|
 |
|
|||||||
|
|||||||||
© 2006 American Public Health Association
DOI: 10.2105/AJPH.2004.049312
RESEARCH AND PRACTICE |
Quanhe Yang is with the Division of Birth Defects and Developmental Disabilities, National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, Atlanta, Ga. Sander Greenland is with the Departments of Epidemiology and Statistics, University of California, Los Angeles. W. Dana Flanders is with the Department of Epidemiology, School of Public Health, Emory University, Atlanta, Ga.
Correspondence: Requests for reprints should be sent to Quanhe Yang, PhD, Division of Birth Defects and Developmental Disabilities, National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, 1600 Clifton Rd, Mail Stop E-86, Atlanta, GA 30333 (e-mail: qay0{at}cdc.gov).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Objectives. We assessed the effects of changes in the maternal age–parity distribution and age-and parity-specific low-birthweight rates on low-birthweight trends in the United States.
Methods. We used natality file data from 1980 through 2000 to assess very-low-birthweight and low-birthweight rates among singleton live-born infants.
Results. Changes in age-and parity-specific low-birthweight rates were the main contributor to the overall trend in rates. However, changes in the age–parity distribution, primarily delayed childbearing, had a smaller but noticeable impact. The very-low-birthweight rate increased 27% among Black women, and changes in the age–parity distribution were associated with, on average, more than 20% of the increased rate during the 1990s. Among His-panic and non-Hispanic White women, on average, more than 10% of the rate increase observed during the 1990s was associated with changes in the age–parity distribution.
Conclusions. Assuming minimal changes in age-specific rates, delayed childbearing may play an increasingly important role in low-birthweight trends in the United States.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Rates of low birthweight (LBW) in the United States increased from 1980 through 2000.1,2 During the same period, percentages of births among teenage mothers declined, whereas percentages among mothers 35 years or older increased.2–5 Because there is a U-shaped relationship between maternal age and LBW,6–10 the decreasing percentages observed among teenage mothers should have lowered crude LBW rates over the period, and the increasing percentages among older mothers should have led to a rise in these rates. A number of studies4,6,8,9,11–19 have examined changes in maternal age, parity status, and LBW separately; in the present investigation, we assessed changes in these factors simultaneously.
The relationships between maternal age, parity status, and LBW are important. If age- and parity-specific LBW rates are constant over time, changes in secular LBW trends may predominantly reflect changes in maternal age and parity, and there may be little intervention potential apart from preventing teenage pregnancies. However, if age- and parity-specific LBW rates change over time, this may reflect shifts in medical practice, environmental exposures, socioeconomic status, or personal lifestyles. We assessed these 2 possible sources of change separately because age- and parity-specific rates are the primary target of public health interventions (e.g., prenatal care clinics) and can be used to assess racial disparities.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
We analyzed National Center for Health Statistics (NCHS) natality files for the period 1980 through 2000. We restricted the analysis to singleton live-born infants of mothers who: (1) were aged 15 to 49 years, (2) had delivered fewer than 16 infants, and (3) had complete information on birthweight, parity, race, and age. In the NCHS natality files, birthweights below 1500 g are classified as very low, and birthweights below 2500 g are classified as low. We calculated, separately for Black women and White women, rates for each age (15–49 years), calendar year (1980–2000), and parity level (0–15).
In conducting our analyses, we used a standardization and decomposition method introduced 50 years ago in the social sciences20 but as yet little used in epidemiology. This method can be used to factor the difference between 2 observed rates in a population at 2 separate time points into 2 components. Here one of these components reflected differences in age- and parity-specific LBW rates, and the other reflected differences in age-parity distribution.20 The former component addressed the extent to which rates would have changed if age- and parity-specific rates had changed as in fact observed but the age-parity distribution had remained constant (as, e.g., in the 1980 population); the latter component addressed the extent to which rates would have changed if the age-parity distribution changed as observed but age- and parity-specific LBW rates had remained constant. The first component indicates the effects of changes in age- and parity-specific LBW rates, and the second indicates the effects of changes in the age–parity distribution (of course, age–parity distribution "effects" include effects of factors associated with age and parity as well as age and parity themselves).
Our goal was to separate the difference between 2 given crude LBW rates into components associated with changes in the age–parity distribution and changes in age- and parity-specific LBW rates. The decomposition method can be described as follows: Let L1 and L2 be 2 crude rates for 1980 and 1990, respectively; let Rij1 and Rij2 be age- and parity-specific rates for 1980 and 1990 (i= 15, 16, 17, . . . 49 years of age and j = parity 1, 2, 3, . . . 15); let Nij1 and Nij2 be the number of births at the ith age and j th parity in 1980 and 1990; and let N++1 and N++2 be the total number of births in 1980 and 1990. Then L1 and L2 equal
 | (1) |
 | (2) |
A crude rate can be expressed as a weighted average of category-specific rates with a weight equal to the actual population studied.21 Thus, the difference between 2 LBW rates can be separated into differences resulting from changes in age- and parity-specific rates and differences resulting from changes in age–parity distributions20,22:
 | (3) |
Equation 3 is obtained via adding and subtracting
 |
from the difference L2 – L1. The proportions Nij1/N++1 and Nij2/N++2 for ages i = 15, 16,17, . . . 49 years and parity j= 1, 2, 3, . . . 15 represent the age–parity distributions in 1980 and 1990, respectively. The first term on the right-hand side of Equation 3 is the 1980-weighted average difference in rates within each age–parity subgroup. It represents the LBW rate change from 1980 to 1990 that would have ensued from the observed changes in age- and parity-specific rates if the age–parity distribution had remained the same as in 1980 (Nij1/N++1) (i.e., using the 1980 population as a standard). The second term is the difference in one rate standardized to the 1980 age–parity distribution and the same rate standardized to the 1990 distribution. It represents the LBW rate change from 1980 to 1990 that would have ensued from the observed changes in the maternal age–parity distribution if age- and parity-specific LBW rates had remained constant (Rij1 = Rij2).
Because both LBW and very-low-birthweight (VLBW) rates among White women increased after 1990 and the LBW rate among Blacks decreased after 1990, we applied the decomposition approach from 1980 through 1990 using the 1980 population as the standard; for 1990 through 2000, we used the 1990 population as the standard. From 1990 through 2000, when natality file data were available on Hispanic origin of mothers, we calculated results for non-Hispanic White women and Hispanic women separately.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
The NCHS natality files for the period 1980 through 2000 contain records on 78023 668 singleton births. As mentioned, we excluded births missing information on birthweight (0.12%) and parity (0.33%), as well as births among mothers who were younger than 15 or older than 49 years (0.28%), had delivered more than 15 infants (0.96%), and whose race/ethnicity was classified as "other" (4.7%). These exclusions left 73628288 births for the analysis. Of the 33533 795 infants included in the 1980 through 1989 period, 83% were White and 17% were Black. Of the remaining infants included during 1990 through 2000, 65.0% were non-Hispanic White, 16.4% were Black, and 18.6% were Hispanic.
Very-Low-Birthweight Rates
Among non-Hispanic White women, the VLBW rate increased 2% from 1980 through 1990 and increased 11% from 1990 through 2000; among Black women, it increased 27% from 1980 through 2000 (from 2.06% to 2.61%); and, among Hispanic women, it increased 10% from 1990 through 2000 (Table 1
). Rates of teenage births declined 21% (from 13.6% to 10.7%) among Whites and 26% (from 26.9% to 20.0%) among Blacks from 1980 through 2000. From 1980 to 2000, birth rates among White women 35 years or older increased 3-fold (from 4.5% to 13.6%), and rates among Black women 35 years or older more than doubled (from 4.0% to 9.6%).
|
As can be seen in Table 1
, the percentage point difference between the VLBW rate in 1990 and 2000 among non-Hispanic Whites was 0.08 (0.73% vs 0.81%). The (percentage point) portion of this difference due to changes in the maternal age–parity distribution was 0.005 (6% of the total difference), whereas the portion due to changes in the age- and parity-specific LBW rate was 0.074 (94% of the total difference). Among Blacks, the VLBW rate increased 18% from 1980 through 1990, and about 16% of the increase was attributable to changes in the age–parity distribution (Table 1).
Among non-Hispanic Whites, the portion of the increased VLBW rate between 1990 and 2000 due to changes in the age–parity distribution declined from a peak of 40% in 1994 to approximately 9% in 1999 (Figure 1a). Among Hispanics, changes in the age–parity distribution were associated with 10% (2000) to 46% (1991) of the increased VLBW rate (Figure 1a). The portion of the increased VLBW rate due to changes in the age–parity distribution among Blacks declined from a peak of 55% in 1994 to about 17% in 1999 (Figure 1b). Between 1980 and 2000, more than 90% of the increased VLBW rate due to changes in the age–parity distribution was a result of the increased proportion of births among mothers in all racial/ethnic groups who were 35 years or older (data not shown).
|
Low-Birthweight Rates
Similar to the patterns evident in VLBW rates, the relationships between age and LBW were U-shaped among both White and Black women. The percentage of LBW babies born to teenage mothers declined from 1980 through 2000 and increased among mothers 35 years or older during the same period.23
Among Whites, LBW rates declined between 1980 and 1990 and increased thereafter (Table 2). The latter trend was primarily a result of the 91% increase in the age- and parity-specific rate. Among Hispanics, the 3% increase in the LBW rate between 1990 and 2000 was entirely attributable to the change in the age- and parity-specific rate (Table 2). Among Blacks, LBW rates declined slightly between 1990 and 2000 (from 11.5% to 11.1%) (Table 2). If the age–parity distribution among Black women had remained the same in 2000 as in 1990, the decline would have been 3.6% instead of the 0.04% observed.
|
Among Whites, changes in the age–parity distribution were associated with more than 25% of the increased LBW rate during the mid-1990s (Figure 2a
). Among Hispanics, changes in the age–parity distribution were associated with more than a 10% increase during the 1990s (Figure 2a). Among Blacks, the increased proportion of mothers 35 years or older had an effect opposite that of declining age- and parity-specific LBW rates between 1990 and 2000 (Figure 2b). The slight increase in the LBW rate between 1990 and 1994 among Blacks (from 11.5% to 11.6%) was due to changes in the age–parity distribution (Figure 2b).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Changes in age- and parity-specific rates were the main contributor to increases in VLBW rates among both White and Black women during the study period, as well as increases in LBW rates among non-Hispanic White and Hispanic women between 1990 and 2000 and decreases in the LBW rate among Black women during the same interval. Changes in the age–parity distribution made smaller but important contributions to the secular trends observed, especially the VLBW trend among Blacks, for whom more than 20% of the increase in the VLBW rate was due to changes in the age–parity distribution between 1990 and 2000. Among non-Hispanic Whites, VLBW rates increased approximately 11% from 1990 through 1998, and more than 10% of this increase was due to changes in the age–parity distribution (Figure 1a
).
Similar to the results among non-Hispanic Whites, VLBW rates increased about 10% from 1990 through 2000 among Hispanics, and more than 10% of the increase was due to changes in the age–parity distribution (Figure 1a). A major contributor to the age–parity distribution was the increased proportion of births to mothers 35 years or older; the frequency of such births increased 3-fold among Whites and more than doubled among Blacks between 1980 and 2000.23 The expected impact of this increase in births to older mothers appeared to be partially offset by a decrease in the proportion of babies born to teenagers.
LBW infants have greatly elevated risks of morbidity and mortality.24–26 Mortality among LBW infants, who represented 7.6% of infants in the United States in 2000, accounted for 66% of overall infant mortality during that year.26 In addition, LBW infants, and especially VLBW infants, are at heightened risk of growth and developmental problems.27–30 Despite the risks associated with older maternal age at birth (including LBW), more women are delaying having children until relatively late in life,3,5,19,31,32 and the percentage of first births in which the mother was 30 through 40 years of age more than doubled from 1970 to 1990.5 Factors that have contributed to delayed childbearing include an aging population, women’s pursuit of advanced education, expanded roles for women in the workplace, advances in contraceptives, delayed and second marriages, and financial concerns.
Our results revealed that the LBW rate among Blacks was about twice that among Whites but that this racial disparity diminished between 1990 and 2000 as the LBW rate declined among Blacks and increased among Whites (Figure 2). The decline among Blacks apparently would have been greater if the age–parity distribution had remained constant from 1990 through 2000 (Table 2).
The NCHS natality files do not include data on other risk factors for LBW. We were not able to address potential causes of increased age- and parity-specific LBW rates, but possibilities are changes in lifestyles, environmental exposures, or obstetrical practices and decreases in the frequency of fetal deaths (leading to increases in preterm live births). Part of the increased LBW rate observed during the study period, especially in the 1990s, might be attributable to the increased use of assisted reproductive therapies, which, especially among women at relatively advanced ages, have been shown to be associated with increased LBW risk.33
The decomposition approach used in this study allowed us to separate the difference between 2 rates into additive components, and the approach is easy to use and interpret. Compared with statistical techniques such as linear regression analysis, it is less model dependent and involves fewer assumptions.34 Although it can be combined with statistical modeling,22 this approach seemed unnecessary here owing to the large numbers available. Of course, this decomposition does not necessarily reflect causal relationships; instead, it reflects the relative contribution of factors associated with standardization variables (here, age and parity) as opposed to other factors, as well as the changes in these factors over time. Also, the relative sizes of the components associated with the age–parity distribution and with age- and parity-specific rates are not unique; they depend on the choice of standard, which should reflect the targeted population of interest.20–22
As more women choose to delay childbearing, this trend will continue to play an important role in LBW rates. Nonetheless, it appears that trends in age- and parity-specific rates, which might involve much more intervention potential, are the largest contributor to recent changes in LBW rates. This finding underscores the importance of improvements in prenatal care, nutrition programs, and health education for pregnant women. It also suggests the value of programs aimed at older pregnant women, who may have heretofore received less attention than teenage mothers.
|
Acknowledgments |
|---|
We thank Hani Atrash, Owen Devine, Baoping Zhu, and Kenneth Schoendorf for their helpful comments on early drafts of the article.
Human Participant Protection
No protocol approval was needed for this study.
|
Footnotes |
|---|
Peer Reviewed
Contributors
Q. Yang and W. D. Flanders originated and designed the study. S. Greenland provided critical reviews and recommendations on the study design and data analysis. Q. Yang carried out the data analyses and wrote the drafts of the article. All of the authors contributed to interpretation of results and to article revisions.
Accepted for publication February 25, 2005.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
1. Centers for Disease Control and Prevention. Infant mortality and low birth weight among black and white infants—United States, 1980–2000. JAMA. 2002;288:825–826.
2. Martin JA, Hamilton BE, Sutton PD, Ventura SJ, Menacker F, Munson ML. Births: final data for 2002. Natl Vital Stat Rep. December 17, 2003;52(10).
3. Heck KE, Schoendorf KC, Ventura SJ, Kiely JL. Delayed childbearing by education level in the United States, 1969–1994. Maternal Child Health J. 1997;1: 81–88.
4. Branum AM, Schoendorf KC. Changing patterns of low birthweight and preterm birth in the United States, 1981–98. Paediatr Perinat Epidemiol. 2002;16: 8–15.[CrossRef][Web of Science][Medline]
5. Mathews TJ, Hamilton BE. Mean age of mother, 1970–2000. Natl Vital Stat Rep. December 11, 2002; 51(1).
6. Fraser AM, Brockert JE, Ward RH. Association of young maternal age with adverse reproductive outcomes. N Engl J Med. 1995;332:1113–1117.
7. Miller HS, Lesser KB, Reed KL. Adolescence and very low birth weight infants: a disproportionate association. Obstet Gynecol. 1996;87:83–88.[CrossRef][Web of Science][Medline]
8. Cooper LG, Leland NL, Alexander G. Effect of maternal age on birth outcomes among young adolescents. Soc Biol. 1995;42:22–35.[Web of Science][Medline]
9. Aldous MB, Edmonson MB. Maternal age at first childbirth and risk of low birth weight and preterm delivery in Washington State. JAMA. 1993;270: 2574–2577.
10. Ventura SJ, Mathews TJ, Hamilton BE. Births to teenagers in the United States, 1940–2000. Natl Vital Stat Rep. September 25, 2001;49(10).
11. Roth J, Hendrickson J, Schilling M, Stowell DW. The risk of teen mothers having low birth weight babies: implications of recent medical research for school health personnel. J Sch Health. 1998;68:271–275.[Web of Science][Medline]
12. Berkowitz GS, Skovron ML, Lapinski RH, Berkowitz RL. Delayed childbearing and the outcome of pregnancy. N Engl J Med. 1990;322:659–664.[Abstract]
13. Forman MR, Meirik O, Berendes HW. Delayed childbearing in Sweden. JAMA. 1984;252:3135–3139.
14. Cnattingius S, Forman MR, Berendes HW, Isotalo L. Delayed childbearing and risk of adverse perinatal outcome: a population-based study. JAMA. 1992;268: 886–890.
15. Cnattingius S, Berendes HW, Forman MR. Do delayed childbearers face increased risks of adverse pregnancy outcomes after the first birth? Obstet Gynecol. 1993;81:512–516.[Web of Science][Medline]
16. Prysak M, Lorenz RP, Kisly A. Pregnancy outcome in nulliparous women 35 years and older. Obstet Gynecol. 1995;85:65–70.[CrossRef][Web of Science][Medline]
17. Prysak M, Kisly A. Age greater than thirty-four years is an independent pregnancy risk factor in nulliparous women. J Perinatol. 1997;17:296–300.[Medline]
18. Tough SC, Svenson LW, Johnston DW, Schopflocher D. Characteristics of preterm delivery and low birthweight among 113,994 infants in Alberta: 1994–1996. Can J Public Health. 2001;92:276–280.[Web of Science][Medline]
19. Tough SC, Newburn-Cook C, Johnston DW, Svenson LW, Rose S, Belik J. Delayed childbearing and its impact on population rate changes in lower birth weight, multiple birth, and preterm delivery. Pediatrics. 2002;109:399–403.
20. Kitagawa EM. Components of a difference between two rates. J Am Stat Assoc. 1955;50:1168–1194.[CrossRef][Web of Science]
21. Rothman KJ, Greenland S. Modern Epidemiology. Philadelphia, Pa: Lippincott-Raven; 1998.
22. Greenl andand S. Model-based estimation of relative risks and other epidemiologic measures in studies of common outcomes and in case-control studies. Am J Epidemiol. 2004;160:301–305.
23. Martin JA, Hamilton BE, Ventura SJ, Menacker F, Park MM. Births: final data for 2000. Natl Vital Stat Rep. February 12, 2002;50(5).
24. McCormick MC. The contribution of low birth weight to infant mortality and childhood morbidity. N Engl J Med. 1985;312:82–90.[Abstract]
25. McIntire DD, Bloom SL, Casey BM, Leveno KJ. Birth weight in relation to morbidity and mortality among newborn infants. N Engl J Med. 1999;340: 1234–1238.
26. MacDorman MF, Minino AM, Strobino DM, Guyer B. Annual summary of vital statistics—2001. Pediatrics. 2002;110:1037–1052.
27. Brooks-Gunn J, McCarton CM, Casey PH, et al. Early intervention in low-birthweight premature infants: results through age 5 years from the Infant Health and Development Program. JAMA. 1994;272: 1257–1262.
28. McCarton CM, Brooks-Gunn J, Wallace IF, et al. Results at age 8 years of early intervention for low-birthweight premature infants: the Infant Health and Development Program. JAMA. 1997;277:126–132.
29. Ment LR, Vohr B, Allan W, et al. Change in cognitive function over time in very low-birthweight infants. JAMA. 2003;289:705–711.
30. Koller H, Lawson K, Rose SA, Wallace I, McCarton C. Patterns of cognitive development in very low birth weight children during the first six years of life. Pediatrics. 1997;99:383–389.
31. Ventura SJ. Trends and variations in first births to older women, United States, 1970–86. Vital Health Stat 21. 1989;No. 47.
32. Lutz W, O’Neill BC, Scherbov S. Demographics: Europe’s population at a turning point. Science. 2003; 299:1991–1992.
33. Schieve LA, Meikle SF, Ferre C, Peterson HB, Jeng G, Wilcox LS. Low and very low birth weight in infants conceived with use of assisted reproductive technology. N Engl J Med. 2002;346:731–737.
34. Das Gupta P. A general method of decomposing a difference between two rates into several components. Demography. 1978;15:99–112.[Web of Science][Medline]
This article has been cited by other articles:
|
|
 |
|
  A. H. Schempf, A. M. Branum, S. L. Lukacs, and K. C. Schoendorf The Contribution of Preterm Birth to the Black-White Infant Mortality Gap, 1990 and 2000 Am J Public Health, July 1, 2007; 97(7): 1255 - 1260. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Schempf and S. Becker ON THE APPLICATION OF DECOMPOSITION METHODS Am J Public Health, November 1, 2006; 96(11): 1899 - 1899. [Full Text] [PDF]
|
|
|
|
|
|
Q. Yang, S. Greenland, and W. D. Flanders YANG ET AL. RESPOND Am J Public Health, November 1, 2006; 96(11): 1899 - 1901. [Full Text] [PDF]
|
|
eLetters:
Read all eLetters
- This could be due to increasing testostrone: The Secular Trend
- James M. Howard
- AJPH Online, 5 Jun 2006 [Full text]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |