Including age in the model did not change the effect estimates (odds ratios), so we did not retain age in our final model. As the table shows compared to infants born to US mothers (the referent group), infants born to refugees had 58% lower odds of LBW and the odds for infants born to other foreign-born mothers were 41% lower.

Table 5.

Multivariable logistic regression model for the association between maternal natality and delivery of low birthweight infant among singleton births, controlling for maternal race, maternal education, insurance status, maternal employment status, tobacco use, and illicit drug use

Maternal natality	Odds ratio	95% CI
Refugees	0.45	0.24–0.83
Other foreign born	0.63	0.47–0.85
US born	1.00	Referent
Race	1.16	1.07–1.26
Education	0.99	0.89–1.10
Insurance status	0.77	0.61–0.98
Employment status (unemployed)	1.24	1.02–1.52
Tobacco use	1.81	1.46–2.24
Illicit drug use	1.82	1.46–2.26

Discussion

This 3-year secondary data analysis of birth outcomes in Syracuse supports the healthy migrant effect in that foreign-born women had fewer low birth weight births, premature births, and cesarean sections than their US born counterparts. Contributing to the paradoxical nature of those findings, prior to this birth more refugee women (18.7%) were identified as “high risk,” compared to other foreign-born women (15%), and US born women (11.2%). Refugee mothers on average had less education, more poverty as indicated by Medicaid insurance coverage, and higher rates of employment. Yet, recent refugees from the DRC and Somalia had better outcomes on measures listed above than foreign-born women who had resided in the US for a longer time or US born women.

Foreign-born Syracusans have higher fertility than US born residents, as documented by the crude birth rate (2017–2019) of foreign-born Syracuse residents (24.3 per 1000 population), compared to US born (12.5 per 1000 population).1 This higher fertility is likely attributable to the foreign-born population being younger, with more individuals in the reproductive age group. Another factor is that many of the foreign-born mothers come from countries and cultures where fertility rates are higher than the U.S.

Refugees from the DRC and Somalia have better birth outcomes than US born women or foreign-born women who arrived earlier. This may result from lower rates of high-risk behaviors that are linked to poor birth outcomes. Refugee mothers were significantly less likely to be teenagers. No refugee mothers reported use of tobacco, alcohol, or illicit substances during pregnancy. Refugee mothers appeared to be more likely to have support from the infant’s father. Refugees had far fewer teen births, and no refugee women used tobacco, took illicit drugs, or consumed alcohol. In fewer than 10% of refugee and other foreign-born births the father did not sign the DP, compared with 26.5% of US born births. Previous research documented that when the father did not sign the DP, the rate of post neonatal mortality was 3.8 times that of infants in which the father was either married to the woman or he signed the DP (Lane et al., 2004). Intended pregnancy also differed greatly among the three groups, with nearly half of US born women (45.1%) reporting that their pregnancy was unintended, compared with 28.2% of the other foreign-born women, and 6.2% of the refugee women. Participation in WIC (the Special Supplemental Nutrition Program for Women, Infants, and Children) also differed greatly, with over 90% of the refugees, compared to 67% each of the US born and other foreign-born women. WIC participation during pregnancy was previously found to be associated with fewer LBW births (Lane, 2008; Lane et al., 2004).

Limitations

A limitation of the study is that the operationalization of refugees could exclude refugees from other countries, particularly those who arrived prior to 2017. Poverty and partner support are both examined by proxy (Medicaid participation, signed paternity declaration), which introduces a limitation. An additional limitation is that the SPDS secondary database uses self-report for illicit substance and tobacco use, which is not as precise as would be using more objective biological measures.

Author Contributions

GG: identified the research questions and conducted the literature review. MF: conducted the quantitative analysis. SDL: assisted with the quantitative analysis and wrote the review of results. MS: advised on the obstetrical issues involved in the analysis and suggested variables and measures to include. CS: advised on the obstetrical issues involved in the analysis and suggested variables and measures to include. RAR: participated study design and in writing the manuscript. AS: responsible for obtaining IRB approval and for the research design. Also participated in the analysis.

Funding

Not applicable.

Data Availability

The study used secondary data from the NY State Perinatal Data System.

Code Availability

Statistical analysis was done using SPSS.

Declarations

Conflict of interest

The author have no conflicts of interest to declare.

Ethical Approval

This study was approved by the Institutional Review Board of SUNY Upstate Medical University and was designated exempt.

Consent to Participate

Not applicable.

Consent for Publication

Not applicable.