Male Mortality Trends in the United States,1900-2010: Progress,Challenges,and Opportunities

Abstract

Objectives:

Male mortality fell substantially during the past century, and major causes of death changed. Building on our recent analysis of female mortality trends in the United States, we examined all-cause and cause-specific mortality trends at each decade from 1900 to 2010 among US males.

Methods:

We conducted a descriptive study of age-adjusted death rates (AADRs) for 11 categories of disease and injury stratified by race (white, nonwhite, and, when available, black), the excess of male mortality over female mortality ([male AADR − female AADR]/female AADR), and potential causes of persistent excess of male mortality. We used national mortality data for each decade.

Results:

From 1900 to 2010, the all-cause AADR declined 66.4% among white males and 74.5% among nonwhite males. Five major causes of death in 1900 were pneumonia and influenza, heart disease, stroke, tuberculosis, and unintentional nonmotor vehicle injuries; in 2010, infectious conditions were replaced by cancers and chronic lower respiratory diseases. The all-cause excess of male mortality rose from 9.1% in 1900 to 65.5% in 1980 among white males and a peak of 63.7% in 1990 among nonwhite males, subsequently falling among all groups.

Conclusion:

During the last century, AADRs among males declined more slowly than among females. Although the gap diminished in recent decades, exploration of social and behavioral factors may inform interventions that could further reduce death rates among males.

Keywords

male health mortality trends chronic disease sex and gender disparities

From 1900 to 2010, age-adjusted death rates (AADRs) per 100 000 for the US population declined 70%, from 2518 to 747; life expectancy at birth increased from 47.3 to 78.7 years; and major causes of death (CODs) changed.^1,2 Female mortality rates declined substantially during this period,³ likely as a result of broad progress in public health, health behaviors, social and environmental factors, and, more recently, clinical care. During this period, male mortality rates persistently exceeded female mortality rates.¹ Here, we explored trends in male mortality from 1900 to 2010, trends in male excess mortality, and possible explanations.

We extended previous analyses, recognizing changes in death reporting and disease classification. Changes include an increase since the early 20th century in the number of states that report death data, improvements in clinical knowledge that have led to better disease classification, and 9 revisions (approximately decennially) of the original list of International Causes of Death (ICD).³ Racial/ethnic data were limited to the categories of white and nonwhite until 1970, when data on black persons were first reported separately. Consideration of differences in death rates by cause, sex, race, and other factors may help elucidate biological, behavioral, social, and environmental contributions to excess male mortality, suggesting public health and medical interventions.

Methods

We tabulated male death rates: all-cause age-specific death rates, trends in AADRs for specific CODs, and AADRs for major CODs by race. We examined trends in male excess AADRs ([male AADR − female AADR]/female AADR) for major CODs by race. We obtained data for the analysis of decadal trends in death rates (1900-2010) from tabulations of CODs from published compilations and from public-use computer data files, as previously described.⁴ Because all data were from previously published US government sources, no institutional review board review was required.

Analyses

We reported death rates and AADRs for the underlying COD for males and females by race (white and nonwhite) previously,⁴ and we used those rates in these analyses. The CODs examined were the 5 major causes either in 1900 or in 2010—ICD groupings that constitute the largest proportions of deaths at those times. Two physician epidemiologists (S.M.T. and R.G.P.) linked multiple ICD codes. From 1900 to 1940, leukemia was not classified with cancers in ICDs, and rates for leukemia are available for these years only for the whole population (ie, without sex- or race-specific rates). Therefore, to show rates for cancers for the entire period of 1900-2010, leukemia was excluded from the “all cancer” category before 1950.

All analyses reported here examined male mortality, unless otherwise noted. Because age- and cause-specific rates by white race and nonwhite race were not available until 1920, we tabulated cause-specific AADRs by race only for 1920-2010. Because age- and cause-specific data for black persons were not available until 1970, we tabulated data for black males only for 1970-2010. All rates were tabulated per 100 000 US 2000 standard population.

We calculated the following:

Age-specific all-cause death rates for white males and nonwhite males from 1900-2010 in decadal years.

Major CODs in 1900 and in 2010, based on AADRs for each major cause in 1900 and 2010. Direct comparison of ranked causes between the 2 years was not possible, because data for chronic lower respiratory diseases were not available in 1900.

AADR trends for selected major CODs for the whole male population from 1900-2010, for white males and nonwhite males from 1920-2010, and for black males from 1970-2010.

Cause-specific AADRs from 1900-2010 for selected conditions combined: chronic (ie, heart disease, stroke, cancers combined), infectious (ie, influenza and pneumonia, tuberculosis, and enteritis and diarrhea combined), and unintentional injury (ie, motor vehicle injury [MVI)] and unintentional nonmotor vehicle injuries [UNMVIs], such as poisoning [including drug overdose], falls, and drowning).

Relative differences between male and female AADRs ([male AADR − female AADR]/female AADR) for all causes and for selected specific causes for the whole population from 1900-2010, for white males and nonwhite males from 1920-2010, and for black males from 1970-2010. Relative differences >0 were defined as “excess.”

We used Microsoft Excel software for all analyses.

Results

Age-Specific All-Cause Death Rates, by Race, 1900-2010

Age-specific death rates declined from 1900 to 2010 for all age groups among both white and nonwhite males, with greater declines among younger age groups among both white and nonwhite males (Table 1). Age-specific death rates among nonwhite males exceeded those among white males in all decades and all age groups except for males aged ≥75, but declines in death rates were greater among nonwhite males than among white males in all age groups. The greatest declines were among males aged 1-4 years—a decline of 98.6% among white males and 99.2% among nonwhite males.

Table 1.

Age-specific death rates (per 100 000 US 2000 standard population) among males, by race (white and nonwhite), death registration states, 1900-1930, and United States, 1940-2010^a

Age, y, and Race^b	Death Registration States				United States								% Decline 1900 to 2010
Age, y, and Race^b	1900	1910	1920	1930	1940	1950	1960	1970	1980	1990	2000	2010	% Decline 1900 to 2010
<1
White	17 587	14 297	9810	7154	5670	3401	2694	2113	1230	896	668	584	97
Nonwhite	36 926	25 762	16 774	12 226	10 122	5992	5189	4020	2350	1805	1305	974	97
1-4
White	2016	1424	984	554	281	136	105	84	66	46	33	27	99
Nonwhite	4343	3013	1498	996	527	271	207	145	103	79	48	37	99
5-14
White	378	296	271	181	112	67	53	48	35	26	20	14	96
Nonwhite	777	543	365	266	164	97	75	65	45	37	25	17	98
15-24
White	580	469	425	298	197	152	144	171	167	131	106	92	84
Nonwhite	1183	1001	994	781	499	290	214	305	202	217	150	117	90
25-34
White	813	672	594	407	281	185	163	177	171	176	124	136	83
Nonwhite	1249	1264	1215	1214	853	496	386	504	358	350	199	163	87
35-44
White	1059	969	765	651	507	381	333	344	257	268	234	207	81
Nonwhite	1419	1752	1445	1700	1318	861	729	874	594	559	359	235	83
45-54
White	1550	1495	1198	1231	1139	985	932	883	699	549	497	492	68
Nonwhite	2466	2483	2009	2633	2453	1857	1551	1646	1309	1018	797	568	77
55-64
White	2852	2841	2421	2552	2522	2304	2225	2203	1729	1467	1163	1033	64
Nonwhite	4213	3874	3114	4039	3711	3481	3152	3047	2606	2219	1661	1304	69
65-74
White	5913	5858	5419	5512	5400	4865	4848	4810	4036	3398	2906	2232	62
Nonwhite	7164	6632	6016	6727	6283	5795	5664	5474	4747	4349	3545	2547	65
75-84
White	12 823	12 764	12 250	11 921	12 202	10 526	10 300	10 099	8830	7845	6933	5704	56
Nonwhite	13 139	11 379	11 599	11 763	10 877	9030	8663	8981	8689	8316	7352	5615	57
≥85
White	26 918	25 794	25 358	23 762	25 135	22 116	21 750	18 552	19 097	18 268	17 716	15 640	42
Nonwhite	24 929	17 407	24 709	22 845	19 972	16 022	15 239	11 405	15 775	16 081	15 488	13 293	47

^a The states providing mortality data are known as “death registration states”; they included 10 states and the District of Columbia in 1900 (40.5% of the US population) and gradually expanded to include all 48 states and the District of Columbia by 1933. Data source: tabulations of causes of death from published compilations and from public-use computer data files, as previously described in Hahn et al.⁴

^b Disaggregation of race beyond white and nonwhite did not occur until 1970.

Declines in death rates among white and nonwhite males from birth through age 14 years occurred every decade from 1900 to 2010. For white and nonwhite males aged 15-44, death rates increased from 1960 to 1970 and were irregular thereafter. For several white and nonwhite age groups aged ≥35, death rate declines were irregular in decades after 1910, increased in 1930, and declined irregularly thereafter.

Major CODs in 1900 and 2010

Five major CODs for males comprised 44.3% of the all-cause AADRs in 1900: pneumonia and influenza (10.9%), heart disease (10.9%), stroke (9.4%), tuberculosis (8.3%), and UNMVIs (4.8%, minimally greater than diarrhea and enteritis; Table 2). In 2010, 5 major CODs comprised 62.0% of all-cause AADRs: heart disease (25.4%), all cancers (23.7%), chronic lower respiratory diseases (5.4%), stroke (4.4%), and UNMVIs (4.0%).

Table 2.

Age-adjusted death rates (per 100 000 US 2000 standard population) among males for all causes and selected causes, by race (white, nonwhite, and black) and available years, death registration states, 1900-1930, and United States, 1940-2010^a

Cause/Race^b	Death Registration States				United States
Cause/Race^b	1900	1910	1920	1930	1940	1950	1960	1970	1980	1990	2000	2010
All causes
All	2631	2460	2213	2089	1976	1674	1609	1542	1348	1203	1054	887
White	2613	2442	2167	2014	1925	1643	1586	1514	1318	1166	1029	879
Nonwhite	3577	3088	2748	2846	2459	1950	1778	1755	1569	1443	1187	911
Black	—	—	—	—	—	—	—	1874	1698	1645	1404	1104
Selected infectious diseases^c
All	631	520	458	285	200	96	83	60	44	49	29	19
White	—	—	436	261	182	86	77	57	42	48	29	19
Nonwhite	—	—	718	514	374	191	135	89	54	57	33	22
Black	—	—	—	—	—	—	—	93	56	61	35	23
Tuberculosis
All	218	181	130	86	61	35	12	5	2	1	0	0
White	—	—	117	73	51	29	10	4	1	1	0	0
Nonwhite	—	—	275	210	154	91	29	14	5	5	2	1
Black	—	—	—	—	—	—	—	15	6	5	2	1
Influenza and pneumonia
All	287	239	281	172	128	55	66	54	42	48	29	18
White	—	—	272	162	120	51	62	52	41	47	28	18
Nonwhite	—	—	382	263	201	90	99	72	48	52	31	21
Black	—	—	—	—	—	—	—	77	50	56	34	21
Diarrhea and enteritis
All	126	100	48	27	12	5	5	1	0	0	0	1
White	—	—	46	25	11	5	5	1	0	0	0	1
Nonwhite	—	—	61	41	20	10	7	2	0	0	0	1
Black	—	—	—	—	—	—	—	2	0	0	0	1
Selected chronic diseases^d
All	624	730	753	874	999	1094	1099	1039	913	762	631	474
White	—	—	759	873	1006	1094	1101	1039	904	747	620	469
Nonwhite	—	—	648	861	894	1065	1056	1033	977	869	700	497
Black	—	—	—	—	—	—	—	1106	1057	986	828	607
All cancers except leukemia
All	91	118	140	162	187	208	225	248	271	280	249	210
White	—	—	145	168	192	210	225	245	265	272	244	208
Nonwhite	—	—	61	81	118	180	225	273	324	347	286	219
Black	—	—	—	—	—	—	—	292	353	398	346	269
Cancers of respiratory system
All	—	—	—	—	—	29	48	72	89	95	80	63
White	—	—	—	—	—	30	48	72	88	93	78	62
Nonwhite	—	—	—	—	—	21	47	76	103	113	88	63
Black	—	—	—	—	—	—	—	81	114	132	107	78
Major cardiovascular diseases
All	579	716	720	792	877	965	950	850	683	511	407	282
White	—	—	724	786	880	963	952	853	680	505	401	277
Nonwhite	—	—	663	846	832	961	912	812	692	552	444	302
Black	—	—	—	—	—	—	—	870	745	623	518	367
Heart disease
All	286	371	383	508	632	699	688	634	539	412	320	225
White	—	—	383	505	638	701	695	640	540	409	317	223
Nonwhite	—	—	381	530	552	653	601	568	521	429	334	229
Black	—	—	—	—	—	—	—	607	561	485	393	281
Stroke
All	248	241	230	204	180	186	186	157	102	69	62	39
White	—	—	231	200	176	182	182	154	99	66	60	38
Nonwhite	—	—	206	250	225	231	230	192	131	92	80	49
Black	—	—	—	—	—	—	—	206	142	103	90	57
Chronic lower respiratory diseases
All	—	—	—	—	—	—	—	39	50	56	56	49
White	—	—	—	—	—	—	—	41	52	57	57	51
Nonwhite	—	—	—	—	—	—	—	26	32	44	43	34
Black	—	—	—	—	—	—	—	27	34	48	48	40
Unintentional injuries^e
All	—	156	125	147	130	102	86	90	71	54	49	52
White	—	—	125	146	129	100	83	86	68	53	49	54
Nonwhite	—	—	124	154	137	119	111	123	89	65	50	40
Black	—	—	—	—	—	—	—	127	94	71	58	46
Unintentional motor vehicle injuries
All	—	3	17	47	45	39	35	42	34	27	22	16
White	—	—	18	48	45	38	35	40	34	26	22	17
Nonwhite	—	—	9	43	45	44	41	51	34	28	22	14
Black	—	—	—	—	—	—	—	51	34	30	24	17
Unintentional nonmotor vehicle injuries
All	127	152	108	100	85	64	51	49	37	28	28	35
White	—	—	107	98	84	62	49	46	35	26	27	37
Nonwhite	—	—	115	111	91	75	70	73	55	37	29	26
Black	—	—	—	—	—	—	—	76	59	42	34	29
Ill-defined conditions
All	—	—	—	—	—	30	20	20	18	12	13	12
White	—	—	—	—	—	23	15	16	15	11	13	12
Nonwhite	—	—	—	—	—	109	66	58	44	22	16	13
Black	—	—	—	—	—	—	—	64	49	26	20	17
Senility
All	—	—	—	—	—	—	—	1	1	1	1	2
White	—	—	—	—	—	—	—	1	1	1	1	2
Nonwhite	—	—	—	—	—	—	—	4	2	1	1	1
Black	—	—	—	—	—	—	—	4	2	1	1	2
Other ill-defined conditions
All	—	—	—	—	—	—	—	19	17	12	12	10
White	—	—	—	—	—	—	—	15	14	10	11	10
Nonwhite	—	—	—	—	—	—	—	54	42	12	15	12
Black	—	—	—	—	—	—	—	60	47	25	19	15

Abbreviation: —, data not available.

^b Disaggregation of race beyond white and nonwhite did not occur until 1970.

^c Selected infectious disease totals are the sum of the rates for influenza and pneumonia, tuberculosis, and enteritis and diarrhea.

^d Selected chronic disease totals are the sum of the rates for heart disease, stroke, and all cancers. They do not include chronic lower respiratory diseases.

^e Unintentional injury totals are the sum of the rates for unintentional motor vehicle injuries and unintentional nonmotor vehicle injuries.

Cause-Specific AADRs

From 1900 to 2010, the greatest decreases in AADRs were for selected infectious diseases: pneumonia and influenza declined by 93.7% (from 287 to 18 per 100 000 population), tuberculosis by 99.9% (from 218 to 0.2 per 100 000 population), and diarrhea and enteritis by 99.4% (from 126 to 1 per 100 000 population; Table 2; Figure 1). By 1950, these infectious conditions combined already had declined by 84.9% (from 631 to 95 per 100 000 population). AADRs for major cardiovascular diseases (dominated by heart disease) increased by 66.7% from 1900 (579 per 100 000 population) to 1950 (965 per 100 000 population) and then declined 70.8% (to 282 per 100 000 population) by 2010. AADRs for UNMVIs peaked in 1910 (152 per 100 000 population) and then declined 76.8% by 2010 (to 35 per 100 000 population).

Figure 1.

Age-adjusted death rates by cause of death among US males, 1900-2010. Rates are per 100 000 US 2000 standard population. Data for 1900-1930 are from death registration states; data for 1940-2010 are from the United States. The states providing mortality data are known as “death registration states”; they included 10 states and the District of Columbia in 1900 (40.5% of the US population) and gradually expanded to include all 48 states and the District of Columbia by 1933. Data source: tabulations of causes of death from published compilations and from public-use computer data files, as previously described.⁴ Abbreviations: MVI, motor vehicle injuries; UNMVI, unintentional nonmotor vehicle injuries.

The AADR trends among males for 3 additional major CODs varied during the study period (Table 1). For cancers, the AADR almost tripled from 1900 to 1990 and then declined by 25.1% thereafter. Chronic lower respiratory conditions, first reported in 1970 at 39 per 100 000 population, increased by 41.6% in 2000 and then declined by 12.7% in 2010. The AADR for MVIs, first recorded in 1910 at 3 per 100 000 population, increased rapidly to a peak of 47 per 100 000 population in 1930 and then declined by 65.7% to 16 per 100 000 population in 2010, after 4 decades of fluctuation.

All-Cause and Cause-Specific AADRs Among White and Nonwhite Males (1920-2010) and Black Males (1970-2010)

From 1900 to 2010, the decline in all-cause AADR was greater among nonwhite males (74.5%) than among white males (66.4%), but the AADR among white males was lower than the AADR among nonwhite males throughout the study period (Table 1). The AADRs among black males exceeded the rate among white and nonwhite males from 1970 to 2010, but they declined 41% overall during this period.

The AADRs for heart disease among white and nonwhite males rose 83.1% and 71.5%, respectively, from 1920 to a peak in 1950 and subsequently fell by 68.2% and 64.9% in 2010 (Table 2). The AADRs for stroke started at their peak of 231 per 100 000 population in 1920 among white males; they peaked at 250 per 100 000 population in 1930 among nonwhite males. Stroke rates then declined by 83.7% and 80.5% among white and nonwhite males, respectively, by 2010. In 2010, the AADR for stroke among nonwhite males (49 per 100 000 population) was 30.1% higher than among white males (38 per 100 000 population).

The AADRs for cancer among both white and nonwhite males rose from 145 and 61 per 100 000 population, respectively, to peaks in 1990—272 and 347 per 100 000 population, respectively—and declined thereafter. The AADRs for cancer were substantially greater among white males than among nonwhite males until 1960, when the AADRs for cancer among nonwhite males surpassed the AADRs for cancer among white males (Table 2).

The AADRs declined similarly among white and nonwhite males for enteritis and diarrhea, tuberculosis, and pneumonia and influenza; the AADRs among nonwhite males exceeded those among white males in all decades for tuberculosis and for pneumonia and influenza but only until 1980 for enteritis and diarrhea (Table 2).

The AADRs for UNMVIs generally declined among nonwhite males from 1920 to 2010; among white males, the AADRs for UNMVIs declined until 1990 and increased sharply until 2010. The AADRs for UNMVIs among nonwhite males exceeded those among white males in all decades except 2010. The AADRs for MVIs peaked at 48 per 100 000 population among white males in 1930 and generally declined thereafter; the AADRs for MVIs among nonwhite males reached a peak of 51 per 100 000 population in 1970 and then declined (Table 2).

In 1920, the AADRs for major CODs among nonwhite males exceeded those among white males only for infectious diseases and for unintentional UNMVIs. By 2010, the AADRs among nonwhite males exceeded those among white males for major cardiovascular diseases, cancers, and infectious diseases. From 1970 through 2010, the AADRs among black males generally exceeded those among nonwhite males for all conditions and followed similar temporal trends (Table 2).

Trends Among Males for Selected Combinations of Chronic, Infectious, and Unintentional Injury Conditions Combined (1900-2010)

The selected chronic diseases, infectious diseases, and unintentional injuries assessed in this analysis accounted for 54.3% of the AADRs among males in 1900 and 62.4% in 2010 (Table 1; Figure 2). Trends in the AADRs for these combined conditions varied markedly. Rates per 100 000 population for the selected chronic conditions rose from 670 in 1900 to a peak of 1165 in 1960, an increase of 75.4%, and then declined by 56.8% to 483 in 2010. Rates per 100 000 population for infectious conditions decreased from 631 in 1900 to 96 in 1950, a decrease of 84.9%, and then declined another 80.8% thereafter to 19 in 2010. Unintentional injury rates remained relatively steady through 1930, declining from 147 to 52 per 100 000 population from 1930 to 2010.

Figure 2.

Age-adjusted death rates for selected chronic diseases, selected infectious diseases, and unintentional injury, US males, 1900-2010. Selected infectious disease totals are the sum of the rates for influenza and pneumonia, tuberculosis, and enteritis and diarrhea. Selected chronic disease totals are the sum of the rates for heart disease, stroke, and all cancers; they do not include chronic lower respiratory diseases. Unintentional injury totals are the sum of the rates for unintentional motor vehicle injuries and unintentional nonmotor vehicle injuries. Rates are per 100 000 US 2000 standard population. Data for 1900-1930 are from death registration states; data for 1940-2010 are from the United States. The states providing mortality data are known as “death registration states”; they included 10 states and the District of Columbia in 1900 (40.5% of the US population) and gradually expanded to include all 48 states and the District of Columbia by 1933. Data source: tabulations of causes of death from published compilations and from public-use computer data files, as previously described.⁴

Trends in Relative Differences in AADRs Between Males and Females for All Causes (1900-2010) and Specific Causes by Race (1920/1970-2010)

The relative differences between male and female all-cause AADRs began a steady increase after 1920, with greater excesses among white males than among nonwhite males until 1990 (Figure 3). Excess all-cause death rates peaked in 1980 among white males and in 1990 among nonwhite and black males and declined thereafter. Despite recent declines, the excess in male AADRs was greater in 2010 (39.7%) than in 1900 (9.1%; Figure 3).

Figure 3.

Relative difference between male and female age-adjusted death rates (AADRs) for all causes and selected causes, by race (white, black, and nonwhite), calculated as ([male AADR − female AADR]/female AADR), 1900-2010, United States. Rates are per 100 000 US 2000 standard population. Data for 1900-1930 are from death registration states; data for 1940-2010 are from United States. The states providing mortality data are known as “death registration states”; they included 10 states and the District of Columbia in 1900 (40.5% of the US population) and gradually expanded to include all 48 states and the District of Columbia by 1933. Not all data were available in all years. Disaggregation of race beyond white and nonwhite did not occur until 1970. Data source: tabulations of causes of death from published compilations and from public-use computer data files, as previously described.⁴

The relative differences in AADRs between males and females varied markedly by cause (Figure 3). Trends for heart disease, cancers, and pneumonia and influenza followed all-cause trends. The AADRs for stroke differed little by sex. Of all categories of disease and injuries studied, excess male death rates were greatest for respiratory system cancers (peaking at an excess of 5.3-fold in 1960), chronic lower respiratory disease (3.9-fold excess for the first year of data in 1970), and MVIs (3.0-fold excess in 1910). AADRs for MVIs showed the greatest excess in male mortality during the entire period, with large excesses more recently for UNMVIs, respiratory system cancer, chronic lower respiratory disease, and tuberculosis.

Discussion

Our analysis of AADRs and major CODs by sex and race across 110 years in the United States demonstrates substantial progress in increasing longevity. However, among several age groups and racial populations, progress was interrupted early in the 20th century and again in 1970. Trends also suggest a need to further explore the factors contributing to sustained (and even increased) excess death rates among males compared with females. Changes in diagnostic and classification acumen may have influenced trends, particularly for chronic diseases,⁴ and improvement in the representation of the US population in vital records by the mid-20th century may also have affected trends. These changes, however, were unlikely to have substantially affected the major trends or trend differences observed in our analysis.

For several age groups of white and nonwhite males, death rate declines between 1900 and 2010 were interrupted by increases. One cluster of increases occurred among men aged ≥35 beginning in 1910 and continuing until 1940. Another cluster occurred among white and nonwhite males aged 15-44, beginning in 1970 and continuing into the next 2 decades for a few subgroups. This cluster, along with a similar cluster among females at the same time,⁴ precedes by more than a decade the onset of the US opioid overdose epidemic. The increase in UNMVI deaths among white men after 1990 may be associated with the opioid overdose epidemic.⁵

Although the AADRs for most major conditions declined during the study period, we found 3 principal exceptions—heart disease, all cancers, and MVIs, which rose substantially before declining. During the study period, male AADRs exceeded female AADRs in every decade. The greatest increase in the relative difference occurred during 5 decades: from 16% in 1930 to almost 65% in 1980—about 4 times the 1930 excess. Although the excess then dropped to about 40% in 2010, it was still higher than the excess of 9% in 1900. Death rates for respiratory system cancer had the largest excess, followed by MVIs, then, to a lesser extent, UNMVIs, and, more recently, chronic lower respiratory diseases.

Although data are scarce for the first half of the 20th century, we have potential explanations for differences in AADRs between males and females, including changes in social determinants, biological markers, and public health and medical interventions. Women have made relative gains in the 20th century in education, employment, and income—all associated with long-term health benefits.⁴ From 1900 to 1950, high school graduation rates among females increased from <30% to approximately 80%.⁶ Postsecondary education for women also increased.⁷ Between 1940 and 1970, the median years of schooling increased for nonwhite and white females.⁸ Increases in proportions of women in the workforce, married women working outside the home, and women in professions—and a narrowing of the gap between earnings among men and women⁷—all may have contributed to improved economic well-being for women.⁹ These factors may partially explain the increased mortality gap between males and females. The health of women has improved more rapidly overall than the health of men, in part because of women’s gains in socioeconomic position.

Another major factor contributing to excess male deaths throughout the 20th century is smoking. Waldron estimates that approximately half of the male–female difference in mortality in the United States is attributable to smoking.¹⁰ Men took up cigarette smoking earlier in the 20th century than women did; more men smoked, and they smoked more cigarettes. Men also began to quit smoking before women did.¹¹ These differences may largely account for the differences in death rates for cardiovascular disease, cancers (respiratory system cancers in particular), and lower respiratory disease—the initial rise and subsequent decline as well as the earlier peak for heart disease than for cancers. Despite reductions in smoking prevalence among males since the 1960s, smoking prevalence is higher among males than among females.¹¹

During the 20th century, rates of hypertension declined more rapidly among females than among males.¹² Men generally have had higher rates of hypertension than women since 1960.¹³ Recent data¹⁴ suggest that rates of high total cholesterol among males and females are nearly equal but that significantly more males have low levels of high-density lipoprotein. Combined, these excesses among males are consistent with higher death rates for heart disease among males. In contrast, females have had higher rates of obesity since 1960,¹⁵ a risk factor also associated with heart disease mortality. Males and females report similar rates of physical activity.¹⁶

The apparent lack of association of trends in smoking and hypertension with stroke mortality is puzzling. However, a study from Britain indicates that 2 distinct forms of stroke have different risk factors and that trends in cerebral infarction mortality follow trends in coronary heart disease, while trends in hemorrhagic stroke do not.¹⁷

Male engagement with the health care system differs from that of females and may also contribute to excess male mortality. Among persons aged >18, females are more likely than males to report having a usual place of health care¹⁸ and more likely than males to visit physician offices and hospital outpatient departments.¹⁹ Males without a usual source of care may not receive preventive services available for 9 of the 10 major CODs.²⁰ The use of 8 recommended preventive services is greater among females than among males.²¹

Remarkable among the comparisons of male and female mortality trends is the male excess in MVI deaths. Although all-cause excess trends do not surpass 69%, MVI trends generally range from 122% to 300% (ie, a male maximum rate that is 4 times that of females). In 2017, men drove approximately 1.09 times as many miles as women,²² and male drivers were associated with 2.43 times as many motor vehicle fatalities.²³

Plausible explanations for trends in male excess MVI fatalities are available. In a US sample in 2006 and 2008, men reported 3.1 times as many episodes of alcohol-impaired driving as did women.²⁴ Men are also approximately 50% more likely than women to drive above the speed limit,²⁵ and 9%-13% of men are less likely than women to use seat belts.²⁶ Moreover, male and female drivers involved in speeding-related motor vehicle fatalities are more likely to have elevated blood-alcohol levels and to not use seat belts than nonspeeding drivers.²⁷ Diverse programs and policies are effective in reducing motor vehicle fatalities.^28,29

Although male mortality declined markedly from 1900 to 2010, substantial room for improvement exists. As recognized in Healthy People 2020, much can be achieved by the equitable distribution of social determinants of health.³⁰ Further health gains can be achieved by reductions in behaviors such as smoking, physical inactivity, and the consumption of foods that promote obesity, hypertension, and high cholesterol. Finally, the full use of multiple preventive and curative health interventions would also contribute to increased longevity.

Limitations

This study had several limitations. First, mortality is not an optimal measure of the burden of disease and suffering and, thus, is not ideal for planning and monitoring interventions to improve public health. However, mortality is a measure available over the time span examined. Second, available mortality data are imperfect for many reasons. Because data on all states in the United States became available only in 1933, early rates may not be representative of the US population, and early trends are assessed in different populations over time. Third, analysis of mortality rates is limited by the consistency and completeness of data available. Completeness of population and mortality reporting is likely to have increased substantially since 1900. Incomplete reporting leads to underestimates of death rates.

Fourth, the comparability of ICD categories between revisions is commonly assessed by the comparability ratio, which compares the number of deaths attributed to a given cause using the previous ICD revision to the number of deaths in the same year attributed to the comparable cause for the subsequent ICD revision. For example, comparability ratios among 56 causes for the sixth and fifth revisions range from 0.57 (for “sepsis of pregnancy, childbirth, and puerperium”) to 1.79 (for “acute nephritis and nephritis with edema, including nephrosis”).³¹ There is also evidence that coronary heart disease is often overdiagnosed,³² which may inflate coronary heart disease death rates and, if overdiagnosis has changed over time, make apparent trends inaccurate. Finally, racial data were not specific for nonwhite groups (eg, black, Asian, American Indian) until recent decades, and racial classification in health statistics is problematic (eg, inconsistency of coding in numerator and denominator data).³³ The specification of black populations among nonwhite populations in mortality data became available only in 1970.

Conclusion

The mortality data presented in our study demonstrate that both males and females are likely to have been harmed by exposures such as smoking and motor vehicles early in the 20th century but are likely also to have benefited from the environmental, behavioral, social, technological, and medical advances of the century—especially in the second half. Both sexes gained >30 years of life expectancy during the century. Nonetheless, male death rates continue to exceed those of females at every life stage, and the relative difference in 2010 exceeded that in 1900. Some excess may be attributable to risk-taking behavior among men, which may include lower rates of use of health care services and greater participation in harmful behaviors. Males also have higher rates of some biological markers than women. All of these features provide corresponding opportunities for health improvement, and evidence-based interventions are available to do so.³⁴

Footnotes

Acknowledgments

The authors recognize the contributions of Dorit Carmelli, PhD, and Li Shu-Chuan Cheng, PhD, from SRI International for compiling the mortality data prior to 1968 and participating in the design and preliminary analysis of these data. We are grateful for the guidance of Daniel Friedman for methodological advice and the CDC WONDER team for their technical support and assistance.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The work of Dr Parrish for this publication was supported by cooperative agreement grant number U38OT000216 from the Centers for Disease Control and Prevention (CDC). The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of the US Department of Health and Human Services or CDC.

ORCID iD

Wanda K. Jones, DrPH

References

Centers for Disease Control and Prevention. Death rates and life expectancy at birth. 2018. https://data.cdc.gov/NCHS/NCHS-Death-rates-and-life-expectancy-at-birth/w9j2-ggv5. Accessed June 26, 2019.

Jones

Podolsky

Greene

. The burden of disease and the changing task of medicine. N Engl J Med. 2012;366(25):2333–2338. doi:10.1056/NEJMp1113569

Wilbur

King

. Manual of International Classification of Causes of Death. Washington, DC: US Government Printing Office; 1902.

Hahn

Chang

Parrish

Teutsch

Jones

. Trends in mortality among females in the United States, 1900-2010: progress and challenges. Prev Chronic Dis. 2018;15:E30. doi:10.5888/pcd15.170284

Dowell

Arias

Kochanek

, et al. Contribution of opioid-involved poisoning to the change in life expectancy in the United States, 2000-2015. JAMA. 2017;318(11):1065–1067. doi:10.1001/jama.2017.9308

Heckman

LaFontaine

. The American high school graduation rate: trends and levels. Rev Econ Stat. 2010;92(2):244–262. doi:10.1162/rest.2010.12366

Caplow

Hicks

Wattenberg

. The First Measured Century: An Illustrated Guide to Trends in America, 1900-2000. Washington, DC: American Enterprise Institute; 2001.

US Census Bureau. Historical Statistics of the United States, Colonial Times to 1970. Washington, DC: US Department of Commerce, US Census Bureau; 1975.

Katz

Stern

Fader

. Women and the paradox of economic inequality in the twentieth century. https://repository.upenn.edu/cgi/viewcontent.cgi?article=1047&context=spp_papers. Accessed June 26, 2019.

10.

Waldron

. Patterns and causes of gender differences in smoking. Soc Sci Med. 1991;32(9):989–1005. doi:10.1016/0277-9536(91)90157-8

11.

Jamal

Phillips

Gentzke

, et al. Current cigarette smoking among adults—United States, 2016. MMWR Morb Mortal Wkly Rep. 2018;67(2):53–59. doi:10.15585/mmwr.mm6702a1

12.

Goff

Howard

Russell

Labarthe

. Birth cohort evidence of population influences on blood pressure in the United States, 1887-1994. Ann Epidemiol. 2001;11(4):271–279. doi:10.1016/S1047-2797(00)00224-6

13.

Burt

Cutler

Higgins

, et al. Trends in the prevalence, awareness, treatment, and control of hypertension in the adult US population data from the health examination surveys, 1960 to 1991. Hypertension. 1995;26(1):60–69. doi:10.1161/01.hyp.26.1.60

14.

Carroll

Fryar

Nguyen

. Total and high-density lipoprotein cholesterol in adults: United States, 2015-2016. NCHS Data Brief. 2017;(290):1–8.

15.

Flegal

Carroll

Kuczmarski

Johnson

. Overweight and obesity in the United States: prevalence and trends, 1960-1994. Int J Obes Relat Metab Disord. 1998;22(1):39–47.

16.

Brownson

Boehmer

Luke

. Declining rates of physical activity in the United States: what are the contributors? Annu Rev Public Health. 2005;26:421–443. doi:10.1146/annurev.publhealth.26.021304.144437

17.

Lawlor

Smith

Leon

Sterne

Ebrahim

. Secular trends in mortality by stroke subtype in the 20th century: a retrospective analysis. Lancet. 2002;360(9348):1818–1823. doi:10.1016/S0140-6736(02)11769-7

18.

Blackwell

Villarroel

Clarke

. Tables of Summary Health Statistics for US Adults: 2014 National Health Interview Survey. Hyattsville, MD: National Center for Health Statistics; 2015.

19.

National Center for Health Statistics. Health, United States, 2016: With Chartbook on Long-Term Trends in Health. Report no. 2017-1232. Hyattsville, MD: National Center for Health Statistics; 2017.

20.

Fox

Shaw

. Receipt of selected clinical preventive services by adults—United States, 2011-2012. MMWR Morb Mortal Wkly Rep. 2015;64(27):738–742.

21.

Centers for Disease Control and Prevention. Use of selected clinical preventive services among adults—United States, 2007-2010. MMWR Morb Mortal Wkly Rep. 2012;61(suppl):1–2.

22.

US Department of Transportation, Federal Highway Administration. National Household Travel Survey. 2017. https://nhts.ornl.gov. Accessed June 26, 2019.

23.

Insurance Institute for Highway Safety. Fatality facts 2017: gender. https://www.iihs.org/topics/fatality-statistics/detail/gender . Accessed October 24, 2019.

24.

Bergen

Shults

Beck

Qayad

. Self-reported alcohol-impaired driving in the US, 2006 and 2008. Am J Prev Med. 2012;42(2):142–149. doi:10.1016/j.amepre.2011.10.015

25.

National Highway Traffic Safety Administration. National Survey of Speeding and Unsafe Driving, Attitudes and Behaviors. Volume II—findings report. http://www.trb.org/Main/Blurbs/153203.aspx. Published 2003. Accessed June 26, 2019.

26.

Centers for Disease Control and Prevention. Vital signs: nonfatal, motor vehicle-occupant injuries (2009) and seat belt use (2008) among adults—United States. MMWR Morb Mortal Wkly Rep. 2011;59(51):1681–1686.

27.

US Department of Transportation, National Highway Traffic Safety Administration. Traffic safety facts: 2014 data. Speeding. https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/812265. Accessed June 26, 2019.

28.

The Guide to Community Preventive Services. Findings for motor vehicle injury. https://www.thecommunityguide.org/content/task-force-findings-motor-vehicle-injury. Published 2019. Accessed June 26, 2019.

29.

Goodwin

Thomas

Kirley

Hall

O’Brien

Hill

. Countermeasures That Work: A Highway Safety Countermeasure Guide for State Highway Safety Offices: 2015. Washington, DC: US Department of Transportation, National Highway Traffic Safety Administration; 2015.

30.

US Department of Health and Human Services, Office of Disease Prevention and Health Promotion. Healthy People 2020. Washington, DC: Office of Disease Prevention and Health Promotion; 2010.

31.

Faust

Dolman

. Comparability Ratios Based on Mortality Statistics for the Fifth and Sixth Revisions, United States, 1950. Washington, DC: US Department of Health, Education, and Welfare, Public Health Service, National Center for Health Statistics; 1964.

32.

Agarwal

Norton

Konty

, et al. Overreporting of deaths from coronary heart disease in New York City hospitals, 2003. Prev Chronic Dis. 2010;7(3):A47.

33.

Rosenberg

Maurer

MacDorman

Hoyert

Scott

. Quality of death rates by race and Hispanic-origin; a summary of current research, 1999. Vital Health Stat 2. 1999;(128):1–13.

34.

Community Preventive Services Task Force. The community guide. https://www.thecommunityguide.org. Published 2019. Accessed June 26, 2019.