Synthesis of Results
This review identified 8 SVFs that are commonly assessed in association to GWG: race/ethnicity, age, parity, marital status, income, education, immigration status, and abuse (physical, psychological or sexual). GWG was reported either as a continuous outcome or categorized and compared between groups of pregnant women (i.e., one group having a higher or lower GWG than the other) or between GWG adequacy groups (i.e., insufficient, adequate, excessive). The characteristics of each study and their association with the 8 SVFs are reported in Supplemental Table 3. Most studies (60%) had a retrospective design, 32% were prospective, and 9% were cross-sectional. GWG data were retrieved from medical charts (39%) or birth certificates (20%), participant self-report (20%), or measured by the study team (17%); 5 studies (3%) used two different methods for collecting GWG data, and 2 studies (1%) did not clearly state their GWG collection method. A synthesis of studies that examined the relationship between each SVF and GWG is detailed below.
Race and ethnicity. Ninety-one studies (58%) assessed the relationship between race/ethnicity and GWG (Supplemental Table 4). Seventy-two studies found a significant association between GWG and race/ethnicity, while 19 studies found no significant association. Studies were published from 1976 to 2019 and were conducted in the USA (75 studies), Canada (3 studies), Australia, the Netherlands (2 studies each), Belgium, Czech Republic, New Zealand, Norway, Singapore, Spain, Sweden, Switzerland (1 study each), and one study was conducted in multiple countries (New Zealand, Australia, and Ireland). Samples sizes ranged from 56 to 7,966,573 individuals. Twenty-eight studies included a population that was primarily affected by a SVF (adolescents, women with a low income, unmarried women, and/or women from a racial or ethnic minority group). The most studied racial or ethnic groups were White, Black, and Hispanic, with some studies specifying both a woman’s racial group and their Hispanic/non-Hispanic identity. One of the least studied groups was Indigenous women.
Age. Eighty-seven studies (55%) considered the association between age and GWG (Supplemental Table 5). Forty-six studies found a significant association and 41 studies found no significant association. A comparison of GWG between adolescents and adults was conducted in 36 of these studies, with 25 studies finding a significant association and 11 finding no significant association. Articles were published from 1977 to 2019 and took place in the USA (67 studies), Taiwan (3 studies), Canada, Japan, Korea (2 studies each), Australia, Austria, Belgium, the Netherlands, Norway, Saudi Arabia, Singapore, Slovenia, Sweden, Switzerland (1 study each), and one study involved 3 countries (New Zealand, Australia, Ireland). Sample sizes ranged from 55 to 3,960,796 individuals; thirty-five studies were specific to women with a SVF (adolescents, women with a low income, women of a racial or ethnic minority group, unmarried women, low education, and/or multiparous women). Age was primarily assessed as a categorical variable, with a diverse range of age categories being used (e.g., < 25 years vs. ≥25 years; and < 20 years, 20–29 years, 30–39 years, and ≥ 40 years). The age range for adolescents and adults varied between studies, with some studies categorizing individuals as young as 17 as adults.
Parity. Forty-eight studies (31%) examined the relationship between parity and GWG (Supplemental Table 6). Significant associations between these factors were reported in 38 studies, while 10 studies reported none. Publication years ranged from 1990 to 2019. Studies were conducted in the USA (37 studies), Canada, Sweden (2 studies each) Australia, Belgium, Korea, the Netherlands, New Zealand, Singapore, and Switzerland (1 study each). The smallest sample size was 55 individuals, while the largest was 2,976,805 individuals. Twenty-two studies had a population that primarily included women with a SVF (racial or ethnic minority, adolescents, and/or women with a low income). The terms ‘nulliparous’ and ‘primiparous’ were often used to refer to women whose first delivery was the one examined in the study. The GWG of nulliparous or primiparous women was generally compared to that of multiparous women. A more in-depth assessment of parity was performed in 15 studies, where either more categories of parity were included (e.g., 0, 1, 2, or ≥ 3 previous births) or parity was assessed as a continuous variable.
Marital status. Twenty-eight studies (18%) assessed the association between marital status and GWG, with 15 studies finding a significant association and 13 studies found no significant association (Supplemental Table 7). Studies were published between 1985 and 2019 in the USA (25 studies), Belgium, Canada, and Sweden (1 study each). Sample sizes ranged from 101 to 251,342 individuals, with half the studies involving a population affected by one or more SVF (women with a low income, a racial or ethnic minority group, multiparous, and/or adolescents). Categories used to assess marital status varied between studies; the general categories of ‘married’ and ‘unmarried’ were commonly used, but some studies opted for more specific categories such as ‘married/cohabitating with partner’, and ‘single/separated’, or ‘married/partnered’, ‘separated/divorced’, or ‘single/no partner’.
Income. Thirty-nine studies (25%) examined the impact of income on GWG (Supplemental Table 8). Twenty studies found a significant association, while 19 found no significant association. Article publication years ranged from 1995 to 2019. Studies were conducted in the USA (33 studies), Canada, Korea (2 studies each), the Netherlands, and Singapore (1 study each). The smallest sample size was 75 individuals and the largest was 515,148 individuals. Thirteen studies were restricted to a population that had one or more SVF (adolescents, women from a racial or ethnic minority group, and/or women with a low income). Studies seldom used the same income categories (e.g., $0-500, $501-1,000, and ≥$1,000; <$20,000 v ≥$20,000; or < 100% of federal poverty line vs. >100% federal poverty line). Nearly half (18 of 39) of all studies used a proxy measure to assess a woman’s income status (e.g., use of Medicaid, or enrollment in the Special Supplemental Nutrition Program for Women, Infant and Children (WIC) in the USA).
Education. Forty-four studies (28%) examined the association between GWG and education level (Supplemental Table 9). Of these, 34 reported a significant association while 10 reported no significant association. Studies were conducted between 1992 and 2019 in the USA (33 studies), Sweden (3 studies), Canada (2 studies), Belgium, Croatia, Korea, the Netherlands, Norway, and Spain (1 study each). Sample sizes ranged from 55 to 2,796,805 participants. Fifteen studies were specific to women with a SVF (racial or ethnic minority group and/or women with a low income). Education was measured as a continuous variable in 3 studies, with the remaining studies measuring education as a categorical variable. Categories varied between studies, with some studies only including two categories (e.g., less than high school vs. high school or more), and others including up to 5 categories (e.g., 0–8 years, 9–11 years, 12 years, 13–15 years, or ≥ 16 years of education).
Immigration status. Nineteen studies (12%) assessed the relationship between immigration status and GWG, with all but one finding significant associations between these factors (Supplemental Table 10). Studies were published from 1997 to 2018 in the USA (14 studies), Canada (3 studies), France (1 study), and one study included multiple countries (New Zealand, Australia, and Ireland). The smallest sample size was 46 individuals and the largest was 250,857. Nine studies were limited to women with an SVF (racial/ethnic minority, adolescent, and/or low-income). Approximately half of the studies considered immigration status as a dichotomous variable (i.e. born in their country of residence or foreign-born), and 6 studies categorized individuals according to their length of time in their country of residence (e.g., US-born, lived in the USA ≥ 10 years, lived in the USA < 10 years). In all 6 studies, the length of time since an individual’s immigration had a significant effect on GWG.
Physical, psychological, and sexual abuse. Twelve studies (8%) considered the impact of experiencing abuse on GWG (Supplemental Table 11). Nine studies found a significant association and 3 found no significant association. Studies were published from 1996 to 2017 in the USA (10 studies), Iceland, and Taiwan (1 study each). Sample sizes ranged from 337 to 251,342 individuals. Half of the studies included a population with one or more SVF (adolescents, low-income, and/or racial/ethnic minority). Studies primarily assessed the impact of physical abuse on GWG, but sexual and psychological abuse were also examined. The timing of abuse (e.g., during childhood, before pregnancy, or during pregnancy) and the perpetrator of abuse (e.g., intimate partner) were considered in some studies.