Vol. 41, No. 7, Supplement ● January 7, 1993 Monthly Vital StatisticsRe~ort — * Final Data From the CENTERS FOR DISEASE CONTROL AND PREVENTION/National Center for Health Statistics Advance Report of Final Mortality Statistics, 1990 Selected contents Highlights .................. 1 Deaths and death rates ..... 2 Expectation of life .......... 3 Cause ofdeath . . . . . . . . . . ...4 HIVinfection . . . . . . . . . . . ...8 Hispanic deaths ........... 8 Drug-induced deaths ...... 9 Alcohol-induced deaths .... 9 Marital status . . . . . . . . . . . . ...9 Educational attainment. ..... 9 lnfantmortality, . . ..o . . . . ...9 Hispanic infant mortality . . . 11 Maternal mortality ......... 11 Report of autopsy ......... 11 References ............... 11 List of detailed tables ...... 13 Technical notes ........... 47 Highlights In 1990, 2,148,463 deaths were registered in the United States, 19,536 fewer than the record high number of 2,167,999 deaths recorded in 198S and 2,003 fewer deaths than in 19S9. The death rate for 1990 was 863.8 deaths per 100,000 population, 1 percent below the rate of 871.3 in 19S9 and 3 percent below the rate in 1988. Pro- visional data for 1991 suggest that the death rate continued to decline. The age-adjusted death rate, which eliminates the effects of the aging of the population, was at a record low of 520.2 per 100,000 popu- lation, 1.5 percent below the 19S9 rate of 528.0 and 4 percent below the 19SS rate of 539.9. The age-adjusted rate decreased for the white and black pop- ulations between 1989 and 1990. For most of the 10-year age groups for males and all of the age groups for females, death rates declined between 19S9 and 1990, but they increased for males aged 15–24 years, 35-44 years, and slightly for those aged S5 years and over. The cause of death contrib- uting most to the increase in death rates for the white population and the black population aged 15–24 years was Homicide and legal intervention; and for the white population aged 35-44 years, Human immunodeficiency virus infection (HIV infection). In 1990, life expectancy at birth reached a record high of 75.4 years. The difference in life expectancy for males and females widened 0.2 years in 1990 after narrowing between 19S4 and 19S9. Women currently are expected to outlive men by an average of 7.0 years. For the white and black populations, life expectancy increased, narrowing the gap between the two race groups from 7.1 in 1989 to 7.0 years in 1990. Although life expectancy for bIack males increased in 1990 for the first time since 1984, it is still below the peak attained in 19S4. As in previous years, the first four leading causes of death accounted for almost 70 percent of deaths-Diseases of heart; Malignant neoplasms, including neopIasms of lymphatic and hematopoietic tissues (cancer); Cere- brovascular diseases (stroke); and Accidents and adverse effects. Age- adjusted death rates for 7 of the 15 Acknowledgments This report was prepared in the Division of Vital Statistics. Brenda S. Gillum and Bcttic L. Hudson, Nlortality Statistics Branch, wrote the report. Kenneth Kochanek provided special analysis tables. Thomas Dunn provided statistical tabks and content review, Charles E. Royer provided computer programming support. The Registration k4ethods Brmch and the Technical Scrviccs Branch provided consultation to State vital statistics offices regarding collection of the death crxtificatc data on which this report is based. This report was edited by Patricia Keaton-Williams and typeset by Jacqueline M. Davis of the Publications Branch, Division of Data Scrviccs. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES Public Health Service Centers for Disease Control and Prevention CK National Center for Health Statistics CENTSRS FOR DISEASE CONTROL AND PREVENTION
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Vol. 41, No. 7, Supplement ● January 7, 1993
Monthly VitalStatisticsRe~ort—
*
Final Data From the CENTERS FOR DISEASE CONTROL AND PREVENTION/National Center for Health Statistics
Advance Report of Final Mortality Statistics, 1990
In 1990, 2,148,463 deaths wereregistered in the United States, 19,536fewer than the record high number of2,167,999 deaths recorded in 198S and
2,003 fewer deaths than in 19S9. Thedeath rate for 1990 was 863.8 deathsper 100,000 population, 1 percentbelow the rate of 871.3 in 19S9 and3 percent below the rate in 1988. Pro-visional data for 1991 suggest that thedeath rate continued to decline.
The age-adjusted death rate,which eliminates the effects of theaging of the population, was at arecord low of 520.2 per 100,000 popu-lation, 1.5 percent below the 19S9 rateof 528.0 and 4 percent below the 19SSrate of 539.9. The age-adjusted ratedecreased for the white and black pop-ulations between 1989 and 1990. Formost of the 10-year age groups formales and all of the age groups forfemales, death rates declined between19S9 and 1990, but they increased formales aged 15–24 years, 35-44 years,and slightly for those aged S5 yearsand over. The cause of death contrib-uting most to the increase in deathrates for the white population and theblack population aged 15–24 years wasHomicide and legal intervention; and
for the white population aged 35-44years, Human immunodeficiency virusinfection (HIV infection).
In 1990, life expectancy at birthreached a record high of 75.4 years.The difference in life expectancy formales and females widened 0.2 yearsin 1990 after narrowing between 19S4and 19S9. Women currently areexpected to outlive men by an averageof 7.0 years. For the white and blackpopulations, life expectancy increased,narrowing the gap between the tworace groups from 7.1 in 1989 to 7.0years in 1990. Although life expectancyfor bIack males increased in 1990 forthe first time since 1984, it is still belowthe peak attained in 19S4.
As in previous years, the first fourleading causes of death accounted foralmost 70 percent of deaths-Diseasesof heart; Malignant neoplasms,including neopIasms of lymphatic andhematopoietic tissues (cancer); Cere-brovascular diseases (stroke); andAccidents and adverse effects. Age-adjusted death rates for 7 of the 15
Acknowledgments
This report was prepared in the Division of Vital Statistics. Brenda S. Gillum and Bcttic L. Hudson, Nlortality Statistics Branch, wrote the report.Kenneth Kochanek provided special analysis tables. Thomas Dunn provided statistical tabks and content review, Charles E. Royer provided
computer programming support. The Registration k4ethods Brmch and the Technical Scrviccs Branch provided consultation to State vital
statistics offices regarding collection of the death crxtificatc data on which this report is based. This report was edited by Patricia Keaton-Williamsand typeset by Jacqueline M. Davis of the Publications Branch, Division of Data Scrviccs.
U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICESPublic Health Service
Centers for Disease Control and Prevention CKNational Center for Health Statistics CENTSRS FOR DISEASE CONTROL
leading causes of death declinedbetween 1989 and 1990, led by reduc-tions for Atherosclerosis. Heart dis-ease, the leading cause of death in theUnited States, and stroke, the thirdleading cause of death, continued theirlong-term declines. Mortality forMotor vehicle accidents, a componentof Accidents and adverse effects, alsodeclined between 1989 and 1990,
Increases in age-adjusted deathrates for leading causes of deathbetween 1989 and 1990 were led byHIV infection and Homicide and legalintervention. Mortality from HIVinfection, which became the 10thleading cause of death in the UnitedStates, increased by 13 percent; andfrom Homicide and legal intervention,the llth leading cause, by 9 percent,Age-adjusted death rates alsoincreased for Suicide (1.8 percent);Pneumonia and influenza (1.4 per-cent); Diabetes mellitus (0.9 percent);Chronic pulmonary diseases and alliedconditions (0.5 percent); and Malig-nant neoplasms, including neoplasmsof lymphatic and hematopoietic tissues(0.4 percent). Among other causes, theage-adjusted death rate for alcohol-induced causes decreased by 1 percentbetween 1989 and 1990, while the ratefor drug-induced causes fell by12 percent.
The age-adjusted death rate formales was about 74 percent higherthan that for females for all causes ofdeath combined. For each of the 15leading causes of death, male mortalityalso was higher. The greatest sex dif-ferential was for HIV infection, wherethe age-adjusted rate for males was 8.4times that for females. The smallestsex differential was for Diabetes mel-Iitus, with a male-to-female ratio of1.1.
Overall, the black population hadage-adjusted death rates that exceededthose of the white population by about60 percent. Rates also were higher formost of the leading causes of death.The largest race differential continuedto be for Homicide and legal interven-tion, for which the age-adjusted ratefor the black population was aboutseven times that of the white popula-tion. The two leading causes that hadlower rates for the black population
were Chronic obstructive puhnona~ dis-eases and allied conditions and Suicide.
In 1990, there were differences inthe leading causes of death betweenthe Hispanic and the non-Hispanicwhite populations in an area com-prised of 45 States, New York State(excluding New York City), and theDistrict of Columbia, Although thetwo leading causes of death – Diseasesof heart and cancer —were the samefor the two population groups, sub-stantial differences exist in the rankingof other leading causes.
The infant mortality rate (9.2infant deaths per 1,000 live births)reached a record low in 1990. Thedecline in the rate occurred for whiteand black infants. Neonatal mortali~rates declined for white infants whilepostneonatal mortality rates declinedfor white and black infants. In 1990,the infant mortality rate for blackinfants remained at more than twicethat for white infants. The ratio ofblack-to-white infant mortality rateshas increased since the early 1970’swhen the mortality rates for blackinfants was 1.8 times that for whiteinfants.
Deaths and death rates
Death rates and life table valuesfor 1990 are based on the 1990 censusenumeration. All death rates and lifetable values for 1981–89 contained inthis report have been revised, usingintercensal population estimates basedon the 1990 census of population.Therefore, they are not comparablewith the death rates and life tablevalues in previous National Center forHealth Statistics (NCHS) reports andpublications containing those data for1981-89 (see Technical notes).
In 1990, a total of 2,148,463 deathsoccurred in the United States, 19,536fewer than in 1988 and 2,003 fewerdeaths than in 1989. The 1988 total of2,167,999 deaths is the largest finalnumber ever recorded. The decreasein the number of deaths between 1989and 1990 reflects lower mortality for anumber of age groups and causes ofdeath. The death rate for 1990, 863.8per 100,000 population, was 3 percentlower than the rate of 886.7 in 1988.
Provisional data for 1991 suggest thatthe death rate continued to decline (l).
The age-adjusted death rate of520.2 deaths per 100,000 population in1990 was a record low, about 4 percentbelow the rate of 539.9 for 1988 and11 percent below the rate for 1980 (fig-ure 1). Age-adjusted death rates showwhat the level of mortality would be ifno changes occurred in the age compo-sition of the population from year toyear. Thus, they are better indicatorsthan unadjusted death rates forshowing changes in the risk of deathover a duration when the age distribu-tion of the population is changing.Also, they are better indicators of riskwhen comparisons of mortality arebeing made for sex or race subgroupsof the population that have differentage compositions. Since 1980, the age-adjusted death rate has decreasedevery year except 1985 and 1988, yearswhen major influenza outbreaksincreased mortality in the UnitedStates.
Death rates by age, sex, and race
Between 1989 and 1990, deathrates for both sexes combinecl declinedfor all of the 10-year age groups except15–24 and 35-44 years (table A). Thelargest decrease (7 percent) occurredfor the age group 5-14 years, and thelargest increase (2 percent) occurredfor those aged 15–24 years.
For the white population, deathrates declined between 1989 and 1990for all age groups except 15-24 and35-44 years. The largest decrease(6 percent) occurred for those aged5–14 years, and the largest increase(1 percent) occurred for those aged35-44 years. For the black population,death rates declined for all age groupsexcept those aged 15–24 and 25-34years. The largest decline occurred forthe age group 1–4 years (6 percent),and the largest increase occurred forthose aged 15-24 years (6 percent).
The death rate for males alsodeclined between 1989 and 1!390for allage groups except 15–24, 25–34, 35-44years, and 85 years and over, Thelargest increase was for those aged15–24 years, although the change formales aged 85 years and over was very
1Death rateSunder I year (baSedon population estimates) differ from infant mortality rateS(based OnHve~~hs); see table Efor 1990 infant mortality rates and Technical notes for further discussion of the d!fferenca.
small. The increase in death ratesbetween 1989 and 1990 for males aged35-44 years (0.9 percent) continuedthe increase begun between 1983 and1984, a reversal of the downward trendfor this age group since the late 1960’s(figure 2 and table A). The death ratefor males aged 25–34 years wasunchanged.
The largest decreases for maleswere for those aged 5–14 years (7 per-cent), under 1 year (5 percent), and1–4 years (4 percent). Death rates forthe younger population aged 1-14years are subject to substantial fluctu-ation from year to year because of therelatively small number of deathsoccurring in these age groups. Deathrates at these ages are very low com-pared with other ages.
For females, death rates declined“ :tween 1989 and 1990 for all age
groups. This pattern is consistent withthe general downward trend inage-specific rates for females observedsince 1950. The largest decreases indeath rates between 1989 and 1990occurred for females aged 1–4 years(8 percent), under 1 year (7 percent),and 5–14 years (6 percent).
Age-adjusted death rates for whitemales decreased by 1 percent between1989 and 1990 and for white femalesby 2 percent. The age-adjusted deathrate decreased by 2 percent for blackmales and females. Age-adjusteddeath rates have decreased almostevery year since 1980 for white malesand females, resulting in record lowrates in 1990. Rates increased from1985 through 1989 for black malesbefore decreasing in 1990. Rates forblack females have fluctuated since1979.
In 1990,
~3
the age-adjusted deathrate for males of all races was 1.7 timesthat for females. In 1950 the male-to-femaIe ratio was 1.5. The 1970 ratio(1.7) increased to 1.8 during the late1970’s until 1987 when the ratio againdeclined to 1.7. For 1990, the ratiobetween male and female age-adjusteddeath rates was 1.7 for the white popula-tion and 1.8 for the black population.
In 1990, the age-adjusted deathrate for the black population was 1.6times that for the white population,the same ratio that has prevailed since1987. From 1960 through 1986, theratio was 1.5.
Expectation of life at birth andat specified ages
In 1990, the average expectationof life at birth reached a record high of75.4 years. This continued the generalupward trend in life expectancy in theUnited States.
The expectation of life at birth for1990 represents the average number ofyears that a group of infants would liveif the infants were to experiencethroughout Iife the age-specific deathrates prevailing in 1990. In 1990 lifeexpectancy for females was 78.8 yearscompared with 71.8 years for males;both figures represent increases over1989. The difference in life expectancybetween the sexes was 7.0 years in1990, higher than the difference of 6.8years in 1989. Historically, the differ-ence in life expectancy between thesexes has narrowed since 1979, in con-trast to widening from 1900 to 1972(table 4). The difference was 6.8 yearsin 1989 compared with 7.1 in 1984, 7.8years in 1975, 7.7 and 7.8 yearsthroughout the period 1972–79, 6.5years in 1960, and 2.0 years in 1900.
Between 1989 and 1990, lifeexpectancy for the white populationincreased from 75.9 years to a recordhigh of 76.1, and for the black popula-tion from 68.8 years to 69.1 years.Although the difference in life expect-ancy between the white and bIack pop-ulations narrowed from 7.6 years in1970 to 5.7 years in 1982, it increasedto 7.1 years in 1989 before declining to7.0 years in 1990. Although Iifeexpectancy for bIack males increased
~Death ra[es under 1 year (based on population estimates) differ from infant mortality rates (based on live births); see figure 5 for infant mortality rates and Technicalnotes ior further discussion of the dllference,
Figure 2. Death rates by age and sex: United States, 1950-90
for the first time since 1984, it is still0.8 years below the peak life expect-ancy of 65.3 years attained in 1984,
Among the four race-sex groups,white females continued to have thehighest life expectancy at birth (79,4years), followed by black females (73.6years), white males (72.7 years), andblack males (64.5) (figure 3). Between1989 and 1990, life expectancy increasedfor black males (born 64.3years in 1989to64.5 years in 1990) and for black females(from 73.3 in 1989to 73.6 in 1990).Before1990, life expectancy for black malesdeclined eve~ year from 1984 through1989. Life expectancy for black femaleshas fluctuated since 1982, showing noclear trend. Overall, the largest gain in life
expectancy behveen 1980 and 1990 was cause of deathfor white males (2,0 years), followed bywhite females (1:3 yems), ‘black female_s(1.1 years), and black males (0.7 years).
The 1990 life table may be used tocompare life expectancies at any agefrom birth onward. For example, aperson who has reached age 65 yearsmay look forward to living to an olderage, on the average, than one who hasreached 50 years. On the basis of mor-tality experienced in 1990, a personaged 50 years could expect to live anaverage of 29.0 more years for a totalof 79.0 years, and a person aged 65years could expect to live an average of17.2 more years for a total of 82.2 years.
Deaths assigned to the 15 leadingcauses accounted for 86 percent of thetotal number of deaths in 1990(table B). (For ranking procedures,see Technical notes.) The ranking ofthe first 9 of the 15 leading causes ofdeath for 1990 was unchanged fromthe list based on data for 1989. Theother six leading causes of death werethe same as for 1989; however, therankings changed for four of the six.HIV infection, the llth leading causein 1989, increased to 10th in 1990.Septicemia, the 14th leading cause in1989, increased to 13th in 1990.Nephritis,nephrotic syndrome, and nephrosis and
Figure 3. Life expectancy by sex United States, 1970-90
Certain conditions originating in theperinatal period retained the sameranks. Homicide and legal interventiondecreased in rank from 10th to Ilth,and Atherosclerosis decreased from13th to 14th.
Trends immortality basedon age-adjusted death rates are shown in fig-ure4 and table C for 14 of the 15leading causes. Because deaths fromCertain conditions originating in theperinatal period occur mainly among
infants under 1 year, age-adjustedrates for this cause are not discussed.
For seven of the leading causes ofdeath, the age-adjusted death ratesdecreased between 1989 and 1990(table C). These causes were Diseasesof heart; Cerebrovascular diseases;Accidents and adverse effects; Chronicliver disease and cirrhosis; Nephritis,nephrotic syndrome, and nephrosis;Septicemia; and Atherosclerosis. Thelargest decline was for Atherosclerosis
(10 percent).
~5
The declines for Dis-eases of heart, Cerebrovascular dis-eases, and Atherosclerosis areconsistent with the generally down-ward trends observed since 1950. Acci-dents and adverse effects has exhibiteda general downward trend since 1968,while Chronic liver disease and cir-rhosis has exhibited a downward trendsince 1973. Death rates for Pneumoniaand influenza increased slightly(1.4 percent) in 1990, a year of excessinfluenza-associated mortality in theelderly (2). Pneumonia and influenza,which had a general downward trendfor 1969–82, has shown an upwardtrend since the early 1980’s, marked bythe periodic effects of outbreaks thathave occurred each year (except 1987)since 1982. Septicemia decreased in1989 and 1990, in contrast to theupward trend since 1950.
Increases in the age-adjusteddeath rates between 1989 and 1990occurred for seven of the leadingcauses of death —Malignant neo-plasms, including neoplasms of lym-phatic and hematopoietic tissues;Chronic obstructive pulmona~ dis-eases and allied conditions; Pneu-monia and influenza; Diabetesmellitus; Suicide; HIV infection; andHomicide and legal intervention. Thelargest increase for 1990, 13 percent,was for HIV infection, which showed asmaller increase than in 1989 (30 per-cent) or in 1988 (22 percent). The age-adjusted death rate for Diabetesmellitus showed a much smallerincrease for 1990 (1.0 percent) com-pared to the 14 percent increase in1989. Diabetes mortality decreasedfrom the late 1960’s throughout the1970’s and remained at a plateau until1988. The age-adjusted death rate forHomicide and legal interventionincreased by 9 percent between 1989and 1990, but the rate was still 6 per-cent lower than the record high of1980. The age-adjusted death rate forChronic obstructive pulmona~ dis-eases and allied conditions increasedslightly between 1989 and 1990 (1 per-cent), continuing the long-termupward trend for this cause of death.There continued to be a large decreasein the age-adjusted death rate for Ath-erosclerosis; a decrease of 10 percent
Table C. Age-adjusted death rates for 1990 and percent changes In age-adjusted deathrates for the 15 leading causes of death from 1989 to 1990 and 1979 to 1990:United States
[Rates per 100,000 population]
Percent
change from -
Age-a@sted 1989 1979
Rank Cause of death (Ninth Revision International death rates to to
orderl Classification of Diseases, 1975) for 1990 1990 1990
Certain conditions originating intheperinatal period2. . . -8.5 -37.0
i Rank based on number of deaths see Technical nOteS.
‘Inesmuch as deaths from this cause occur mainly among infants, percent changaa are based on infant mortslity ratesInstsad of age.adjusted rates.
occurred between 1989 and 1990,aIthough it was not as large as thedecrease of 14percent between 1988and 1989. The very large decrease inmortality for Atherosclerosis and thevery large increase for Diabetes mel-Iitusthat occurred between 1988 and1989mayreflectin part changes intheway in which physicians report causesof death on the death certificate(3).
Mortality levels for each of the 15leading causes of death were higherformaIes than for females (table D).Ten of the leading causes of deathshowed differentials in which deathrates for males were at least 1.5 timesthose for females. The age-adjusteddeath rate for HIV infection, in whichthe death rate for males was8.4 timesthat for females, was the largest differ-ential. Other large differentials werefor Suicide (4,2); Homicide and legalintervention (3.9); Accidents andadverse effects (2.7); Chronic Iiverdis-ease and cirrhosis (2.3); Diseases ofheart and Chronic obstructive pulmo-nary diseases and aIIied conditions (1.9each); Pneumonia and influenza (1.7);Nephritis, nephrotic syndrome, andnephrosis (1,5); and Malignant neo-plasms, including neoplasms oflymphatic and hematopoietic tissues
(1,5). The smallest sex difference inmortality was for Diabetes mellitus,with a male-to-female ratio of 1.1.
For females, the age-adjusteddeath rate for all causes combineddecreased by 2 percent between 1989and 1990. Decreases in rates occurredfor six of the leading causes of death–
Atherosclerosis, Accidents and ad-verse effects, Chronic liver disease andcirrhosis, Diseases of heart, Septi-cemia, and Cerebrovascular diseases.Increases in rates occurred for fiveleading causes led by HIV infection;Homicide and legal intervention;Pneumonia and influenza; Diabetesmellitus; and Malignant neoplasms,including neopIasms of lymphatic andhematopoietic tissues. Rates for threecauses were unchanged —Chronicobstructive pulmonary diseases andallied conditions; Suicide; andNephritis, nephrotic syndrome, andnephrosis. Among the leading causesof death, HIV infection had the largestpercent increase in age-adjusted ratesbetween 1989 and 1990 for women,16.7 percent.
For males, the age-adjusted deathrate for aIl causes combined decIinedby 1 percent between 1989 and 1990.Decreases occurred in the rates for sixof the Ieading causes of death –Ath-erosclerosis; Chronic liver disease andcirrhosis; Nephritis, nephrotic syn-drome, and nephrosis; Accidents andadverse effects; Diseases of heart; andCerebrovascular diseases. Increases inthe age-adjusted death rate occurredfor HIV infection; Homicide and legalintervention; Pneumonia and influ-enza; Suicide; Chronic obstructive pul-monary diseases and allied conditions;
Table D. Ratio of age-adjusted death rates for the 15 leadhrg causes of death by sex andrace: United States, 1990
Rat/o of-
Rank Cause of death, race, and sex (Ninth Revision International Maie toorder 1
Black toClassification of Diseases, 1975) female white
1.33 1.15Certain conditions originating In the perinatal period 2 . . . . . 1.25 3.10
1Rank based on number of deaths see Technical notes.Pinasmuch as deaths from these causas occur mainly among infants, ratios ara baaed on infant r_fIOtifity ratas Instead Of a9a-
Diabetes mellitus; and Malignant neo-plasms, including neoplasms of lym-phatic and hematopoietic tissues. Formales, the largest increase was forHIV infection (12 percent). Between1989 and 1990 the age-adjusted deathrate for Motor vehicle accidents con-tinued the decline observed between1988 and 1989. The increase in thedeath rate for males aged 15–24 yearsbetween 1989 and 1990 was principallydue to a 23-percent increase for Homi-cide and legal intervention for this agegroup (from 26.8 to 32.9 per 100,000population). For males aged 35-44years, the increase was primarily dueto a 15-percent increase in death ratesdue to HIV infection.
Mortality was higher for the blackpopulation than for the white popula-tion for most of the leading causes(table D). The largest differential wasfor Homicide and legal intervention,for which the age-adjusted death ratefor the black population was 6.7 timesthat of the white population. Othercauses for which the differential waslarge include HIV infection (3.2);Nephritis, nephrotic syndrome, andnephrosis (3.0); Septicemia (2.7); Dia-betes mellitus (2.4); Cerebrovasculardiseases (1.9); Chronic liver diseaseand cirrhosis (1,7); Pneumonia andinfluenza (1.5); and Diseases of heart(1.5). Age-adjusted rates for the blackpopulation were lower than those forthe white population for two leadingcauses of death —Chronic obstructivepulmonary diseases and allied condi-tions and Suicide.
For the white population,decreases in age-adjusted ratesoccurred for seven leading causes ofdeath. The largest decrease was forAtherosclerosis (10 percent). Of theleading causes that increased for thewhite population, the largest increaseswere for HIV infection (12.7 percent)and Homicide and legal intervention(9.3 percent). The increase in thedeath rate for the white populationaged 15–24 years between 1989 and1990 was primarily due to an increasein the rate for Homicide and legalintervention (from 8.2 to 9.9) for thisage group. For those aged 35-44 years,the increase in the death rate wasprimarily due to HIV infection.
For the black population, age-adjusted death rates were lower in1990 than in 1989 for 10 of the leadingcauses, including Accidents andadverse effects (8.5 percent); Nephri-tis, nephrotic syndrome, and nephrosis(8.5 percent); Diseases of heart(4.7 percent); Cerebrovascular dis-eases (4.3 percent); Chronic liver dis-ease and cirrhosis (4.2 percent);Septicemia (4 percent); Atheroscle-rosis (3.1 percent); Suicide (2.8 per-cent); Pneumonia and influenza(2.5 percent); and Chronic obstructivepulmonaq diseases and allied condi-tions (1,7 percent). Of the leadingcauses that increased for the blackpopulation, the largest increase wasfor HIV infection (11.7 percent).Other increases were for Homicideand legal intervention (9.7 percent);Malignant neoplasms, including neo-plasms of lymphatic and hematopoietictissues (1.4 percent); and Diabetesmellitus (1.2 percent). The increase inthe death rate for the black populationaged 15–24 years between 1989 and1990 was primarily due to an increasein the rate for Homicide and legalintervention (from 65.4 to 78.2).
HIV infection
In 1990, there were 25,188 deathsdue to HIV infection, 14 percent morethan the 22,082 deaths recorded in1989. Of these deaths, 64 percent(16,106 deaths) were for white males,24 percent (6,097 deaths) for blackmales, 6 percent (1,633 deaths) forblack females, and 5 percent (1,149deaths) for white females (table 14),The largest numbers for males andfemales were for age groups 25-34 and35-44 years. Although the numbers ofdeaths were greatest for white males,the age-adjusted death rates andalmost all age-specific death rates werehighest for black males, followed bywhite males, black females, and whitefemales.
Overall, HIV infection was rankedas the 10th leading cause of death in1990. For the black population itranked 7th, and for the white popula-tion it ranked 10th among the leadingcauses of death. For males of all racescombined, HIV infection was theeighth leading cause of death; for
females it did not rank among the 10leading causes.
By age, HIV infection rankedamong the 10 leading causes for ages1-4, 5-14, 15-24, 25-44, and ,45-64years. HIV infection ranked 8th forages 1-4 years, 10th for ages 5-14years, 6th for ages 15-24 years, 3rd forages 25-44 years, 10th for ages 45–64years. Although HIV infection was the8th leading cause for ages 1-4 years,the number of deaths due to this causewas relatively small – 123 deaths, or2 percent of deaths from all causes forthat age group.
Hispanic deaths
Leading causes of death for aII agegroups combined for the Hispanicpopulation differ from those for thenon-Hispanic white population in anarea comprised of 45 States, New YorkState (excluding New York City), andthe District of Columbia (table 17)(see Technical notes). Although thetwo leading causes of death – Diseasesof heart and Malignant neoplasms,including neoplasms of lymphatic andhematopoietic tissues – are the samefor both groups, they accounted forthree-fifths of ail deaths in 1990 for thenon-?+spanic white population but foronly two-fifths of the deaths for theHispanic population.
Major differences in leading
causes of death between the twogroups are as follows: Three of the 10leading causes of death for the His-panic population – Homicide and legalintervention (5th leading cause), HIVinfection (8th), and Certain conditionsoriginating in the perinatal period(lOth) – are not among the 10 leadingcauses for the non-Hispanic white pop-ulation. Conversely, the 3 leadingcauses for the non-Hispanic white pop-ulation not among the 10 leadingcauses for the Hispanic populationwere Chronic obstructive pulmonarydiseases and allied conditions (4thleading cause), Suicide (8th), and Ath-erosclerosis (lOth).
Differences in ranking the leadingcauses of death between the two pop-ulation groups largely reflect differ-ences in age composition between thetwo groups; within broad age groups,leading causes are similar. However,
even within age categories some differ-ences exist. Homicide and legal inter-vention consistently ranked higher forthe Hispanic population than for thenon-Hispanic white population for allage groups under 45 years. HIV infec-tion for the Hispanic population aged15-24, 25-44, and 45-64 consistentlyranked higher than for the non-Hispanic white population for thesesame age groups. Diabetes mellitusand Chronic liver disease and cirrhosisranked higher for Hispanics aged45-64 years.
Drug-induced deaths
In 1990, a total of 9,463 personsdied of drug-induced causes in theUnited States (table 18). The catego~drug-induced causes includes not onlydeaths from dependent and nondepen-dent use of drugs (legal and illegaluse), but also poisoning from medicallyprescribed and other drugs; it excludesaccidents, homicides, and other causesindirectly related to drug use (fordrug-induced causes, see Technicalnotes). The age-adjusted death ratefor drug-induced causes in 1990 was3.6 deaths per 100,000 population. Therate increased by 32 percent from 1983to 1988, then declined 2 percent in1989 and 12 percent in 1990. The age-adjusted death rate for drug-inducedcauses for males was 1.8 times the ratefor females, and the rate for the blackpopulation was 1.7 times that for thewhite population.
Alcohol-induced deaths
In 1990, a total of 19,757 personsdied of alcohol-induced causes in theUnited States (table 19). The categoryalcohol-induced causes includes notonly deaths from dependent and non-dependent use of alcohol, but alsoaccidental poisoning by alcohol; itexcludes accidents, homicides, andother causes indirectly related toalcohoI use (for alcohol-inducedcauses, see Technical notes). The age-adjusted death rate for alcohol-induced causes in 1990 was 7.2 deathsper 100,000 population, which is lowerthan the rate of 7.3 for the previousyear. The rate decreased by 20 percentfrom 1980 to 1986, increased by
9 percent from 1986 to 1989, thendecreased by 1 percent in 1990. Theage-adjusted death rate for alcohol-induced causes for males was 3.4 timesthe rate for females, and the rate forthe black population was 2.6 times therate for the white population.
Marital status
About 90 percent of the personswho died in 1990 had been married(table 20). The proportion was largerfor females (92 percent) than formales (87 percent) and for the whitepopulation (90 percent) than for theblack population (81 percent). Theproportion who were widowed at thetime of death was considerably greaterfor women (57 percent) than for men(18 percent) but about the same forboth major race groups: 38 percent ofthe white population and 33 percent ofthe black population.
Educational attainment
About 55 percent of the personswho died in 1990 had completed highschool in an area comprised of 28States and the District of Columbia(table 21), (see Technical notes). Thepercent was about the same for males(54 percent) and females (57 percent),and similar for the white population(56 percent) compared to the blackpopulation (53 percent). About thesame proportion of white females(57 percent) and white males (55 per-cent) who died in 1990 had completedhigh school. The proportion who hadcompleted 4 years of college was sim-ilar for white females (3.8 percent) andfor white males (2.4 percent). Agreater proportion of black females(60 percent) had completed highschool than black males (51 percent).Slightly more black females (2.9 per-cent) than black maIes (1 percent) hadcompleted 4 years of college at thetime of death.
Infant mortality
A change in the tabulation by raceof live births beginning with data for1989 affects infant mortali~ rates byrace, because counts of live birthscomprise the denominator of infant
mortality rates. The tabulation of racefor live births changed from race ofchild to race of mother (see Technicalnotes). As a result, infant mortalityrates by race for 1989 and 1990 are notcomparable with those of previousyears. In general, changing the basisfor tabulating birth data from race ofchild to race of mother results in morewhite births and fewer black births andbirths of other races. Therefore, infantmortality rates based on live birthstabulated by race of mother tend to belower for white infants and higher forblack infants and infants of other racesthan they are when computed on thebasis of live births tabulated by race ofchild. (For additional detail, see Tech-nical notes.)
To facilitate comparison with pre-vious years’ data and analysis of currentpatterns, key tabulations and text anal-ysis in this and other reports fromNCHS for 1989 and 1990 infant mor-tality will show data computed on thebasis of live births tabulated by race ofmother and race of child. This willmake it possible to distinguish theeffects of this change in tabulationfrom real changes in the data. Accord-ingly, rates for 1989 and 1990 are pre-sented in figure 5 and table 22 by bothmethods of tabulating race of livebirths. In this report, the discussions ofchanges in various measures among1990 and years before 1989 are basedon measures computed by race ofchiId. Text references to white andblack infant mortality for 1990 onlyrefer to rates with Iive births tabulatedaccording to race of mother (table E).
In 1990, there were 38,351 deathsof infants under 1 year of age,(table E) compared with 39,655 in theprevious year. The infant mortalityrate of 9.2 infant deaths per 1,000 livebirths was the lowest final rate everrecorded for the United States; it rep-resented a decline of 6 percent fromthe rate of 9.8 for the previous year. In1990, the mortali~ rate for whiteinfants declined 6 percent (8.1 in 1989,compared with 7.6 in 1990) and forblack infants, by 3 percent (18.6 in1989, compared with 18.0 in 1990).
In 1990, the infant mortality ratefor black infants (18.0) was 2.4 timesthe rate for white infants (7.6), slightly
NOTE: Infant dealhsare classll,ed byraceof de.-endent. Beginning in 1989. live birlhsare classified by
race of molhec from 1950-89, by race 01 child; see Technical notes,
Figure 5. Infant mortality rates by race: United States, 1950-90
Table E. Infant, neonatal, and postneonatal deaths and mortality rates by race and sex:United States, 199o
[Rates are Infant (under 1 year), neonatal (under 28 days), and postneonatal (28 days-11 months) deatha per100,000 live births in specified group. Beginning in 1969, race for live births IS tabulated according to race ofmothep see Technical notes]
higher than the ratio of 2,3 in theprevious year, When based on livebirths by race of child, the ratio ofblack-to-white infant mortality in 1990was 2.2; the ratio has increased sincetheearIy1970’s, when it was 1.8.
Between 1989 and 1990, the neo-natal mortality rate declined by6.5 percent, from 6.2 to 5.8 deaths forinfants under 28 days per J.,000 livebirths. For white infants the ratedeclined from 5.1 to 4.8, while forblack infants, the change in the ratebetween the tsvo successive years wasnot statistically significant (11.9 in 1989compared with 11.6 in 1990), Neonatalmortality rates historically havedeclined for both races – an averageannual decrease of 4 percent per yearfor white infants compared with 3 per-cent for black infants between 1960and 1990 by race of child,
The postneonatal mortality rate–deaths to infants 28 days-n monthsper 1,000 live births– declined 5.6 per-cent for 1990 from 3,6 to 3.4 deaths.For white infants the postneonatalmortality rate declined 3 percent, from2.9 to 2.8 deaths per 1,000 live births,while for black infants the ratedeclined 4 percent, from 6.7 in 1989 to6.4 in 1990, This follows the historicaltrend, by race of child, of more rapiddeclines in postneonatal mortality forblack than for white infants. From1960 to 1990 postneonatal mortalitydeclined by an average of 3 percentper year for black infants comparedwith 2 percent for white infants.
Although the list of the 10 leadingcauses of infant death was unchangedin 1990, the rankings for 2 of theleading causes of infant death changedslightly between 1989 and 1990, New-born affected by complications of pla-centa, cord, and membranes increasedfrom the seventh leading cause in 1989to the sixth, Accidents and adverseeffects, the sixth leading cause in 1989,decreased to the seventh. Among the10 Ieading causes of infant death, thefirst 4– Congenital anomalies, Suddeninfant death syndrome, Disordersrelating to short gestation and unspec-ified low birthweight, and Respiratorydistress syndrome – accounted for justover half (54 percent) of all infantdeaths in 1990; and the remaining 6
causes accounted for only 15 percentof all infant deaths.
Between 1989 and 1990, the infantmortality rate decreased for 8 of the 10leading causes of infant death andincreased for the other 2 causes. Thelargest decreases were for Respiratorydistress syndrome (24 percent), Acci-dents and adverse effects (9 percent),and Sudden infant death syndrome(7 percent). The two increases werefor Newborn affected by maternalcomplications of pregnancy (5 percent)and Intrauterine hypoxia and birthasphyxia (2 percent).
Differences between infant mor-tality rates for white and black infantsby cause are reflected in differences inranking of the leading causes of infantdeath as well as in differences incause-specific infant mortality rates.Congenital anomalies was the leadingcause of death for white infants, fol-lowed by Sudden infant death syn-drome, Disorders relating to shortgestation and unspecified low birth-weight, and Respiratory distress syn-drome. Together these four causesaccounted for 56 percent of whiteinfant deaths. In contrast, for blackinfants the leading cause of death wasDisorders relating to short gestationand unspecified low birthweight, fol-lowed by Sudden infant death syn-drome, Congenital anomalies, andRespiratory distress syndrome. To-gether these four causes accounted for49 percent of all black infant deaths.
AIthough the difference betweenblack and white infant mortality ratesvaried by cause, the risk was higher forblack than for white infants for all theleading causes. Expressed as the ratioof the infant mortality rate for blackinfants to that for white infants, theleading causes ranked beginning withthe highest ratio are: Disordersrelating to short gestation and unspec-ified low birthweight (4,6); Pneumoniaand influenza (3.0); Respiratory dis-tress syndrome and Newborn affectedby maternal complications of preg-nancy (2.6 each); Infections specific tothe perinatal period (2,5); Accidentsand adverse effects (2.3); Intrauterinehypoxia and birth asphyxia (2.2); New-born affected by complications of pla-centa, cord, and membranes (2.1);
Sudden infant death syndrome (2.1);and Congenital anomalies (1.1).
Hispanic infant mortality
The infant mortality rate was 7.8deaths to Hispanic infants under 1year of age per 1,000 live births in anarea comprised of 45 States, New YorkState (excluding New York City), andthe District of Columbia. This was5 percent higher than the rate of 7.4for non-Hispanic white infants(table 26). Among specified subgroupsof the Hispanic population, the mor-tality rate for Mexican infants was 7.7deaths to infants under 1 year of ageper 1,000 live births, 10.2 for PuertoRican infants, and 7.6 for Cubaninfants. Infant mortality rates by spec-ified Hispanic origin and race for non-Hispanic origin may be somewhatunderstated (see Technical notes).
Maternal mortality
In 1990,343 women were reportedto have died of maternal causes(table F) compared with 320 in 1989.As in previous years, the number doesnot include all deaths occurring topregnant women, but only thosedeaths assigned to Complications ofpregnaney, childbirth, and the puerpe-rium (ICD-9 Nos. 630–676). Thematernal mortality rate for 1990 was8.2 deaths per 100,000 live births com-pared with a rate of 7.9 in 1989. Thedifference in the rate between the 2years was not statistically significant.
Black women have a higher risk ofmaternal death than white women. In1990, the maternal mortality rate forblack women was 22.4, 4.1 times the
Table F. Maternal deaths and maternalmortality rates by race: United States, 1990
[Maternal deaths are those assigned to Complicationsof pregnancy, childbirth, and the puerperium, cate-gory numbers 630-676 of the Ninth Revision /ntema-tional Classification of Diseases, 1975. Rates per100,000 live births in specified group. Beginning in1989, race for live births is tabulated according to raceof mother; see Technical notes]
for white women. Thechange in the tabulation by race of livebirths beginning with data for 1989affects maternal mortality rates by raceas it does infant mortality rates,because counts of live births comprisethe denominator of maternal mortalityrates (see Technical notes).
Report of autopsy
For 1990, all States requestedinformation on the death certificate asto whether autopsies were performed.They were reported as performed on239,591 deaths, or 11.2 percent of thedeaths that occurred in 1990, a reduc-tion from the 11.5 percent reported forthe previous year (table 28). This con-tinues the downward trend in the per-cent of deaths autopsied. The percentautopsied for all causes of death com-bined is heavily influenced by the lowrates for the three leading causes ofdeath –Diseases of heart (8 percent);Malignant neoplasms, including neo-plasms of Iymphatic and hematopoietictissues (4 percent); and Cerebrovas-cular diseases (4 percent). Among the15 leading causes of death, the highestpercents reported were for traumaticcauses —Homicide and legal interven-tion (97 percent), Suicide (55 percent),and Accidents and adverse effects(50 percent). For nontraumatic causes,the highest percents were for Certainconditions originating in the perinatalperiod (24 percent), Chronic liver dis-ease and cirrhosis (17 percent), andHIV infection (10 percent).
References
1.
2.
3.
4.
National Center for Health Statistics.Annual summary of births, marriages,divorces, and deaths: United States,1991. Monthly vital statistics report; vol40 no 13. Hyattsville, Maryland: PublicHealth Service. 1992.Centers for Disease Control. Morbidityand mortality weekly repor~ vol 41 noSS-3.Washington: Public Health Ser-vice. 1992.National Center for Health Statistics.Advance report of final mortality sta-tistics, 1989. Monthly vital statisticsrepor~ vol 40 no 8, suppl 2. Hyattsville,Maryland: Public Health Service. 1992.Tolson GC, Barnes JM, Gay GA,Kowaleski JL. The 1989 revision of the
U.S. standard certificates and reports.National Center for Health Statistics.Vital Health Stat 4(28). 1991.National Center for Health Statistics.Vital statistics of the United States,1989, vol II, mortality, part A. Wash-ington: Public Health Service. 1992.World Health Organization. Manual ofthe International Statistical Classifica-tion of Diseases, Injuries, and Causesof Death, based on the recommenda-tions of the Ninth Revision Confer-ence, 1975. Geneva: World HealthOrganization. 1977.Klebba AJ, Scott JH. Estimates ofselected comparability ratios based ondual coding of 1976 death certificatesby the Eighth and Ninth Revisions ofthe International Classification of Dis-eases. Monthly vital statistics report;vol 28 no 11, suppl. Hyattsville, Mary-land: Public Health Service. 1980.IUebba AJ, Dolman AB. Compara-bility of mortality statistics for the Sev-enth and Eighth Revisions of theInternational Classification of Dis-eases, United States. National Centerfor Health Statistics. Vital Health Stat2(66). 1975.National Center for Health Statistics.Comparability of mortality statistics forthe Sixth and Seventh Revisions,United States, 1958. Vital statis-tics – special reports; vol 51 no 4.Washington: Public Health Service.1965.National Center for Health Statistics.Vital statistics, instructions for
11<
12.
13.
14.
15.
classifying the underlying cause ofdeath. NCHS instruction manual; part2a. Hyattsville, Maryland: PublicHealth Service. Published annually.National Center for Health Statistics.Vital statistics, instructions for classi-fying multiple causes of death. NCHSinstruction manual; part 2b. Hyatts-ville, Maryland: Public Health Service.Published annually.National Center for Health Statistics.Vital statistics, nonindexed terms, stan-dard abbreviations, and State geo-graphic codes used in mortality dataclassification. NCHS instructionmanual; part 2e. Hyattsville, Maryland:Public Health Service, Published annu-ally.National Center for Health Statistics.Vital statistics, ICD-9 ACME decisiontables for classifying underlying causesof death. NCHS instruction manual;part 2c. Hyattsville, Maryland: PublicHealth Service. Published annually.National Center for Health Statistics.Vital statistics, data entry instructionsfor the mortality medical indexing,classification, and retrieval system(MICAR). NCHS instruction manual;part 2g. Hyattsville, Maryland: PublicHealth Service. Published annually.National Center for Health Statistics.Vital statistics, dictionary of validterms for the mortality medicalindexing, classification, and retrievalsystem (MICAR). NCHS instruction
16.
17.
18.
19.
20.
21.
22.
23.
manual; part 2h. Hyattsville, Maryland:Public Health Service. Publishedannually.National Center for Health Statistics.Advance report of final mortality sta-tistics, 1984. Monthly vital statisticsreport; vol 35 no 6, suppl 2. Hyattsville,Maryland: Public Health Service. 1986.National Center for Health Statistics.Advance report of final mortality sta-tistics, 1985. Monthly vital statisticsreport; vol 36 no 5, SUPPI.Hyattsville,Maryland: Public Health Service. 1987.National Center for Health Statistics.Advance report of final mortality sta-tistics, 1986. Monthly vital statistics ~report; VOI37 no 6, SUPPI.Hyattsville,Maryland: Public Health Service. 1988.U.S. Bureau of the Census, Unpub-lished data from the 1990 census, byage, race and sex.U.S. Bureau of the Census. Unpub-lished data from the 1990 census, byState.National Center for Health. Statistics.Vital statistics of the United States,1989, vol I, natality. Washington:Public Health Service. 1992.Sirken MG. Comparison of two
methods of constructing abridged lifetables by reference to a “standard”table. National Center for Health Sta-tistics. Vital Health Stat 2(4). 1966.U.S. Bureau of the Census. Unpub-lished data from the 1990 census forpersons of Spanish origin by State.
Deaths and death rates from16 selected subcategories ofMalignant neoplasms, in-cluding neoplasms of lym-phaticand hematopoietic tis-sues, byrace and sex UnitedStates, 1990 . . . . . . . . . . . . . . .
Deaths and death rates bylo-year age groups and age-adjusted death rates for Humanimmunodefieiency virus infec-tion, by race and sex UnitedStates,1989-$Xl . . . . . . . . . . . . .
Deaths and death rates formajor causes of death for theUnited States, each division,and State; and by race andsex for the United States,1990 . . . . . . . . . . . . . . . . . . . . .
Deaths by age according tospecified Hispanic origin,race fornon-Hispanic origin,and sex: Total of 45 States,New York State (excludingNew York City), and the Dis-trict of Columbia, 1990 . . . . .
Deaths for the 10 leadingcauses of death for Hispanicand white non-Hispanicorigin, for specified agegroups: Total of 45 States,New York State (excludingNew York City), and the Dis-trict of Columbia, 1990 . . . . .
Deaths and age-adjusteddeath rates for drug-inducedcauses, by race and sex:United States, 1979-90 . . . . .
Deaths and age-adjusted deathrates for alcohol-inducedcauses, by race and sex UnitedStates, 197>90 . . . . . . . . . . . . .
Deaths and percent distribu-tion of deaths for ages 15years and over, by marital
Deaths and percent distribu-tion of deaths for ages 15years and over, by educa-tional attainment, race, andsex: Total of 28 States andthe District of Columbia,1990 . . . . . . . . . . . . . . . . . . . . .
Deaths under 1 year andinfant mortality rates for 61selected causes by race:United States, 1990 . . . . . . . .
Deaths under 1 year andinfant mortali~ rates for the10 leading causes of infantdeath: United States, 1990..
Total deaths and death rates,and infant and neonataldeaths and mortality ratesfor the United States, eachdivision, and State; and byrace and sex for the UnitedStates, 1990 . . . . . . . . . . . . . . .
Infant, neonatal, and post-neonatal deaths and mor-tality rates, by specifiedHispanic origin and race fornon-Hispanic origin: Total of45 States, New York State(excluding New York City),and the District of Columbia,1990 . . . . . . . . . . . . . . . . . . . . .
Table 1. Deaths, death rates, and age-adjusted death rates, by race and sex: United States, 1940, 1950, 1960, 1970, and 1975-90
[Rates per 100,000 population in specified group. Rates are baaed on populations enumerated as of Aoril 1 for census years and estimated as of July 1 for all other years.Beginning 1970, excludes deaths of nonresidents of the United States]
All other
A// races White Total Black
BothYear
Bothsexes Ma/e Female
BothMale
Bothsexes Female sexes Male Fema/e sexes Male Female
‘Rates are revised and, therefore, may differ from those published in Advance Repo,d of Fir?tIf Mortality Statistics for 1989 and earlier years; see Technical notes,
1figura~ for agenotstated are includedin All ages but not distributed amon9 a9e 9rOUPa.
2Death rates under 1 year (based on population estimatas) differ from infant mortality rates (baaed on five bfrlhs); see tables E and 22-26 for infant mortality ratea and Technical notes for further
1l-lie table values are ravised and, therefore, may differ from those published in Advance Report of Final Mortality Statistics for 1989 and earllar yaarq aea Tachnbal notes.2Deaths basad on a 5C-parcent sample.
[Rates per 100,000 populatim in specified group. I% explanation of asterisk Precedin9 cause-of-death codes, see Technical notes]
Age
Cause of death (Ninth Revision International All Under l-t 5-14 15-24Age-
25-34 35-44 45-54 55-64Classification of Oiseases, 1975)
65–74 75-84 85 years adjustedYear ages i 1 yearz years yaars years years yaars years years years years and ovar rate3
Table 5. Age-wecific and age-adjusted death rates for’ the 15 leading causes in 1990 and selected components: United States, 1979, 1989, and 1990
.. . . . .. . . .. . . . .. .---., ––.––.–,–– . . ...- . .. -—- ----- .-_--Table 5. Age-specific and age-actjustedrieatrtrates lor me 15 Ieaalng causes In 1YYU ana seiecIea components: Urmea states, 1979, 1YtrY, ana 1YYU– Len.
[Rates per 100,000 population in specified group. For explanation of asterisk preceding cause-of-death codes, see Technical notes]
Age
Cause of death (Ninth Revision International All Under 1+ 5–14 15-24 25-34 3544 45-54 55-64 65-74 7s4Classification of Oiseases, 1975) Year ages 1 1 yearz years years years years years years years years years
Malignant neoplasm of breast. . . . . ..174-175
Malignant neoplasms of genital organs . ..179-187
Malignant neoplasms of urinary organs ..1 88–189
Malignant neoplasms of all other andunspecified sites .170–173,190–t99
1Ffgures for age not statedare includedin Afl ages but not distributed among age groups.2Dealhretesunderl year (based onpopulation estimates) dtierkm infant monal~rates (based ontive Mtihs): see bbles Eand22-26f0r infant mofialfiY rates and Technical notes fordiscussbn of the fiKerence.3F0r method of computation, see Technicaf notes.4f3ecause deaths from this cause occur primarity among infants, age-adjusted ratea are not shovm.
10.29.4
10.24.34.54.34.14.22.32.73.05.7
(?(’)(’)
NOTE: Rates for 1989 have been revieed; see Technical notes.
lF@ures for age not stated included in All agas but not distributed am0n9 a9e 9r0uPs.2Deathrates underl year (based on population estimates) tiffer from infant mmtafitymtas (basedon We b[rlhs); see tables Eand 22-26 for infant morfafii ratea, and Technical notes for further discussion of the difference.
NUr
Table 10. Deaths from 72 selected causes by race and sex: United States, 1LWO[For explanation of asterisk preceding cause-of-death codes, see Technical notes]
A// ofher
All races White Total Black
Cause of death (Ninth Revision InternationalClassification of Diseases, 1975)
Table 12. Age-adjusted death rates for72selected causes byraceand sex: United States, 1990[Rates per 100,000 population in specified group. Computed by the direct method, using as the standard population the age distribution of the total population of the United States as enumerated in IWO. For explanation
ofasterisk preceding cause-of-death codes, see Technical notes]
All other
All races White Total Black
Cause of death (Ninth Revision International Both BothClassification of Diseases, 1975) sexes Male Female sexes
Table 14. Deaths and death rates by 10-year age groups and age-adjusted death rates for Human immunodeficiency virus infection, byrace and sex United States, 1989-90
[Human immunodeficiency virus infection deaths are those assigned to categov numbers ‘042-’044 introduced in the United States in 198Z see Technical notes]
Age,4ge-
All Under 1=$ 5-14 15-24 25-34 35-44 45-54 55-64 65–74 75-64 85 years adjusted
Year, race, and sex ages 1 1 year2 years years years years years years years years years and over rate3
i Figures for age not stated included in All ages but not distributed among age groups.‘Death rates underl year (based onpopulation estimates) tifferfrom inkntmotihty rates (based on Kve Mflhs); seetables Eand22-26for infant motiti~rates and TechniMl notes fortiflherdiscussion of the difference.%rmethodof computetion,se eTechnical nOteS.
NOTE Fiates for 1989 have been revised; see Technical notes.
Table 15. Deaths and death rates for major causes of death for the United States, each division, and State; and by race and sex for theUnited States, 1990
[Rates per 100,000 population in each race-sex group and area. Numbers after cause of death are category numbers of the Ninth Revision /rrternationa/ Classification ofDiseases, Ig751
Diseases of heart
(390-398, 402,404-429)
Ma/ignantneop/asms,
inc/udingneOp/asms
of /ympfra tic andhernatopoletic
tissues(140-208)
Race, sex, and area Number Rate Number Rate
Cerebrovascular Accidents and Motor vehiclediseases adverse effects accidents Suicide
Table 15. Deaths and death rates for major causes of death for the United States, each division, and State; and by race and sex for theUnited States, 1990 -Con.
[Rates per 100,000 population in each race-sex group and area. Numbers affer cause of death are category numbers of the Ninth Revision /ntema’Ilona/ C/assii7cstion ofDiseases, 19751
Malignantneoplasms,
includingneop/asms oflymphatic and
Diseases of Frean’ hematopoietk Cerebrovascular Accidents and Motor vehicle
NOTE Caution should be used in comparing crude death rates by State.Death rates are affected by the population composition of the area.
Table 16. Deaths byageaccording tospecified Hispanic origin, race fornon-Hispanic origin, andsex: Total of45Statesj New York State(excluding New York City), and the District of Columbia, 1990
[For a listing of States, see Technical notes]
Hispanic origin, race for Under 1-4 5–14 15-24 25-34 3544 45=54 55-64 65-74 75-84 85 years Notnon-Hispanic on”gin, and sex Total 1 year years years years years years years years years years andover stated
Table 17. Deaths for the 10 leading causes of death for Hlspanlc and white non-Hispanic origin, for specified age groups: Total of 45States, New York State (excluding New York City), and the District of Columbia, 199o
[For explanation of asterisk preceding cause-of-death codes, see Technical notes. For a Iistino of States. see Technical nnte~l
Hispanic
Cause of death and ageRank (Ninth Revision International Classificationorder of Diseases, 1975) Number
infection ..,,..,....,,. ,., .,. .* O42–*O44Congenital anomalies. .740-759Cerebrovascular diseases. . .430-438Pneumonia and influenza ..480-487Complications of pregnancy, childbirth, and the
Table 17. Deatha for the 10 leading causes ‘of death for f+spanic and white non-1 -lispanlc origin, for specified age groups: Total of 45States, New York State (excluding New York City), and the District of Columbia, 1990–Con.
[For explanation of asterisk preceding cause-of-death codes, see Technical notes. For a listing of States, see Technical notes]
1For method of computation, ses Technical notes.2Ratesare revised and, therefore, may differ from those published in Advance Report o/ Final Mortality Statistics for 1989 and earlier years; see Technical notes.
Table 19. Deaths and age-adjusted death rates for alcohol-induced causes, by race and sex: United States, 1979-90
[Rates per 100,000 population in specified group, For listing of alcohol-induced causes, see Technical notes]
All other
A// races White Total Black
BothYear
BothMale
Bothsexes
BothFemale sexes Male Female sexes Male Female sexes Male Female
1For method of computation, see Technical notes,2Ratesare revised and, therel~re, may differ from those pubhshed m Advance Report of Final Mortality .%alistics for 1989 and earfier year% See Technical nOtes
1Excludes figures for age not stated.2Denominators.cfpercent distribution exclude deaths of persons of unknown marital status.
Table 21. Deaths and percent distribution of deaths for ages 15 years and over, by educational attainment, race, and sex Total of 28States and the District of Columbia, 1990
AU other
A// races White Total Black
\ Both Both BothYeare of schoo/ comp/eteo’ Male
Bothsexes Female sexes Male Female sexes Male Female sexes Male Female
Table 22. Infant, neonatal, and postneonatal mortality rates, by race and sex: United States, 1940, 1950, 1960, 1970, and 1975-90[Rates are infant (under 1 year), neonatal (under 28 days), and postneonatal (28 days-11 months) deaths per 1,000 live births in specified group. Beginning in 1989, racefor live births is tabulated according to race of mothe~ see Technical notes]
W other
A// races White Total Black
Both BothYear
BothMale
Bothsexes Female sexes Male Female sexes Male Female sexes Male Female
‘Infant deaths based on race of decedent, and Ilva blrih$based on race of molhec see Technical notes,‘Infant daatha based on rata C4decadant, and live births based on race of child; see Technical notes.
Table 23, Deaths under 1 year and infant mortality rates for 61 seleoted causes by race: United States, 1990
[Rates per 100,000 live births in specified group. For explanation of asterisk preceding cause-of-death codes, see Technical notes. Beginning in 1989, race for live births1stabulated according to race of mothec see Technical notes]
All AflCause of death (Ninth Revision International Classification of Diseases, 1975) races 1 White Black rsces~ White Black
Table 23. Deaths under 1 year and infant mortality rates for 61 selected causes by race: United States, 1990– Con.
[Rates per 100,000 live births in specified group. For explanation of asterisk preceding cause-of-death codes, see Technical notes, Beginning In 1989, race for live birthsis tabulated according to race of motheq see Technical notes]
AllCause of death (Ninth Revision international Classification of Diseases, 1975)
Allraces 1 White Black racesi White Black
Number Rate
Accidents and adverse effects . . . . .E800-E949Inhalation and ingestion of food or other object
causing obstruction of respiratory trad or suffocation. .E91 1-E912Accidental mechanical suffocation . . . . . .E913O:~cccidental causes and adverse
Table 25. Total deaths and death rates, and infant and neonatal deaths and mortality rates for the United States, each division, and State;and by race and sex for the United States, 1990
I Per 100,000 population In each race-sex group and area,Zper 1,000 KW ~flhs in e~ch race.sex group a“d area, Beginning in 1989, race for hve births is tabulated according !0 race of mothec Sae Technical nOteS.
I NOTE Caution should be used in comparing crude death rates by State.Oeath rates are affectad by the population composition of the area, I
Table 26. infant, neonatal, and postneonatal deaths and mortality rates, by specified Hispanic origin and race for non-Hispanic origin:Total of 45 States, New York State (excluding New York City), and the District of Columbia, 1990
[Rates per 1,000 live births in specified group. Live births based on race and Hispanic origin of mothec see Technical notes. For a listing of States, see Technical notes]
Hispanic Non-Hispar7/c
All Puerto Other NotAge origins Total Mexican Rican Cuban Hispanic 1 Totalz White Black stated3
Number
Under l year.............,.. 35,030 4,228 2,952 363 81 832 30,231 18,745 10,465 571
1Includes Central and South Ametican and other and unknown Hispanic,‘Includes races other than white and black.31ncludes infant daaths that occurred in States that d!d not report Hispanic origin on the death certificate.4Figures for origin not stated included in All origins but not distributed among origin groups.
Tabie 27. Maternal deaths and maternal mortality rates for selected causes by race: United States, 1990
[Maternal deaths are those assigned to Complications of pregnancy, childbirth, and the puerperium, category numbers 630-676 of the N/nfh Revision hrternationa/Classification of Diseases, 1975. Rates per 100,000 live births in specified group. Beginning In 1969, race for live births is tabulated according to race of mothec seeTechnical notes]
All other A// otherCause of death (Ninth Revision International All
Data shown in this report arebased on information from all deathcertificates filed in the 50 States andthe District of Columbia. The U.S.Standard Certificate of Death wasrevised in 1989; for additional detailssee the 1989 revision of the U.S. stan-dard certificates and reports, andTechnical Appendix fiotn Wal Statisticsof the United States, vol II (4,5).
Mortality statistics are based oninformation coded by the States andprovided to the National Center forHealth Statistics (NCHS) through theVital Statistics Cooperative Program(VSCP) and from copies of the orig-inal certificates received from theState registration offices. In 1990 allthe States and the District ofColumbia participated in this programand submitted part or all of the mor-tality data for 1990 on computer tapeto NCHS. The 25 States in the VSCPthat submitted preceded medical datafor all deaths on computer tapes areCalifornia, Colorado, Florida, Georgia,Indian% IovJ%Kansas, Louisiana, Maine,Maryland, Massachusetts, Michigan, Min-nesota, Mississippi, Nebraska, NewHampshire, New York State (excludingNew York City), North Carolina, Pennsyl-vania, South Carolina, Texas, Vermont,Virginia, Washington, and Wisconsin.NCHS also contracted with Colorado andKansas to precede medical data for alldeaths on computer tape for Alaska, Del-aware, Idaho, North Dakota, and Wyo-ming. Vermont subcontracted withPennsylvania to code its medical data.The remaining 20 VSCP States, NewYork Chy, and the District of Columbiasubmitted copies of the original certifi-cates from which NCHS coded the med-ical data. For 1990 all States submittedpreceded demographic data for all deaths.
Data for the entire United Statesrefer to events occurring within theUnited States. Data shown for geo-graphic areas are by place of resi-dence. Beginning with 1970, mortalitystatistics exclude deaths of nonresi-dents of the United States. All dataexclude fetal deaths.
Cause-of-death classification
The mortality statistics presentedhere were compiled in accordance withthe World Health Organization(WHO) regulations, which specify thatmember nations classify causes ofdeath by the current Manual of theInternational Statistical Classification ofDiseases, Injuries, and Causes of Death(6). Causes of death for 1979-90 wereclassified according to the manual. Forearlier years, causes of death wereclassified according to the revisionsthen in use – 1968-78, Eighth Revi-sion; 195S’-67, Seventh Revision; and1949–57, Sixth Revision. Changes inclassification of causes of death due tothese revisions may result in disconti-nuities in cause-of-death trends. Con-sequently, cause-of-death comparisonsamong revisions require cctnsiderationof comparability ratios and, whereavailable, estimates of their standarderrors. Comparability ratios betweenthe Eighth and Ninth Revisions,between the Seventh and Eighth Revi-sions, and between the Sixth and Sev-enth Revisions may be found in otherNCHS reports (7-9).
Besides specifying the classifica-tion, WHO regulations outline theform of medical certification and theprocedures to be used in coding causeof death. Cause-of-death data pre-sented in this publication were codedby procedures outlined in annualissues of the NCHS InstmctionalManual (10-15).
Before 1968, mortality medicaldata were based on manual coding ofan underlying cause of death for eachcertificate in accordance with WHOrules. Effective with data year 1968,NCHS converted to computerizedcoding of the underlying cause andmanual coding of al[ causes (multiplecauses) on the death certificate. In thissystem, called “Automated Classifica-tion of Medical Entities” (ACME)(13), the multiple cause codes serve asinputs to the computer sofhvare thatemploys WHO rules to select theunderlying cause. Since 1968, manyStates also have implemented ACME
and provide multiple cause and under-lying cause data to NCHS in electronicform.
Beginning with data year 1990,another computer system is beingimplemented. This system, called“Mortality Medical Indexing, Classifi-cation, and Retrieval” (MICAR)(14,15), automates the coding of themultiple causes of death. In addition,MICAR ultimately will provide muchmore detailed information on the dis-eases reported on death certificatesthan is available through the Interna-tional Classification of Diseases (ICD)code structure. In this first year ofimplementation, only about 5 percent(94,372) of the Nation’s death recordswere multiple cause coded usingMICAR with subsequent processingthrough ACME. This includes a por-tion of the records for Alabama, Ken-tucky, Oregon, Rhode Island, andWest Virginia. The remainder of thenational file was processed by eitherNCHS or the State using only theACME system.
In this report, tabulations ofcause-of-death statistics are basedsolely on the underlying cause ofdeath. The underlying cause is definedby WHO as the disease or injury thatinitiated the sequence of eventsleading directly to death or as thecircumstances of the accident or vio-lence that produced the fatal injury, Itis selected from the conditions enteredby the physician in the cause-of-deathsection of the death certificate. Whenmore than one cause or condition isentered by the physician, the under-lying cause is determined by thesequence of conditions on the certif-icate, provisions of the ICD, and asso-ciated selection rules. Generally, moremedica[ information is reported ondeath certificates than is directlyreflected in the underlying cause ofdeath.
Codes for HIV infection
Beginning with data for 1987,NCHS introduced category numbers
infection. The asterisk before the cat-egory numbers indicates that thesecodes are not part of the Ninth Revi-sion of the International Classificationof Diseases (ICD-9). Deaths classifiedto HIV infection for 1989–90 areshown in table 14. They are included,but not shown separately, in the cate-gory All other infectious and parasiticdiseases in the List of 72 SelectedCauses of Death and in the categoryRemainder of infectious and parasiticdiseases in the List of 61 SelectedCauses of Infant Deaths. Before 1987,deaths involving HIV infection wereclassified to Deficiency of cell-mediated immunity (ICD 279.1), con-tained in the category All otherdiseases; to Pneumocystosis (ICD-9No. 136.3), contained in the categoryAII other infectious and parasitic dis-eases; to Malignant neoplasms,including neoplasms of lymphatic andhematopoietic tissues; and to anumber of other causes. As a conse-quence, cause-of-death data beginningwith 1987 are not strictly comparablewith data for previous years.
For data years 1983-86, acquiredimmunodeficiency syndrome (AIDS)and HIV infection, when reported onthe death certificate, were assigned tothe category Deficiency of cell-mediated immunity (ICD-9 No. 279.1).Because the selection rules for under-lying cause of death were developedbefore the identification of AIDS,other conditions mentioned on thedeath certificate and not category No.279.1 were often selected as the under-lying cause of death during this period,Alsoj this category was not uniquelyspecific for HIV conditions. As men-tioned in more detail in previousreports for 1984-86 (16-18) thenumber of death certificates that hadmention of conditions coded to ICD-9No. 279.1 was 2,943 for 1984, 6,040 for1985, and 10,900 for 1986. It isbelieved that HIV infection wasinvolved in most of these deaths.
Drug-induced deaths
Causes of death attributable todrug-induced mortality include ICD-9No. 292, Drug psychoses; No. 304,
Drug dependence; Nos. 305.2-305.9,Nondependent use of drugs, notincluding alcohol and tobacco; Nos.E850-E858, Accidental poisoning bydrugs, medicaments, and biological;Nos. E950.0-E950.5, Suicide by drugs,medicaments, and biological; No.E962.0, Assault from poisoning bydrugs and medicaments; and NOS.
E980.O-E980.5, Poisoning by drugs,medicaments, and biological, unde-termined whether accidentally or pur-posely inflicted. Drug-induced causesexclude accidents, homicides, andother causes indirectly related to druguse.
Alcohol-induced deaths
Causes of death attributable toalcohol-induced mortality includeICD-9 No. 291, Alcoholic psychoses;No. 303, Alcohol dependence syn-drome; No. 305.0, Nondependentabuse of alcohol; No. 357.5, Alcoholicpolyneuropathy; No. 425.5, Alcoholiccardiomyopathy; No. 535.3, Alcoholicgastritis; Nos. 571 ,0–571.3, Chronicliver disease and cirrhosis, specified asalcoholic; No, 790.3, Excessive bloodlevel of alcohol; No. E860, Accidentalpoisoning by alcohol, not elsewhereclassified. Alcohol-induced causesexclude accidents, homicides, andother causes indirectly related toalcohol use.
Educational attainment
Beginning with the 1989 data year,mortality data on educational attain-ment are being tabulated from infor-mation reported on the deathcertificate. As a result of the revisionof the U.S. Standard Certificate ofDeath (4), this item was added to thecertificates of numerous States. Mor-tality data on educational attainmentfor 1990 are based on deaths to resi-dents of the 28 States and the Districtof Columbia whose data were at least90 percent complete on a place-of-occurrence basis. These 28 States areAlabama, Arizona, California, Colo-rado, Delaware, Florida, Hawaii,Idaho, Illinois, Iowa, Kansas, Massa-chusetts, Michigan,souri, Montana,Hampshire, North
One index of the quality ofreporting causes of death is the pro-portion of death certificates coded tothe Ninth Revision Chapter XVISymptoms, signs, and ill-defined condi-tions (ICD-9 Nos. 780-799). Althoughdeaths occur for which the underlyingcauses are impossible to determine,this proportion indicates the care andconsideration given to the certificationby the medical certifier. This propor-tion also may be used as a rough mea-sure of the specificity of the medicaldiagnoses made by the certifier in var-ious areas. In 1990, a record low of1.12 percent of all reported deaths inthe United States were assigned toSymptoms, signs, and ill-defined condi-tions (ICD-9 Nos. 780–799) comparedwith 1.25 percent for 1989. However,trends in the percent of deathsassigned to this category vary by age,Although the percent of deaths fromthis cause for all ages combined gener-ally has remained stable since 1980,decreases have occurred for the agegroup 55-64 years since 1983; for agegroup 65–74 years since 1982; for agegroup 75-84 years since 1986; and for10-year age groups from 15 to 54 yearssince 1988. Between 1989 and 1990,the percent decreased for all agegroups, except for those under 1 yearof age, which was unchanged.
Population bases forcomputing rates
The population used for com-puting death rates and life tablesshown in this report (furnished by theU.S. Bureau of the Census) representsthe population residing in the specifiedarea, The populations for 1990 arebased on the April 1, 1990 censusenumeration. Death rates and lifetables for 1981–89 have been recom-puted, based on revised populationsthat are consistent with the 1990census levels (19,20). They are, there-fore, not comparable with death ratesand life table values published in the
Advance Repoti of Final Mortali~ Sta-tistics, in Vital Statktics of the UnitedStates, Volume I& Mortality and inother NCHS publications for thoseyears.
Infant mortality rates shown in fig-ure5 and tabIes E and 22–26 are themost commonly used index for mea-suring the risk of dying during the firstyear of life. They are calculated bydividing the number of infant deaths ina calendar year by the number of livebirths registered for the same periodand are presented as rates per 1,000 orper 100,000 live births. Infant mortalityrates use the number of live births inthe denominator to approximate thepopulation at risk of dying before thefirst birthday.
In contrast to infant mortalityrates based on live births, infant deathrates shown in figure 2 and tables A, 2,5,9, and 14 are based on the estimatedpopulation under 1 year of age. Forintercensal years infant death ratesthat appear in tabulations of age-specific death rates are calculated bydividing the number of infant deaths ina calendar year by the midyear popu-lation of infants under 1 year of age(estimated from births occurring in the12-month period ending with June)and are presented as rates per 100,000population in this age group. Becauseof differences in the denominators,infant death rates may differ frominfant mortality rates.
Race for Infant and MaternalMortality Rates
Beginning with the 1989 data year,the method of tabulating live births byrace was changed from race of child torace of mother. This change affectsinfant and maternal mortality ratesbecause live births comprise thedenominator of these rates. In 1989and 1990, as in previous years, infantand maternal deaths continue to betabulated by the race of the decedent.
As noted in detail in TechnicalAppendix to Vital Statistics of theUnited States, Volume 1, Natality (21),beginning with 1989, live births aretabulated primarily by race of mother,as reported directly on the birth certif-icate. Before 1989 birth tabulations
were by race of child, as determinedstatistically by an algorithm based onrace as reported for the mother andfather. Briefly, in cases of mixed par-entage where only one parent waswhite, the child was assigned to theother parent’s race. When neitherparent was white, the child wasassigned the race of the father, exceptif either parent was Hawaiian, thechild was assigned to Hawaiian. If racewas not reported for one parent, thechild was assigned the race of theparent for whom race was given.
The change in tabulating livebirths by race reflects three factorsover the past two decades – the topicalcontent of the birth certificate hasbeen expanded to include considerablehealth and demographic informationrelated to the mother, the increasingincidence of interracial parentage, andthe growing proportion of births forwhich no information on the father isreported.
Quantitatively, the change in thebasis for tabulating live births by raceresults in more white births and fewerblack births and births of other races.Consequently, infant and maternalmortality rates under the new tabu-lating procedure tend to be about2 percent lower for white infants andabout 5 percent higher for blackinfants than they are when computedby the previous method of tabulatinglive births by race of child. Rates formost other minority races also arehigher when computed by race ofmother (5).
The change in tabulating race oflive births presents challenges to thoseanalyzing infant and maternal mor-tality data, particularly trend data. Tofacilitate continui~ and ease of inter-pretation, key published tables for1989 and 1990, incIuding alI trendtables, wilI show data computed on thebasis of live births tabulated by race ofmother and race of child. This willmake it possible to distinguish theeffects of this change from reaIchanges in the data. The text in thisreport focuses on live birth data tabu-lated by race of mother except wheretrends are discussed. In the latter case,the analysis is based on data tabulatedby race of child.
Cause-of-death rankings
The cause-of-death rankingsshown in figure 4 and tables B–D, 5–7,17, and 28 are based on the List of 72SeIected Causes of Death and HIVinfection (ICD-9 Nos. *042–*044). Thecause-of-death ranking for infants intable 24 is based on the List of 61Selected Causes of Infant Death andHIV infection. HIV infection wasadded to the lists of rankabIe causeseffective with data year 1987.
The group titles Major cardiovas-cular diseases and Symptoms, signs,and ill-defined conditions are notranked from the List of 72 SelectedCauses; Certain conditions originatingin the perinatal period and Symptoms,signs, and ill-defined conditions arenot ranked from the List of 61Selected Causes of Infant Death. Inaddition, category titles that begin withthe words “Other” and “All Other”are not ranked to determine theleading causes of death. When one ofthe titIes that represents a subtotal isranked (for example, Tuberculosis), itscomponent parts are not ranked (inthis case, Tuberculosis of respiratorysystem and Other tuberculosis).
Age-adjusted rates
The age-adjusted rates presentedin this report were computed by thedirect method, that is, by applying theage-specific death rates for a givencause of death to the standard popula-tion distributed by age. The total pop-ulation as enumerated in 1940 wasselected as the standard. By using thesame standard population, the ratesfor the total population and for eachrace-sex group were adjusted sepa-rately. The age-adjusted rates werebased on 10-year age groups. It isimportant not to compare age-adjusteddeath rates with crude rates.
Random variation
Although the mortality data in thisreport (except data for 1972) are notsubject to sampiing error, they maybeaffected by random variation in thenumber of deaths involved. When thenumber of events is small (perhaps lessthan 100) and the probability of such
an event is small, considerable cautionmust be observed in interpreting thedata. Such infrequent events may beassumed to follow a Poisson proba-bility distribution. For this distribu-tion, a simple approximation may beused to estimate the confidenceinterval, as follows:
If N is the number of registereddeaths in the population and R is thecorresponding rate, the chances are 19in 20 (approximate 95-percent confi-dence interval) that
1. N-2~and N +2fl
covem the “true” number ofevents.
2, R-2& and R+2&
m m
covers the “true” rate,
If the rate R, corresponding to N,events is compared with the rate R2corresponding to Nz events, the dif-ference between the two rates may beregarded as statistically significant if itexceeds
3“2%’T%Additional information on random
variation may be found in the Tech-nical Appendix of V7tal Statistics of theUnited States, Volume II (5).
Infant and maternal mortalityrates —Comparisons made in the textamong infant, neonatal, postneonatal,and maternal mortality rates, unlessotherwise specified, are statisticallysignificant at the 0.05 level of signifi-cance. Lack of comment in the textabout any two rates does not meanthat the difference was tested andfound not to be significant at this level.
Rates, proportions, and ratios
Beginning with 1989 data, anasterisk is shown in place of a rate
based on 20 or fewer deaths. Theserates have a relative standard error of23 percent or more and are, therefore,considered statistically unreliable. Forage-adjusted death rates, this criterionis applied to the sum of the age-specific deaths.
Life tables
U.S. abridged life tables are con-structed by reference to a standard lifetable (22).
Hispanic origin
For 1990 mortality data for theHispanic-origin population are basedon deaths to residents of 45 States,New York State (excluding New YorkCity), and the District of Columbiawhose data were at least 90 percentcomplete on a place-of-occurrencebasis and considered to be sufficientlycomparable to be used for analysis.The 45 States are Alabama, Alaska,Arizona, Arkansas, California, Colo-rado, Delaware, Florida, Georgia,Hawaii, Idaho, Illinois, Indiana, Iowa,Kansas, Kentucky, Maine, Maryland,Massachusetts, Michigan, Minnesota,Mississippi, Missouri, Montana,Nebraska, Nevada, New Jersey, NewMexico, North Carolina, NorthDakota, Ohio, Oregon, Pennsylvania,Rhode Island, South Carolina, SouthDakota, Tennessee, Texas, Utah, Ver-mont, Virginia, Washington, West Vir-ginia, Wisconsin, and Wyoming. Datafor New York City are excluded for1990 because more than 10 percent ofits death certificates were coded to“unknown origin. ” Because about halfof the deaths to Puerto Ricans areaccounted for by New York City, theresulting mortality data may not becomparable with that of previousyears.
Infant mortality rates for theHispanic-origin population are basedon numbers of resident infant deaths
reported to be of Hispanic-origin andnumbers of resident live births by His-panic origin of mother for 45 States,New York State (excluding New YorkCity), and the District of Columbia.The 45 States are the same as listedabove. In computing infant mortalityrates, deaths and live births ofunknown origin are not distributedamong the specified Hispanic and non-Hispanic groups. Because the percentof infant deaths of unknown origin was1,6 and the percent of live births ofunknown origin was 1,0 in thereporting area for 1990, infant mor-tality rates by specified Hispanic originand race for non-Hispanic origin maybe somewhat underestimated,
Small numbers of infant deaths forspecific Hispanic-origin groups canresult in infant mortality rates subjectto relatively large random variation(see section, “Random variation”).
In 1990 the 45 States, New YorkState (excluding New York City), andthe District of Columbia accounted forabout 89 percent of the Hispanic pop-ulation in the United States, includingabout 99 percent of the Mexican pop-ulation, 58 percent of the PuertoRican population, 92 percent of theCuban population, and 81 percent ofthe “Other Hispanic” population (23).
Computation of percentdistributions
Deaths of persons of unknownmarital status and unknown educa-tional attainment were subtractedfrom figures for total deaths used asdenominators before percent distribu-tions were computed.
This report represents summary tabulations from the final mortali~ statistics for 1990. Moredetailed tabulations for 1990 will be pubIished in J’%alStatistics of the United States, 1990Volume II – Mortality. Prior to the publication of that volume, the National Center for HealthStatistics will respond to requests for unpublished data whenever possible.
Suggested citation Copyright information National Center for Health :Statlstlcs
National Center for Health Statistics. Advance This report may be reprinted without furtherreport of final mortality statistics, 1990. permission.Monthly vital statistics report; VOI41 no 7,suppl, Hyattsville, Maryland: Public HealthService. 1993.
DirectorManning Feinleib, M.D., Dr. P.H
Acting Deputy DirectorJack R. Anderson
U.S. DEPARTMENT OF HEALTHAND HUMAN SERVICESPublic Health ServiceCenters for Diseese Control and PreventionNational Center for Health Statistics6525 Belcrest RoadHyattsville, Maryland 20782
OFFICIAL BUSINESSPENALTY FOR PRIVATE USE, $300
To receive the Monthly Vital Statistics Reportregularly, contact the National Center for HealthStatistics by calling 301-436-8500