Objectives. We examined the extent to which area socioeconomic inequalities in overall and cause-specific mortality among US children aged 1–14 years changed between 1969 and 2000.
Methods. We linked a census-based deprivation index to US county mortality data from 1969 to 2000. We used Poisson and log-linear regression and inequality indices to analyze temporal disparities.
Results. Despite marked declines in child mortality, socioeconomic gradients (relative mortality risks) in overall child mortality increased substantially during the study period. During 1969–1971, children in the most deprived socioeconomic quintile had 52%, 13%, 69%, and 76% higher risks of all-cause, birth defect, unintentional injury, and homicide mortality, respectively, than did children in the least deprived socioeconomic quintile. The corresponding relative risks increased to 86%, 44%, 177%, 159%, respectively from 1998–2000.
Conclusions. Dramatic reductions in mortality among children in all socioeconomic quintiles represent a major public health success. However, children in higher socioeconomic quintiles experienced much larger declines in overall, injury, and natural-cause mortality than did those in more deprived socioeconomic quintiles, which contributed to the widening socioeconomic gap in mortality. Widening disparities in child mortality may reflect increasing polarization among deprivation quintiles in material and social conditions.
Mortality among US children aged 1–14 years has declined sharply over the past 3 decades, from a rate of 54.3 deaths per 100000 population in 1969 to 21.2 in 2002.1,2 Conversely, 99.7% of children aged 1 year survived to age 15 in 2002, compared with 99.2% in 1969–1971 and 90.2% in 1900–1902.3 Despite the overall improvement in child mortality over the long term, contemporary social disparities remain quite marked. Black children experience approximately twice the mortality rate of White, Asian, and Hispanic children,1,2 and children in low-income families experience a higher mortality that is 2 to 3 times higher than their counterparts in high-income families.4
Reducing and ultimately eliminating health inequalities, including those in childhood mortality, is one of the most important priorities of the US Department of Health and Human Services, as specified in Healthy People 2010.5 Documentation of health disparities between the least and most deprived groups can tell us the extent to which child mortality can be reduced. Monitoring such disparities over time allows us to track progress toward achieving the Healthy People 2010 objectives and to evaluate the role of specific social, behavioral, and public policy interventions in reducing childhood mortality.
Because household socioeconomic data are not available in national mortality statistics, socioeconomic differentials in US child mortality cannot be readily examined. However, socioeconomic differentials can be analyzed by linking census socioeconomic data with child mortality statistics at the small-area level, such as counties. Although there are a few studies that have examined the substantial inverse association between socioeconomic status and child mortality, such analyses are rarely done in a temporal fashion.4,6–8 No studies, to our knowledge, have analyzed how US child mortality rates have changed in recent decades in relation to area deprivation. In this study, we used a comprehensive area-deprivation index to examine the extent to which socioeconomic inequalities in overall and cause-specific mortality among US children aged 1–14 years changed between 1969 and 2000.
METHODSTo analyze temporal socioeconomic inequalities in US childhood mortality, we used data from the National Vital Statistics System (NVSS) and the decennial census.1,9 Area socioeconomic patterns in child mortality were indirectly derived by linking county-level socioeconomic data from the 1990 decennial census with the NVSS data.1,9 We used a factor-based deprivation index that consisted of 17 census-based social indicators, which may be viewed as broadly representative of educational opportunities, labor force skills, economic, and housing conditions in a given county. Selected indicators of education, occupation, wealth, income distribution, unemployment rate, poverty rate, and housing quality were used to construct the index.10–12 The factor loadings (correlations of indicators with the index) ranged from 0.92 for 150% of the poverty rate to 0.45 for household plumbing.10,11 The US deprivation index was constructed at the county level for the 1970, 1980, and 1990 censuses and also at the census-tract and zip-code levels for the 1990 census.10–12 Substantive and methodological details underlying the construction of the US deprivation index are provided elsewhere.10,11
To analyze trends in childhood mortality by deprivation, we used the weighted population quintile distribution of the 1990 deprivation index that classified all US counties into 5 groups of approximately equal population size. The groups thus created ranged from being the most socioeconomically deprived (first quintile) to the least socioeconomically deprived (fifth quintile) population groups. The 1990 deprivation index was used to compute annual or 3-year moving average rates from 1969 through 2000. Specifically, the national mortality database1 was used to obtain county- and cause-specific childhood death data from 1969 through 2000, whereas county-specific population estimates, served as the denominators for computing mortality rates.9,13 Each of the 3097 counties in the mortality database was assigned to 1 of the 5 deprivation quintiles. In the case of Alaska and Hawaii, state-level—rather than county-level—data were used. The underlying cause-of-death categories we used were coded according to the International Classification of Diseases, 8th Revision (ICD–8) from 1969–1978; International Classification of Diseases, 9th Revision (ICD–9) from 1979–1998; and the International Classification of Diseases, 10th Revision (ICD–10) from 1999–2000.1
We used Poisson regression, as estimated by the GENMOD procedure in SAS,14 to estimate relative risks (RR) of mortality for each deprivation quintile and time period. Although mortality rates for each deprivation quintile were computed annually between 1969 and 2000, we modeled childhood mortality as a function of age (1–4, 5–9, 10–14 years), area deprivation, race/ethnicity, and gender for 3-year and 2-year time periods (1969–1971, 1972–1974, . . . 1996–1998, and 1999–2000) in order to reduce variability associated with annual rates and to provide more stable RR estimates. We used log-linear regression to calculate average annual exponential rates of decline in mortality for each deprivation quintile.4,15 We used an index of disparity, which approximated in relative terms the average deviation of the rates from the rate for the least deprived socioeconomic quintile, to summarize disparities over time across all deprivation quintiles.16–18 This relative mean deviation index of disparity was calculated as
(1)where Qri is the rate for the ith quintile (i=1,2,3,4,5), Qr5 is the rate for the fifth quintile, and I is the number of quintiles being compared. Additionally, a population attributable risk measure was computed in order to assess the extent of improvement in child mortality if all quintiles were to have the rate of the least deprived quintile.18,19
RESULTSDescriptive sociodemographic data in Table 1 ▶ show lower education, income, and occupation levels; lower health personnel per capita; and higher proportions of Black, American Indian/ Alaska Native (AIAN), rural, and southern residents in more deprived population groups.
1.Selected Sociodemographic Characteristics, by Socioeconomic Deprivation Quintile: United States, 1990
Socioeconomic Deprivation Quintiles Sociodemographic Characteristic Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 Race/ethnicity, % Black 20.79 14.39 9.63 9.00 7.03 Hispanic 10.35 9.27 5.50 12.27 7.18 Asian/Pacific Islander 1.19 1.72 1.53 4.87 5.24 American Indians/Alaska Natives 1.58 0.60 0.65 0.73 0.36 Foreign-born population, % 5.67 7.16 4.30 12.29 10.00 Non-English speaking population, % 2.90 2.89 1.48 4.44 2.71 Region, % Urban 56.57 72.84 74.19 84.31 87.93 South 62.25 42.31 29.65 21.85 18.60 Educational attainment, % High school graduates 64.21 72.55 76.90 78.40 83.48 College graduates 13.45 17.23 18.92 22.75 28.81 Median family income, $ 23 774 28 231 32 025 36 853 45 754 Median home value, $ 39 500 45 500 58 000 77 700 125 650 Employed in white-collar occupations, % 48.78 54.49 56.32 60.82 66.67 Families below poverty level, % 17.83 11.47 8.82 7.81 4.35 Population below 150% of the poverty threshold, %a 33.99 24.71 20.56 18.54 11.03 Unemployment rate, % 8.78 7.08 5.83 5.70 4.43 Single-parent households, % 11.68 9.92 8.92 8.91 7.35 No. of doctors per 10 000 population 17.30 21.44 21.31 22.15 25.06 No. of nurses per 10 000 population 55.29 71.84 79.46 80.35 92.04Figure 1 ▶ shows area socioeconomic differentials in overall child mortality rates between 1969 and 2000. The relevant numerator (death) and denominator (population) data by time period are provided in an appendix table (available as an online supplement to this article at: http://www.ajph.org). Although child mortality declined over time for all deprivation quintiles, more deprived socioeconomic quintiles had higher mortality rates than less deprived socioeconomic quintiles each year, and the socioeconomic gradients (relative mortality risk) generally increased over time. Child mortality in the most deprived through the least deprived socioeconomic quintiles declined at average annual rates of 2.32%, 2.50%, 2.60%, 2.86%, and 3.11%, respectively, between 1969 and 2000. Compared with children in the least deprived socioeconomic quintile, the mortality rate for children in the most deprived socioeconomic quintile was 52% higher (RR=1.52; 95% CI=1.48, 1.55) in 1969–1971, 65% higher (RR=1.65; 95% CI=1.61, 1.70) in 1988–1990, and 86% higher (RR=1.86; 95% CI=1.80, 1.92) in 1998–2000. The relative overall disparity in child mortality across deprivation quintiles widened from 25% in 1969–1970 to 34% in 1988–1990 and 43% in 1998–2000.
FIGURE 1—Trends in All-Cause Mortality Among Children Aged 1–14 Years, by Area Socioeconomic Deprivation Index (3-Year Moving Averages), Mortality Rates (a) and Relative Risks (b): United States, 1969–2000.
Table 2 ▶ summarizes changing area socioeconomic differentials in overall child mortality across different time periods, after adjusting for the effects of age, gender, and race/ ethnicity using Poisson regression. The relative risk of childhood mortality associated with area deprivation increased substantially. In 1969–1971, children in the 2 most deprived socioeconomic quintiles had 38% and 25% higher risks of mortality, respectively, than did children in the least deprived socioeconomic quintile. In 1999–2000, the corresponding relative risks were 72% and 50% higher, respectively. Even after we adjusted for socioeconomic deprivation levels, Black children experienced approximately 50% higher mortality throughout the study period than did White children.
2.Poisson Regression Analysis Showing Age- and Gender-Adjusted Relative Risks (RR) of All-Cause Mortality Among Children Aged 1–14 Years, by Socioeconomic Deprivation Quintile and Race/Ethnicity: United States, 1969–2000
Race/Ethnicity and Time Period Quintile 1, RR (95% CI) Quintile 2, RR (95% CI) Quintile 3, RR (95% CI) Quintile 4, RR (95% CI) Trend, Pa All childrenb 1969–1971 1.38 (1.35, 1.41) 1.25 (1.23, 1.28) 1.21 (1.18, 1.23) 1.16 (1.13, 1.19) <.001 1972–1974 1.38 (1.35, 1.41) 1.27 (1.24, 1.30) 1.22 (1.19, 1.25) 1.19 (1.16, 1.22) <.001 1975–1977 1.39 (1.35, 1.43) 1.28 (1.25, 1.31) 1.24 (1.20, 1.27) 1.17 (1.13, 1.20) <.001 1978–1980 1.41 (1.38, 1.45) 1.31 (1.28, 1.35) 1.24 (1.21, 1.28) 1.21 (1.18, 1.25) <.001 1981–1983 1.40 (1.36, 1.44) 1.28 (1.24, 1.31) 1.22 (1.18, 1.26) 1.17 (1.14, 1.21) <.001 1984–1986 1.44 (1.40, 1.49) 1.29 (1.25, 1.33) 1.25 (1.21, 1.29) 1.21 (1.17, 1.24) <.001 1987–1989 1.48 (1.43, 1.52) 1.37 (1.33, 1.41) 1.28 (1.25, 1.32) 1.24 (1.20, 1.27) <.001 1990–1992 1.64 (1.59, 1.69) 1.47 (1.42, 1.52) 1.37 (1.33, 1.41) 1.27 (1.22, 1.31) <.001 1993–1995 1.66 (1.61, 1.71) 1.47 (1.42, 1.52) 1.36 (1.32, 1.41) 1.27 (1.23, 1.31) <.001 1996–1998 1.72 (1.67, 1.78) 1.47 (1.42, 1.52) 1.40 (1.35, 1.44) 1.24 (1.20, 1.29) <.001 1999–2000 1.72 (1.65, 1.79) 1.50 (1.44, 1.57) 1.43 (1.37, 1.49) 1.25 (1.20, 1.31) <.001 White childrenc 1969–1971 1.38 (1.35, 1.42) 1.25 (1.22, 1.28) 1.20 (1.17, 1.23) 1.16 (1.13, 1.19) <.001 1972–1974 1.38 (1.35, 1.42) 1.25 (1.22, 1.28) 1.21 (1.17, 1.24) 1.17 (1.14, 1.20) <.001 1975–1977 1.37 (1.34, 1.41) 1.26 (1.23, 1.31) 1.21 (1.18, 1.25) 1.14 (1.11, 1.17) <.001 1978–1980 1.41 (1.37, 1.5) 1.29 (1.25, 1.33) 1.21 (1.17, 1.25) 1.19 (1.16, 1.2) <.001 1981–1983 1.41 (1.36, 1.45) 1.26 (1.22, 1.30) 1.20 (1.16, 1.24) 1.15 (1.11, 1.19) <.001 1984–1986 1.49 (1.44, 1.54) 1.29 (1.25, 1.34) 1.24 (1.20, 1.29) 1.21 (1.17, 1.25) <.001 1987–1989 1.53 (1.48, 1.59) 1.36 (1.32, 1.41) 1.28 (1.24, 1.33) 1.22 (1.18, 1.27) <.001 1990–1992 1.70 (1.64, 1.76) 1.45 (1.40, 1.51) 1.38 (1.33, 1.43) 1.27 (1.22, 1.31) <.001 1993–1995 1.72 (1.66, 1.79) 1.45 (1.40, 1.50) 1.37 (1.33, 1.43) 1.26 (1.21, 1.30) <.001 1996–1998 1.79 (1.73, 1.86) 1.48 (1.42, 1.54) 1.39 (1.34, 1.45) 1.26 (1.21, 1.31) <.001 1999–2000 1.78 (1.70, 1.87) 1.53 (1.46, 1.60) 1.45 (1.38, 1.52) 1.26 (1.20, 1.32) <.001 Black childrenc 1969–1971 1.17 (1.09, 1.24) 1.09 (1.02, 1.16) 1.05 (0.98, 1.12) 1.00 (0.93, 1.08) <.001 1972–1974 1.24 (1.16, 1.32) 1.24 (1.16, 1.33) 1.18 (1.10, 1.27) 1.21 (1.12, 1.31) <.001 1975–1977 1.29 (1.20, 1.38) 1.23 (1.14, 1.33) 1.23 (1.14, 1.33) 1.20 (1.10, 1.30) <.001 1978–1980 1.33 (1.23, 1.43) 1.33 (1.24, 1.44) 1.32 (1.22, 1.43) 1.22 (1.12, 1.33) <.001 1981–1983 1.30 (1.20, 1.40) 1.29 (1.19, 1.39) 1.24 (1.14, 1.35) 1.24 (1.14, 1.35) <.001 1984–1986 1.18 (1.10, 1.27) 1.15 (1.07, 1.24) 1.16 (1.07, 1.26) 1.12 (1.03, 1.22) <.001 1987–1989 1.20 (1.12, 1.29) 1.25 (1.16, 1.35) 1.17 (1.08, 1.27) 1.21 (1.11, 1.31) <.001 1990–1992 1.36 (1.27, 1.46) 1.40 (1.30, 1.51) 1.26 (1.16, 1.36) 1.17 (1.08, 1.27) <.001 1993–1995 1.34 (1.26, 1.44) 1.35 (1.26, 1.45) 1.18 (1.09, 1.27) 1.20 (1.12, 1.30) <.001 1996–1998 1.41 (1.31, 1.51) 1.31 (1.22, 1.41) 1.30 (1.20, 1.40) 1.10 (1.02, 1.20) <.001 1999–2000 1.48 (1.35, 1.62) 1.42 (1.29, 1.56) 1.37 (1.24, 1.51) 1.17 (1.05, 1.29) <.001 Asian/Pacific Islander and American Indian/Alaska Native childrenc 1969–1971 2.95 (2.53, 3.44) 2.28 (1.88, 2.77) 2.18 (1.77, 2.69) 1.67 (1.40, 2.00) <.001 1972–1974 3.40 (2.90, 3.98) 1.55 (1.24, 1.93) 1.81 (1.45, 2.26) 1.71 (1.43, 2.05) <.001 1975–1977 3.58 (3.07, 4.19) 1.44 (1.16, 1.79) 1.82 (1.47, 2.26) 1.88 (1.58, 2.24) <.001 1978–1980 2.90 (2.49, 3.83) 1.71 (1.41, 2.08) 1.92 (1.58, 2.34) 1.68 (1.42, 1.99) <.001 1981–1983 2.26 (1.95, 2.61) 1.26 (1.05, 1.52) 1.64 (1.37, 1.97) 1.35 (1.15, 1.57) <.001 1984–1986 2.21 (1.93, 2.54) 1.31 (1.11, 1.56) 1.44 (1.21, 1.72) 1.32 (1.15, 1.52) <.001 1987–1989 2.27 (1.99, 2.59) 1.51 (1.29, 1.77) 1.62 (1.38, 1.91) 1.39 (1.21, 1.59) <.001 1990–1992 2.38 (1.08, 2.72) 1.49 (1.26, 1.78) 1.44 (1.22, 1.71) 1.45 (1.27, 1.66) <.001 1993–1995 2.64 (2.31, 3.01) 1.83 (1.57, 2.13) 1.69 (1.44, 1.98) 1.56 (1.37, 1.78) <.001 1996–1998 2.30 (2.01, 2.62) 1.80 (1.54, 2.09) 1.56 (1.33, 1.83) 1.35 (1.19, 1.54) <.001 1999–2000 2.10 (1.78, 2.48) 1.16 (0.93, 1.44) 1.10 (0.89, 1.37) 1.42 (1.21, 1.67) <.001To examine whether socioeconomic differentials in child mortality varied by race/ethnicity, we also estimated race-specific models, as shown in Table 2 ▶. Because socioeconomic patterns were similar for boys and girls, gender-specific models are not shown in Table 2 ▶. Although socioeconomic differentials in childhood mortality increased over time for both White and Black children, the gradients were more pronounced and consistent for Whites than for Blacks. The relative risk of child mortality between the most and least deprived socioeconomic quintiles increased from 1.38 in 1969–1971 to 1.78 in 1999–2000 for White children, and it increased from 1.17 in 1969–1971 to 1.48 in 1999–2000 for Black children. Socioeconomic inequalities were largest for Asian/Pacific Islander (API) and AIAN children, but the differentials decreased over time because of changes in the ethnic composition of this heterogeneous group over time. In 1969–1971, API and AIAN children in the most deprived socioeconomic quintile had 3 times the mortality rate of their least socioeconomically deprived counterparts. The relative differential decreased to 2.1 times in 1999–2000. Owing to the unavailability of temporal mortality and population denominator data, mortality rates and relative risks could not be estimated at all for Hispanics or separately for API and AIAN children.
Trends in Mortality from Unintentional Injuries, Homicide, and SuicideInjury and violence are leading causes of childhood mortality and account for nearly 50% of all childhood deaths.1,4 Trends in child mortality from unintentional injuries and homicide are presented in Figure 2 ▶ and Figure 3 ▶, respectively. All deprivation quintiles showed a decreasing trend in unintentional injury mortality during 1969–2000, but socioeconomic gradients were larger for unintentional injury than for overall mortality. Unintentional injury mortality declined faster in the least deprived socioeconomic quintile than in the other deprivation quintiles, which contributed to the widening socioeconomic gap. The average annual rates of decline during 1969–2000 in the most deprived through the least deprived socioeconomic quintiles were 2.47%, 3.02%, 3.21%, 3.85%, and 4.30%, respectively. In 1969–1971, children in the most deprived socioeconomic quintile had a 69% higher rate (RR=1.69; 95% CI=1.63, 1.74) of unintentional injury mortality than did children in the least deprived socioeconomic quintile. The differential widened to 177% (RR=2.77; 95% CI=2.62, 2.93) in 1998–2000. The summary index of inequality showed a consistent increase, with an increase in the overall relative socioeconomic disparity from 37% in 1969–1971 to 83% in 1998–2000.
FIGURE 2—Trends in Unintentional Injury Mortality Among Children Aged 1–14 Years, by Area Socioeconomic Deprivation Index (3-Year Moving Averages), Mortality Rates (a) and Relative Risks (b): United States, 1969–2000.
FIGURE 3—Trends in Homicide Mortality Among Children Aged 1–14 Years, by Area Socioeconomic Deprivation Index (3-Year Moving Averages), Mortality Rates (a) and Relative Risks (b): United States, 1969–2000.
Child homicide rates increased consistently from 1969 through the early 1990s and then declined in the late 1990s. However, the socioeconomic gradients remained large. In 1998–2000, children in the most deprived socioeconomic quintile had a 159% higher homicide rate (RR=2.59; 95% CI=2.51, 2.68) than did children in the least deprived socioeconomic quintile. The rate for the most deprived socioeconomic quintile was only 76% higher in 1969–1971 (RR=1.76; 95% CI=1.73, 1.80) than that for least deprived socioeconomic quintile. The inequality index indicates generally increasing overall socioeconomic disparities in homicide rates during the study period, with the index value increasing from 54% in 1969–1971 to 94% in 1998–2000. The population attributable risk estimates indicated that child homicide rates in 1998–2000 would have declined by 48% if children in the more deprived socioeconomic quintiles experienced homicide rates similar to those of the least deprived socioeconomic quintile.
Suicide is one of the leading causes of death among children aged 5–14 years, although the mortality rate from this cause is considerably lower than those from unintentional injuries and homicide.1,4 Similar to homicide, suicide rates among children aged 5–14 years rose consistently between 1969 and 1994 before declining in the late 1990s. Suicide rates did not vary markedly by deprivation levels until the late 1980s. Although suicide rates were generally similar in 1969–1971, the rate in the most deprived socioeconomic quintile was 51% higher in 1990–1992 (RR=1.34; 95% CI=1.15, 1.88) and 55% higher in 1999–2000 (RR=1.55; 95% CI=1.24, 1.87) compared with the suicide rate for the least socioeconomic deprived quintile. The inequality index showed widening overall socioeconomic disparities in suicide rates, increasing from 6% in 1969–1971 to 37% in 1995–2000.
Trends in Mortality from Leading Medical or Biological CausesBecause of the small numbers of annual deaths from these causes, the second, third, and fourth socioeconomic deprivation quartiles were collapsed into a single category. Trends in childhood mortality from birth defects, cardiovascular diseases (CVD), and cancer are shown in Figures 4 ▶–6 ▶ ▶, respectively. Although birth defects, cardiovascular diseases, and cancer mortality declined substantially for all children, children in the least deprived socioeconomic quintile experienced steeper mortality declines than did those in the most deprived socioeconomic quintile. The average annual rates of decline in birth defects mortality during 1969–2000 for the most deprived through the least deprived socioeconomic quintiles were 2.56%, 2.89%, and 3.38% respectively. The corresponding annual rates of decline were 1.06%, 1.94%, and 1.75% for cardiovascular disease mortality and 2.79%, 3.06%, and 3.29% for cancer mortality. Although both birth defects and cardiovascular disease mortality rates generally increased as deprivation levels increased, the socioeconomic gradients in birth defects mortality increased during 1969–2000. In 1969–1971, children in the most deprived socioeconomic quintile had a 13% higher birth defects mortality rate (RR=1.13; 95% CI=1.04, 1.21) than did children in the least deprived socioeconomic quintile. The differential widened to 44% (RR=1.44; 95% CI=1.26, 1.61) in 1998–2000. The summary index of inequality increased, particularly in the 1990s, with the overall relative socioeconomic disparity in birth defects mortality increasing from 8% in 1969–1971 to 27% in 1998–2000.
FIGURE 4—Trends in Birth Defects Mortality Among Children Aged 1–14 Years, by Area Socioeconomic Deprivation Index (3-Year Moving Averages), Mortality Rates (a) and Relative Risks (b): United States, 1969–2000.
FIGURE 5—Trends in Cancer Mortality Among Children Aged 1–14 Years, by Area Socioeconomic Deprivation Index (3-Year Moving Averages), Mortality Rates (a) and Relative Risks (b): United States, 1969–2000.
FIGURE 6—Trends in Cardiovascular Disease (CVD) Mortality Among Children Aged 1–14 Years, by Area Socioeconomic Deprivation Index (3-Year Moving Averages), Mortality Rates (a) and Relative Risks (b): United States, 1969–2000.
Childhood cancer mortality did not vary appreciably by area deprivation, although socioeconomic patterns appeared to have changed in recent times, with higher mortality rates associated with higher deprivation levels—a pattern similar to that for overall adult cancer mortality rates.12 Pneumonia, influenza, and infectious disease mortality (data not shown) declined sharply for children in all deprivation quintiles, and although socioeconomic gradients remained quite steep throughout the study period, they generally diminished over time. In 1969–1971, children in the most deprived socioeconomic quintile had a 90% higher pneumonia and influenza mortality rate (RR=1.90; 95% CI=1.73, 2.07) and a 106% higher rate of infectious disease mortality (RR=2.06; 95% CI=1.85, 2.26) than did children in the least deprived socioeconomic quintile. The corresponding differentials in 1998–2000 had decreased to 61% (RR=1.61; 95% CI=1.22, 2.01) and 54% (RR=1.54; 95% CI=1.25, 1.82), respectively.
DISCUSSIONSocioeconomic disparities in mortality among children aged 1–14 years are as marked as those among infants and working-age adults.4,10,11,15 Yet, because childhood mortality rates are the lowest of all age-specific mortality rates and because national mortality statistics lack relevant socioeconomic information,1,4 socioeconomic disparities in child mortality are rarely documented, particularly in a temporal fashion. For this study, we used a comprehensive area-based deprivation index to analyze changing socioeconomic inequalities in US childhood mortality in the past 3 decades. Although all deprivation quintiles experienced dramatic reductions in mortality over the long term (which represents a major public health success), relative socioeconomic disparities not only remained large but also increased substantially during 1969–2000. For children in the most deprived socioeconomic group, contemporary patterns indicate an overall mortality rate of approximately twice that of children in the least deprived socioeconomic group, and an unintentional injury and homicide mortality rate of 2.6 to 2.8 times higher compared to those in the least deprived socioeconomic group. In fact, US child mortality rates would decline by one third overall and unintentional injury and homicide mortality rates by almost half if children in the more deprived socioeconomic quintiles experienced mortality rates similar to those of the least deprived socioeconomic quintile.
The pattern of marked and increasing socioeconomic disparities in US childhood mortality during the past 3 decades is consistent with the pattern of increasing disparities in mortality observed for US working-age adults and the elderly.10,11 Increasing or persistent social inequalities in child mortality have also been observed for Britain during 1979–2001.20–22 However, the social patterning in US childhood mortality differed from that observed for urban Canada, where the socioeconomic inequality in child mortality did not increase between 1971 and 1996.19
The socioeconomic differentials in child mortality we present are probably underestimated because we used counties rather than smaller and more homogeneous geographic areas, such as census tracts or neighborhoods, to define deprivation quintiles. Many US counties are large geographic areas or population units with substantial socioeconomic heterogeneity.10–12 Unfortunately, public-use national vital records data do not identify geographic areas smaller than counties to protect the confidentiality of individual information on the death certificates. The 1990 deprivation index has been shown to provide a stable socioeconomic classification of counties over time,10–12 and the use of the 1980 index produced child mortality trends similar to those based on the 1990 index. The deprivation indices for the 1970, 1980, and 1990 censuses were highly correlated.10–12 The correlation of the 1990 index with the 1970 and 1980 indices was 0.90 and 0.94, respectively. Thus, the sole use of the 1990 index to construct deprivation quintiles for the entire study period is unlikely to have caused any substantial area misclassification, and the general trend of increasing inequalities in child mortality holds regardless of which index is used. Caution should be exercised when comparing area variations in mortality with individual-level socioeconomic differentials. Equating differentials at the 2 levels may lead to the ecological fallacy, implying that the socioeconomic effects estimated at the aggregate area level are being interpreted as those occurring at the individual level.10,11
Widening socioeconomic disparities in child mortality may be related to increasing temporal inequalities in the material and social living conditions between deprivation quintiles, both in absolute and relative terms. Absolute differences between deprivation quintiles in income, wealth and assets (as measured by home ownership and median home value), poverty, unemployment, female-headed households, and health care personnel per capita increased between 1970 and 1990. Similarly, the gap in relative income disparity between deprivation quintiles widened markedly.11
Temporal socioeconomic disparities in overall child mortality are driven primarily by trends in injury mortality, although trends in other prominent “biological” causes of child mortality—such as cancer, birth defects, heart disease, pneumonia, and influenza—also have a significant impact.1,4 Although inequalities in injuries, particularly those in motor vehicle crashes, accidental drowning, and residential fires—the 3 most common types of fatal injuries—arise primarily from social and environmental factors (such as material disadvantage, poor housing or living conditions, inadequate attention to motor vehicle safety, nonuse of seat belts, drinking and driving), inequalities in mortality from biological or medical causes might reflect disparities in social conditions, nutritional factors, and health care services.1,4,23
The introduction of Medicaid in 1965, a federally funded public health insurance program for poor and deprived children and socially disadvantaged groups, may be a factor in reducing childhood mortality.4 Without Medicaid, the extent of socioeconomic disparities in child mortality would perhaps be even greater and the increase in overall inequality over time even more rapid.
The data presented here underscore the increasingly important role of area socioeconomic deprivation in producing health disparities in US childhood mortality, which have not only persisted but have widened over time. Such large and growing disparities in childhood mortality run counter to the Healthy People 2010 goals and have important implications for the magnitude of current overall health inequalities as well as for the future course of inequalities in adult health.5,20 Narrowing the socioeconomic gap in child mortality may require designing strategies that are not only aimed at improving child health services but also aimed at mitigating the effects of inequalities in material and social living conditions.4,20
Peer Reviewed
Note. The views expressed are the authors’ and not necessarily those of the Health Resources and Services Administration or the US Department of Health and Human Services.
Contributors G.K. Singh planned and designed the study, analyzed the data, and wrote the article. M.D. Kogan contributed to the analysis and to the writing of the article.
Human Participation Protection No protocol approval was needed for this study.
ReferencesRetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.3