Births to Hispanic women in the United States represent 18% of total births,1,2 and Hispanics are the fastest growing segment of the United States' population. In 1996, 70% of Hispanic births in the United States were to Mexican Americans.2 The prevalence of fetal growth restriction (FGR) might be underestimated for Mexican American infants based on existing references for birth weight for gestational age.3 Existing references might also underestimate excessive fetal growth associated with later risk of overweight in Mexican American children and adults.4
National references for birth weight for gestational age usually combine non-Hispanic whites and Hispanics as “white,”5 or combine all races,6 because of a lack of information on Hispanic ethnicity on birth certificates. Reporting on Hispanic parentage began in 1989, with all states reporting since 1993.1 This analysis derives references percentiles for birth weight for gestational age of United States births to Mexican American mothers that can identify infants either small for gestational age (SGA, below the tenth percentile) or large for gestational age (LGA, at or above the 90th percentile).
Materials and Methods
The data analyzed were from electronic natality files for 1992–1994, obtained from the National Center for Health Statistics.7 Analysis was restricted to singleton births to United States residents. Infant race and ethnicity were specified on the natality files by a standard National Center for Health Statistics' algorithm, based primarily on mother self-identification.7 Previous reference data included Mexican American infants with non-Hispanic whites,5 so we focused only on births classified as Mexican American or non-Hispanic white. A total cell size of about 500–800 births for each week of gestation was necessary to yield stable and precise estimates with about 125–200 births at each week for four maternal parity and gender groupings (Table 1). The number of births in 1994 was sufficient for non-Hispanic whites to meet sample size criteria, but limited numbers at early gestational ages required combining of 1992–1994 data for Mexican Americans.
Gestational age was calculated in completed weeks from the last menstrual period (LMP).7 As in previous studies, the distribution of birth weight from 25–35 weeks' gestation based on the LMP was not normal,5,6 raising questions about biologically implausible birth weights at those gestations. Clinical estimates of gestation were substituted in the births with implausible values at 25–35 weeks.5,6 After review, the optimal cutoff points for substitution were virtually identical to those published by Zhang and Bowes.5 The cutoffs ranged from a maximum of 1250 g at 25 weeks' gestation to 3600 g at 35 weeks' gestation (Table 1).
The data were restricted to gestational ages of 25–42 weeks, and 1641 births with missing birth weights were excluded (409 Mexican Americans, 1232 non-Hispanic whites). Nearly equal percentages of Mexican American and non-Hispanic white births were excluded for gestation exceeding 42 weeks (4.9 and 4.5%, respectively). Relatively more Mexican American births than non-Hispanic white births were excluded for unknown gestation (2.8 and 0.6%, respectively), but the percentile distributions of birth weights excluded were nearly identical. Birth weights were identical at 3430 g at the 50th percentile of excluded observations. A total of 29,229 births (1.0%) were based on the clinical estimates of gestation (10,817 Mexican Americans; 18,412 non-Hispanic whites), with no more than 4% at any gestational age based on the clinical estimate.
Fifth, tenth, 25th, 50th, 75th, 90th, and 95th percentiles for birth weight were determined for each ethnicity, infant gender, and maternal parity. Polynomial models were tested for best fit at the second to fifth degrees.5 Unweighted fourth-degree polynomial equations, which provided the best fit for these data, were derived, and birth weight values at each week of gestation were extrapolated from the smoothed curves. External validity of the fits was determined by comparison with raw data. Although only the tenth, 50th, and 90th percentile data are presented, polynomial equations at the other percentiles had similar statistical fits and graphic patterns.
Birth weights at the tenth, 50th, and 90th percentiles are presented by parity and gender for Mexican Americans in Table 2 and for non-Hispanic whites in Table 3. At 37–42 weeks' gestation, infants of multiparas were heavier than those of primiparas at each percentile. The magnitudes of differences by parity at the 50th percentile were less at earlier gestations, almost disappearing and reversing directions at 25–26 weeks for female infants. Term male infants were 130–160 g heavier than term female infants for both parities.
Comparisons varied somewhat more by gestational age and gender. Mexican American infants were heavier than non-Hispanic white infants between 30 and 37 weeks' gestation for all parities and all percentiles. Consistent crossover occurred when non-Hispanic white infants became heavier as they approached term and through 42 weeks' gestation. Beginning at 37 weeks, the differences in weights for infants of primiparas increased to more than 100 g by 40 weeks; differences were only slightly less for infants of multiparas.
Only selected, but representative, curves are shown, ie, the curves at the 50th percentile for infants born to primiparas (Figure 1). Increased birth weight between 30 and 37 weeks' gestation for Mexican American infants was apparent, but Mexican American and non-Hispanic white birth curves flattened at term.8
The external validity of the smoothed curves was excellent, with numbers classified as SGA or LGA deviating by less than 0.5% overall. Using the smoothed percentiles, 10.5% of Mexican American infants were SGA (n = 126,131) and 9.7% were LGA (n = 116,292), closely approximating the expected 10% at both extremes. A total of 23,947 Mexican American infants were LGA (20.6% of total LGA) who would not have been so classified if non-Hispanic whites had been the reference. For non-Hispanic whites, 10.3% were SGA (n = 230,458) and 9.8% were LGA (n = 219,829).
Mexican Americans have a birth weight distribution somewhat dissimilar from that of non-Hispanic whites, with infants born heavier than non-Hispanic whites between 30 and 37 weeks' gestation but smaller between 37 and 42 weeks' gestation. This crossover was found consistently at all percentile levels, for both infant genders, and by maternal parity.
Currently, only one set of reference data9 includes any appreciable percentage (26%) of Hispanic births and this percentage is based on identification of Hispanic ethnicity from a review of Spanish surnames in 1970–1976 vital statistics records in California. Those California references have been recommended for international use by the World Health Organization10 because they include births to minority women. However, a study of Arizona births3 in 1986–1987 compared births of SGA infants using the California reference9 and found that Mexican American birth weights were significantly higher and birth rates of SGA infants significantly lower (6.3%) than expected. A significant percentage of the Mexican American infants who might have been born with FGR were not identified.
The main effect of classifying Mexican American births using reference percentiles generated from “whites,” when non-Hispanic and Hispanic whites are combined,5,6,9 is to underidentify at term a nontrivial number of Mexican American infants who might be LGA. Those LGA infants might be at increased risk for overweight and obesity in childhood and later life.4
Our findings were consistent with those of other studies of patterns of fetal growth among ethnic groups.11,12 Mexican Americans give birth to fewer very large infants (4000 g or more) compared with non-Hispanic whites,11,13 whereas birth rates of low birth weight (LBW, less than 2500 g) infants among Mexican Americans are similar to those among whites and most other Hispanics (Central and South Americans, Cubans).1,13 This pattern is most similar to that reported for immigrant and native Chinese,12 in which Chinese infants are larger from 30 to 37 weeks' gestation and are smaller at term, compared with Canadian white infants.
It is possible, although not likely, that the crossover could be explained solely by errors in dating due to imprecise recall of the LMP or menstrual irregularities.14,15 However, our verification of inaccurate or incomplete data argues against that explanation. Mexican American and non-Hispanic white births show flattening of curves at term, as usual for curves describing distribution of birth weight for gestational age based on LMP.8
More likely explanations for differences in birth weight distributions are that Mexican American women differ, both physically and behaviorally, in ways that affect birth weight distributions at various points along the curve,1 although outcomes for Mexican American mothers generally are better than expected based on demographic profiles.16 Mexican American mothers are at least three times less likely to smoke cigarettes than are non-Hispanic whites,17,18 possibly accounting for the larger weights between 30 and 37 weeks' gestation, but clearly not explaining the crossover at term.
Lower birth weights at term might be related to physical characteristics of Mexican American women. Mexican American women tend to be short as teenagers and adults, creating possible constricted uterine environments at term.19,20 Birth weights for Mexican Americans are higher at those gestations (30–37 weeks) when constraint due to maternal small stature is unlikely but fall below birth weights of non-Hispanic whites as birth weight increases at term.
Although Mexican American mothers might be shorter, they also tend to be overweight, which might promote fetal growth.21 In keeping with this tendency of being overweight, glucose intolerance and diabetes during pregnancy are more prevalent among Mexican American mothers, exacerbating the risk of delivery of an LGA infant.22 Not only would those conditions affect uterine environment and fetal growth, the associated LGA and fatness in infants might have implications for later development of overweight among Mexican American children.22,23
Mexican American children have higher body mass indices than do non-Hispanic white children, as early as 3 years of age.19,23,24 Recent concern about effects of preterm delivery and FGR on adult obesity and later risk for chronic disease25,26 reinforces the need to evaluate United States childhood growth measures to address prenatal growth influences. Our study provides a reference for future studies in identifying infants at risk for FGR or overgrowth.
1. Mathews TJ, Ventura SJ, Curtin SC, Martin JA. Births of Hispanic origin, 1989–95. Monthly vital statistics report, vol. 46, no. 6, suppl. Hyattsville, Maryland: National Center for Health Statistics, 1998.
2. Ventura SJ, Martin JA, Curtin SC, Mathews TJ. Report of final natality statistics, 1996. Monthly vital statistics report, vol. 46, no. 11, suppl. Hyattsville, Maryland: National Center for Health Statistics, 1998.
3. Balcazar H. The prevalence of intrauterine growth retardation in Mexican Americans. Am J Public Health 1994;84:462–5.
4. Whitaker RC, Dietz WH. Role of the prenatal environment in the development of obesity. J Pediatr 1998;132:768–76.
5. Zhang J, Bowes WA Jr. Birth-weight-for-gestational-age patterns by race, sex, and parity in the United States population. Obstet Gynecol 1995;86:200–8.
6. Alexander GR, Himes JH, Kaufman RB, Mor J, Kogan M. A United States national reference for fetal growth. Obstet Gynecol 1996;87:163–8.
7. National Center for Health Statistics. 1994 natality data set. CD-ROM series 21, no. 4. Hyattsville, Maryland: National Center for Health Statistics, 1997.
8. Wilcox M, Gardosi J, Mongelli M, Ray C, Johnson I. Birth weight from pregnancies dated by ultrasonography in multicultural British population. BMJ 1993;307:588–91.
9. Williams RL, Creasy RK, Cunningham GC, Hawes WE, Norris FD, Tashiro M. Fetal growth and perinatal viability in California. Obstet Gynecol 1982;59:624–32.
10. WHO Expert Committee on Physical Status. Physical status: The use and interpretation of anthropometry. WHO technical report series 854. Geneva, Switzerland: World Health Organization, 1995.
11. Shiono PH, Klebanoff MA, Graubard BI, Berendes HW, Rhoads GG. Birth weight among women of different ethnic groups. JAMA 1986;255:48–52.
12. Wen SW, Kramer MS, Usher RH. Comparison of birth weight distributions between Chinese and Caucasian infants. Am J Epidemiol 1995;141:1177–87.
13. Fuentes-Afflick E, Lurie P. Low birth weight and Latino ethnicity: Examining the epidemiologic paradox. Arch Pediatr Adolesc Med 1997;151:665–74.
14. Kramer MS, McLean FH, Boyd ME, Usher RH. The validity of gestational age estimation by menstrual dating in term, preterm, and postterm gestations. JAMA 1998;260:3306–8.
15. Henriksen TB, Wilcox AJ, Hedegaard M, Secher NJ. Bias in studies of preterm and posterm delivery due to ultrasound assessment of gestational age. Epidemiology 1995;6:533–7.
16. Collins JW Jr, Shay DK. Prevalence of low birth weight among Hispanic infants with United States-born and foreign-born mothers: The effect of urban poverty. Am J Epidemiol 1994;139:184–92.
17. LeClere FB, Wilson JB. Smoking behavior of recent mothers, 18–44 years of age, before and after pregnancy: United States, 1990. Advance data from vital and health statistics, no. 288. Hyattsville, Maryland: National Center for Health Statistics, 1997.
18. Camilli AE, McElroy LF, Reed KL. Smoking and pregnancy: A comparison of Mexican-American and non-Hispanic white women. Obstet Gynecol 1994;84:1033–7.
19. Roche AF, Guo S, Baumgartner RN, Chumlea WC, Ryan AS, Kuczmarski RJ. Reference data for weight, stature, and weight/stature2
in Mexican Americans from the Hispanic Health and Nutrition Examination Survey (HHANES 1982–1984). Am J Clin Nutr 1990;51:917S–24S.
20. Najjar MF, Kuczmarski RJ. Anthropometric data and prevalence of overweight for Hispanics: 1982–1984. Vital and health statistics series 11, no. 239 (DHEW publication [PHSM] 89-1689). Washington, DC: U.S. Government Printing Office, 1989.
21. Winkleby MA, Kraemer HC, Ahn DK, Varady AN. Ethnic and socioeconomic differences in cardiovascular disease risk factors: Findings for women from the Third National Health and Nutrition Examination Survey, 1988–1994. JAMA 1998;280:356–62.
22. Hollingsworth DR, Vaucher Y, Yamamoto TR. Diabetes in pregnancy in Mexican Americans. Diabetes Care 1991;14:695–705.
23. Ogden CL, Troiano RP, Briefel RR, Kuczmarski RJ, Flegal KM, Johnson CL. Prevalence of overweight among preschool children in the United States, 1971 through 1994. Pediatrics [serial online] 1997;99:e1. Available at: http:/www.pediatrics.org/egi/content/full/99/4/e1
24. Mei Z, Scanlon KS, Grummer-Strawn LM, Freedman DS, Yip R, Trowbridge FL. Increasing prevalence of overweight among US low-income preschool children: The Centers for Disease Control and Prevention Pediatric Nutrition Surveillance, 1983 to 1995. Pediatrics [serial online] 1998;101:e12. Available at: http:/www.pediatrics.org/egi/content/full/101/1/e12
25. Strauss RS. Effects of the intrauterine environment on childhood growth. Br Med Bull 1997;53:81–95.
26. Barker DJ. The fetal origins of coronary heart disease. Acta Paediatr Suppl 1997;422:78–82.