Abstract
Objectives. We examined whether education-related inequalities were associated with the occurrence of low birthweight in 6 counties in southern China in the early and late 1990s.
Methods. The study population consisted of 111181 women (65669 in the early 1990s and 45482 in the late 1990s) in a population-based Perinatal Health Care Surveillance System. We used the χ2 test, logistic regression, and concentration index for our analyses.
Results. From the early to late 1990s, the mean maternal education level increased significantly, and the percentage of low-birthweight births declined among all groups, for both male and female births, and at all levels of the mother’s education. Relative to those with less than 9 years of formal education, there was a decreasing risk of low birthweight among those with 9 to 11 years of formal education (range in adjusted odds ratio=0.69–0.82) and with 12 or more years of formal education (range in adjusted odds ratio=0.51–0.74). Between the early and late 1990s, the concentration index changed from −0.0778 to −0.0656 for male births and from −0.0717 to −0.0813 for female births.
Conclusions. Education-related inequalities associated with low birthweight persisted from the early to the late 1990s in surveyed areas.
Low birthweight has been associated with poor growth in childhood and an increased incidence of adult diseases, including type 2 diabetes, hypertension, and cardiovascular disease.1,2 In the past 2 decades, a number of investigators have reported socioeconomic inequalities in low birthweight,3–11 and these inequalities are expected to affect health-related inequalities later in life. Most of these studies were implemented in developed countries,3–8 which differ greatly from developing countries in terms of low birthweight and socioeconomic conditions.
Education, an independent socioeconomic factor that has been associated with low birth-weight, has been widely studied as a predictor of health.12–16 In rural China, both education and health have improved in recent decades. From 1991 to 2000, the overall infant mortality rate in rural China decreased by 36%, from 58 per 1000 live births to 37 per 1000 live births,17 and from 1990 to 2000, the illiteracy rate in China decreased from 15.9% to 6.7%.18 However, little is known about the education-related inequalities associated with low birthweight and their time trend in China. We used an existing population-based Perinatal Health Care Surveillance System in China to examine the education-related inequalities associated with low birthweight in 6 rural counties in southern China.
METHODS
Data
As part of the evaluation of a community folic acid intervention program to prevent neural tube defects in China from 1993 to 1995,19 a population-based Perinatal Health Care Surveillance System was established in 30 project counties by the Institute of Reproductive and Child Health at Peking University Health Science Center. In the areas under surveillance, all women who were planning to be married or become pregnant were enrolled in the Perinatal Health Care Surveillance System and were followed throughout pregnancy. Infants born to these women were evaluated at birth and during a home follow-up visit on the 42nd day after birth by a health professional. The information collected through this system included maternal demographic characteristics, preconception health status, health care services utilization, perinatal health status, pregnancy outcome, and postpartum health status of the mother and infant.
We analyzed data collected during 1993, 1994, 1999, and 2000 in 6 rural counties comprising Jiaxing City, an administrative area in Zhejiang Province, near Shanghai. Because the prevalence of low birthweight in this population is 2.51% (2796 of 111 181), which is relatively low compared with other areas with similar demographics,20 we combined the data from 1993 and 1994 and referred to this period as the “early 1990s,” and combined data from 1999 and 2000 and referred to this period as the “late 1990s.”
Definitions
We defined low birthweight as weight at birth of less than 2500 g, as measured by trained health care professionals at township-level hospitals or higher during the first hour after delivery. We looked at the percentage of babies with low birthweight in each of 3 groups that corresponded to their mothers’ highest level of formal education (less than 9 years, 9–11 years, and 12 or more years). We described women’s characteristics by their education based on the following variables: place of residence (suburb of Jiaxing City or not), parity (primiparous or multiparous), occupation (farmer or nonfarmer), age at delivery (years), prepregnancy weight (kg), height (cm), weight gain during pregnancy (kg), prenatal visits (number), and utilization of prenatal care (at least 5 prenatal visits beginning in the first trimester vs less than 5 prenatal visits or no visits during the first trimester). We considered whether women had any high-risk medical conditions before or during pregnancy, including hypertension, diabetes, heart disease, nephropathy, liver disease, anemia, psychosis, hyperthyroidism, or tuberculosis, as diagnosed by a health care provider at a township-level hospital or higher. The proportions of preterm births and caesarean births were also presented by maternal education level in the early and late 1990s.
Data Collection, Entry, and Cleaning
Birthweight was recorded in the hospital record in the first hour after birth and entered into the Perinatal Health Care Surveillance System no later than 2 days after birth. All data were uploaded to the county data server and sent to the Institute of Reproductive and Child Health at Peking University Health Science Center for checking and cleaning. Records that were incomplete or had logical errors were returned to the Maternal and Child Health Care Units for correction, to ensure completeness and accuracy.
Statistical Analysis
We used the χ2 test to compare the proportion of low-birthweight infants among the 3 educational groups. We conducted multivariate logistic regression to assess the association between maternal education and low birthweight after we controlled for place of residence. Women’s education may have influenced other variables including women’s occupation, parity, age at delivery, utilization of prenatal care, morbidities during the pregnancy, weight gain during pregnancy, and caesarean delivery; these variables were considered as potential mediating factors rather than as confounders. Concentration index and its standard error were calculated to measure the education-related inequalities in low birthweight according to the grouped education variable.21,22
RESULTS
During the study period, 113098 women delivered live-born singleton infants. We excluded a total of 1917 births with missing data for 1 or more of the following variables: birth-weight (1345 missing), mother’s highest educational achievement (1602 missing), and newborn’s gender (110 missing), which left 111181 (98%) women (32217 in 1993, 33482 in 1994, 23828 in 1999, and 21654 in 2000) and newborns in the study group. The number of infants born in each year and their gender distribution are shown in Table 1 ▶.
TABLE 1—
Gender Distribution of Births in 6 Rural Counties in Southern China, by Year: Perinatal Health Care Surveillance System, 1993–1994 and 1999–2000
| Year | Total Births, No. (%)a | Boys, No. (%) | Girls, No. (%) | Boy-to-Girl Ratio |
| 1993 | 32 217 (28.9) | 16 767 (52.0) | 15 450 (48.0) | 1.08 |
| 1994 | 33 482 (30.1) | 17 500 (53.2) | 15 982 (47.7) | 1.12 |
| Total 1993–1994 | 65 699 (59.1) | 34 267 (52.1) | 31 432 (47.8) | 1.09 |
| 1999 | 23 828 (21.4) | 12 205 (51.2) | 11 623 (48.8) | 1.07 |
| 2000 | 21 654 (19.5) | 11 361 (52.5) | 10 293 (47.5) | 1.12 |
| Total 1999–2000 | 45 482 (40.9) | 23 566 (51.8) | 21 916 (48.2) | 1.07 |
| Total | 111 181 (100.0) | 57 833 (52.0) | 53 348 (48.0) | 1.08 |
aPercentage of all births.
The distribution of maternal characteristics is presented in Table 2 ▶. The proportion of women with 12 or more years of formal education increased significantly from 7.0% in the early 1990s to 11.9% in the late 1990s. From the early to the late 1990s, the proportion of women who had both prenatal care initiation during the first trimester and at least 5 prenatal visits increased by approximately 4 times. The proportion of women with caesarean delivery increased markedly from 21.5% in the early 1990s to 58.3% in the late 1990s. Both maternal prepregnancy height and weight were similar during the study period. In addition, more women were aged older than 30 years, multiparous, farmers, and had a high-risk pregnancy in the late 1990s than in the early 1990s. During the study period, higher maternal education was associated with a greater probability of being nulliparous, of being diagnosed with a high-risk pregnancy, with more prenatal visits, and caesarean deliveries, as well as with greater weight gain during pregnancy.
TABLE 2—
Characteristics of Pregnant Women (n = 118 181) in 6 Rural Counties in Southern China, by Year and Maternal Education Level: Perinatal Health Care Surveillance System, 1993–1994 and 1999–2000
| 1993–1994 | 1999–2000 | |||||||
| Mother’s Education, y | Mother’s Education, y | |||||||
| Characteristicsa | < 9 | 9–11 | ≥ 12 | Total | < 9 | 9–11 | ≥ 12 | Total |
| Living in suburb, no. (%) | 6 074 (19.0) | 5 690 (19.5) | 638 (13.8) | 12 402 (18.9) | 2 233 (20.7) | 5 021 (17.2) | 520 (9.6) | 7 774 (17.1) |
| Nulliparous, no. (%) | 22 715 (71.4) | 23 971 (82.3) | 3 899 (84.7) | 50 585 (77.1) | 4 641 (43.1) | 23 026 (78.8) | 5 110 (94.2) | 32 777 (72.2) |
| Farmers, no. (%) | 22 735 (71.3) | 16 316 (55.9) | 1 109 (24.1) | 40 160 (61.2) | 9 001 (83.3) | 21 014 (71.9) | 1 334 (24.6) | 31 349 (69.0) |
| High-risk pregnancy, no. (%) | 1 877 (6.0) | 2 016 (7.0) | 436 (9.8) | 4 329 (6.7) | 757 (7.1) | 2 311 (8.0) | 572 (10.6) | 3 640 (8.1) |
| Prenatal care utilization, no. (%)b | 5 695 (17.8) | 6 240 (21.4) | 960 (20.8) | 12 895 (19.6) | 7 637 (70.7) | 23 123 (79.1) | 4 220 (77.7) | 34 980 (76.9) |
| Preterm rate, no. (%) | 1 773 (5.6) | 1 497 (5.1) | 194 (4.2) | 3 464 (5.3) | 461 (4.3) | 1 057 (3.6) | 189 (3.5) | 1 707 (3.8) |
| Caesarian section rate, no. (%) | 5 873 (18.4) | 6 799 (23.3) | 1 467 (31.8) | 14 139 (21.5) | 5 122 (47.4) | 17 474 (59.7) | 3 898 (71.8) | 26 494 (58.3) |
| Maternal age, y | ||||||||
| No.c | 31 906 | 29 186 | 4 607 | 65 699 | 10 806 | 29 246 | 5 430 | 45 482 |
| Mean (SD) | 26.2 (3.7) | 25.6 (2.9) | 26.4 (2.9) | 26.0 (3.3) | 28.4 (4.0) | 26.0 (3.3) | 26.2 (2.4) | 26.6 (3.6) |
| Height, cm | ||||||||
| No.c | 2 736 | 25 388 | 4 159 | 56 907 | 9 169 | 26 542 | 4 848 | 40 559 |
| Mean (SD) | 157.9 (4.6) | 158.5 (4.5) | 158.9 (4.4) | 158.2 (4.6) | 158.1 (4.6) | 158.8 (4.4) | 159.2 (4.4) | 158.7 (4.5) |
| Prepregnancy weight, kg | ||||||||
| No.c | 26 835 | 24 971 | 4 128 | 55 934 | 9 099 | 26 432 | 4 836 | 40 367 |
| Mean (SD) | 52.8 (6.6) | 52.3 (6.5) | 51.1 (6.3) | 52.4 (6.6) | 54.1 (7.4) | 52.3 (6.8) | 51.2 (6.4) | 52.6 (7.0) |
| Weight gain during pregnancy, kg | ||||||||
| No.c | 26 314 | 24 568 | 4 071 | 54 953 | 9 019 | 26 213 | 4 788 | 40 020 |
| Mean (SD) | 9.2 (5.2) | 10.1 (5.2) | 11.4 (5.3) | 9.8 (5.3) | 11.7 (5.0) | 13.1 (5.1) | 14.3 (4.8) | 12.9 (5.1) |
| Prenatal visits | ||||||||
| No.c | 16 640 | 15 682 | 2 426 | 34 748 | 10 696 | 29 013 | 5 383 | 45 092 |
| Mean times (SD) | 7.6 (3.1) | 8.2 (3.0) | 8.7 (2.8) | 7.9 (3.0) | 9.1 (2.7) | 9.9 (2.4) | 10.3 (2.3) | 9.7 (2.5) |
a The number of women with missing data was 195 for parity, 70 for occupation, 1590 for high-risk pregnancy, 13 715 for prepregnancy height, 14 880 for prepregnancy weight, 16 208 for weight gain during pregnancy, and 31 341 for prenatal visits.
bReferent is those who began prenatal care during the first trimester and had at least 5 prenatal visits.
c The number of women for whom data were available.
The frequency of low birthweight by education and year is shown in Table 3 ▶. There was a decline in the percentage of low-birthweight babies among all groups, in both male and female births, and at all maternal education levels. Mothers with lower educational level were more likely to have infants with low birthweight than were mothers with more education. The percentage reduction in low birthweight from the early to the late 1990s was slightly higher among male than among female births, but it varied by maternal education. Results of the concentration index indicated that there were education-related inequalities in the occurrence of low birthweight for male and female births in both study periods. There was a slight increase in the concentration index from the early to the late 1990s for female births and a slight decrease for male births, but neither of these were statistically significant (2-tailed P >.50 for both male and female births).
TABLE 3—
Occurrence of Low Birthweight in 6 Rural Counties in Southern China, by Maternal Education: Perinatal Health Care Surveillance System, 1993–1994 and 1999–2000
| Male Births | Female Births | |||||||||
| 1993–1994 | 1999–2000 | 1993–1994 | 1999–2000 | |||||||
| Low-Birthweight Births, % (95% CI) | ORa (95% CI) | Low-Birthweight Births, % (95% CI) | ORa (95% CI) | Reduction in Low-Birthweight Births,b % | Low-Birthweight Births, % (95% CI) | ORa (95% CI) | Low-Birthweight Births, % (95% CI) | ORa (95% CI) | Reduction in Low- Birthweight Births,b % | |
| All births | 2.59 (2.42, 2.76) | 1.77 (1.61, 1.95) | 31.6 | 3.18 (2.99, 3.38) | 2.25 (2.06, 2.46) | 29.2 | ||||
| Mother’s education,y | ||||||||||
| <9 | 3.00 (2.75, 3.27) | 1.00 | 2.15 (1.78, 2.56) | 1.00 | 28.3 | 3.58 (3.29, 3.89) | 1.00 | 2.96 (2.52, 3.46) | 1.00 | 17.3 |
| 9–11 | 2.18 (1.96, 2.43) | 0.72 (0.63, 0.83) | 1.70 (1.50, 1.92) | 0.79 (0.63, 0.98) | 22.0 | 2.95 (2.67, 3.24) | 0.82 (0.72, 0.94) | 2.07 (1.84, 2.32) | 0.69 (0.57, 0.84) | 29.8 |
| ≥ 12 | 2.24 (1.69, 2.92) | 0.74 (0.56, 0.98) | 1.36 (0.98, 1.90) | 0.62 (0.43, 0.90) | 39.3 | 1.86 (1.34, 2.52) | 0.51 (0.37, 0.70) | 1.86 (1.38, 2.45) | 0.61 (0.44, 0.85) | 0.0 |
| Concentration indexc (SE) | −0.0778 (0.0169) | −0.0656 (0.0239) | −0.0717 (0.0155) | −0.0813 (0.0226) | ||||||
Notes. OR = odds ratio; CI = confidence interval.
aORs adjusted for maternal place of residence.
bCalculated as ([lbw{1993 – 1994} − lbw{1999 – 2000}]/[lbw{1993 – 1994]}) × 100%, where lbw is the rate of low birthweight.
c The concentration index reflects the experience of the entire population, being sensitive to the distribution of the population across socioeconomic groups.21
DISCUSSION
The occurrence of low birthweight was only 2.87% in 1993–1994 and 2.00% in 1999–2000, both of which were significantly lower than the level of low birthweight in China as a whole. For example, according to the United Nations Children’s Fund report, the proportion of low-birthweight infants from 1998 to 2004 in China was 4%.23 The possible explanation is that we only analyzed the data of singleton births, whereas the United Nations Children’s Fund report was based on the analysis of all births, including multiple births. In addition, the considerably higher economic level in the Jiaxing City over that of the rest of China as a whole may explain the relatively lower level of low birthweight in surveyed areas, because low birthweight is associated with socioeconomic factors.8,24
We also found substantial education-related inequalities associated with the occurrence of low birthweight in surveyed areas both in the early and late 1990s, as has been reported in some developed countries.3–8 Low birth-weight is associated with numerous maternal biological, psychosocial, and behavioral factors including poor nutrition, smoking, physical labor, and poor pregnancy health care.1,11 In China, rural women farmers have a higher risk of delivering a low birthweight infant, likely because of a lower level of knowledge regarding health during pregnancy and poor utilization of prenatal care.24,25 In our study, women with lower levels of education were more likely to be farmers, to have fewer prenatal care visits, and to gain less weight during pregnancy than were women who had more education, which may have resulted in the inequalities in low birthweight by level of education we found.
Our study showed that education-related inequalities associated with low birthweight persisted from the early to the late 1990s, despite improvements in all educational groups. When the concentration indices were compared, there was some indication that inequalities increased among female births and decreased among male births, but these differences were not statistically significant.
One possibility for this discrepancy in inequalities is that son preference in rural China (a cultural preference for a male child) might affect trends in birthweight, because male newborns tend to be heavier than female newborns, as our study confirmed. However, the roughly similar gender ratios at birth in the early and late 1990s in our study do not support the son-preference argument. Also, there were no obvious gender differences in prenatal care utilization, occupation, or preterm rates; the exception was the caesarean rate, which was about 2 percentage points higher among male than among female births in each educational group at the same period (data not shown).
A noteworthy finding that was not related to the main objective of these analyses was the remarkable increase in caesarean sections, from 21.5% to 58.3% in a 7-year period. This affected all educational groups, and women with higher levels of education had markedly higher rates in both time periods. It is interesting that, unlike what was reported in a study from Brazil where caesarean sections are also epidemic, increased rates in China were not associated with greater frequencies of preterm deliveries nor of low birthweight.26
The strengths of our analysis were that we used population-based surveillance data that covered more than 95% of all births, and we used a concentration index to measure the inequalities in low birthweight. The concentration index reflects the experience of the entire population, being sensitive to the distribution of the population across socioeconomic groups.
There are 3 potential weaknesses in our study. First was the lack of information on potential confounders such as income, a factor strongly associated with low birthweight.24 Second, some variables (e.g., prenatal visits and prenatal care utilization) had relatively high proportions of missing values, and no information was available on the content and quality of prenatal care. Third, the relatively low level of low birthweight may limit the generalization of the findings to other areas in which this condition is more common.
Our findings are useful for understanding the education-related inequalities in child health in rural China. Inequalities in low birthweight have health service resource implications, because low-birthweight infants are more likely to require intensive neonatal care.3 In addition, because birthweight affects health in later life, the inequalities in low birthweight may affect future health inequalities. Finally, education-related inequalities in low birthweight persisted in spite of an overall reduction in its frequency, suggesting that strategies to address and to monitor inequalities in child health should be implemented even before a child is delivered.
Acknowledgments
We would like to gratefully acknowledge the work and support of the leaders and health care professionals at all levels responsible for the operation of the Perinatal Health Care Surveillance System in Jiaxing City, Zhejiang Province, China.
Human Participant Protection No protocol approval was needed for this study.
Peer Reviewed
Contributors Y. Liu and Z. Li originated the study and supervised all aspects of its implementation. Both Y. Liu and J. Liu synthesized the analysis and led the writing. R. Ye, S. Li, and A. Ren assisted with the writing and revising.
References
- 1.World Health Organization. Low Birthweight: Country, Regional and Global Estimates. Available at: http://www.who.int/reproductive-health/publications/low_birthweight/low_birthweight_estimates.pdf. Accessed February 10, 2006.
- 2.Barker DJP. Fetal and Infant Origins of Adult Disease. London, England: BMJ Books; 1992.
- 3.Moser K, Li L, Power C. Social inequalities in low birth weight in England and Wales: trends and implications for future population health. J Epidemiol Community Health. 2003;57:687–691. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Muntaner C, Lynch JW, Hillemeier M, et al. Economic inequality, working-class power, social capital, and cause-specific mortality in wealthy countries. Int J Health Serv. 2002;32:629–656. [DOI] [PubMed] [Google Scholar]
- 5.Townsend P, Simpson D, Tibbs N. Inequalities in health in the city of Bristol: a preliminary review of statistical evidence. Int J Health Serv. 1985;15:637–663. [DOI] [PubMed] [Google Scholar]
- 6.Reading R, Jarvis S, Openshaw S. Measurement of social inequalities in health and use of health services among children in Northumberland. Arch Dis Child. 1993;68:626–631. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Oldroyd J. Low birth weight in South Asian babies in Britain: time to reduce the inequalities. Fukushima J Med Sci. 2005;51:1–10. [DOI] [PubMed] [Google Scholar]
- 8.Pattenden S, Dolk H, Vrijheid M. Inequalities in low birth weight: parental social class, area deprivation, and “lone mother” status. J Epidemiol Community Health. 1999;53:355–358. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Krieger N, Chen JT, Waterman PD, Soobader MJ, Subramanian SV, Carson R. Choosing area based socioeconomic measures to monitor social inequalities in low birth weight and childhood lead poisoning: The Public Health Disparities Geocoding Project (US). J Epidemiol Community Health. 2003;57:186–199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Bremberg S. Does an increase of low income families affect child health inequalities? A Swedish case study. J Epidemiol Community Health. 2003;57: 584–588. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Rutter DR, Quine L. Inequalities in pregnancy outcome: a review of psychosocial and behavioural mediators. Soc Sci Med. 1990;30:553–568. [DOI] [PubMed] [Google Scholar]
- 12.Wu J, Liu Y, Rao K, Sun Q, Qian J, Li Z. Education-related gender differences in health in rural China. Am J Public Health. 2004;94:1713–1716. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Wu J, Liu Y, Rao K, et al. Gender differences in education related health inequalities in Chinese northern rural areas: 1993 and 2001. J Epidemiol Community Health. 2003;57:506–507. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Borrell C, Regidor E, Arias LC, et al. Inequalities in mortality according to educational level in two large Southern European Cities. Int J Epidemiol. 1999;28: 58–63. [DOI] [PubMed] [Google Scholar]
- 15.Kalediene H, Petrauskiene J. Inequalities in life expectancy in Lithuania by level of education. Scan J Public Health. 2000;28:4–9. [DOI] [PubMed] [Google Scholar]
- 16.Silventoinen K, Lahelma E. Health inequalities by education and age in four Nordic countries, 1986 and 1994. J Epidemiol Community Health. 2002;56: 253–258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.China’s Health Statistics Yearbook of 2001. Beijing, China: Ministry of Health; 2002.
- 18.China’s Population Statistical Yearbook 2000. Beijing, China: China Statistics Press; 2001.
- 19.Berry RJ, Li Z, Erickson JD, et al. Prevention of neural-tube defects with folic acid in China. China–US Collaborative Project for Neural Tube Defect Prevention. New Engl J Med. 1999;341:1485–1490. [DOI] [PubMed] [Google Scholar]
- 20.Lin L, Liu Y, Zhang X, Mi J, Cao L. Sampling survey on low birth weight in China in 1998. Zhonghua Yu Fang Yi Xue Za Zhi. 2002;36:149–153. [PubMed] [Google Scholar]
- 21.Wagstaff A, Paci P, van Doorslaer E. On the measurement of inequalities in health. Soc Sci Med. 1991; 33:545–557. [DOI] [PubMed] [Google Scholar]
- 22.Kakwani N, Wagstaff A, van Doorslaer E. Socioeconomic inequalities in health: measurement, computation and statistical inference. J Econometrics. 1997; 77:87–103. [Google Scholar]
- 23.United Nations Children’s Fund. The State of the World’s Children 2006 (Table 2). Available at: http://www.unicef.org/sowc06/pdfs/sowc06_table2.pdf. Accessed October 8, 2006.
- 24.Zhang D, Xu X, Lu X, Qiu L. Multinomial logistic regression analysis on factors associated with low birth weight. Chin J Public Health. 2004;20:40–42. [Google Scholar]
- 25.Fang L, Lou C, Gao E. Study on prenatal care services of married childbearing women in China. Maternal Child Health Care China. 2005;20:1424–1426. [Google Scholar]
- 26.Barros FC, Victora CG, Barros AJ, et al. The challenge of reducing neonatal mortality in middle-income countries: findings from three Brazilian birth cohorts in 1982, 1993, and 2004. Lancet. 2005;365:847–854. [DOI] [PubMed] [Google Scholar]