The widening of last century's life expectancy gap between the sexes has been attributed mostly to smoking—men more than women. But in the last few decades, the number of female smokers has increased relative to their male counterparts, effectively narrowing the gap. In 1999, the latest year for which figures are available, a woman's life expectancy at birth was 79.4, and for a man's, 73.9.
These fluctuations are not confined to the United States. For example, while not citing specific reasons, the United Kingdom's Government Actuary's Department reports that in 1971, a man was expected to live until nearly 69; a woman, 75. By 2011, the ages are expected to be 77.4 and 81.6, respectively. The smoking scenario has been an important factor in explaining demographics in developed countries, and now it is becoming important in developing countries as well, notes Ingrid Waldron, professor of biology, University of Pennsylvania.1
Of the sex-based differences discussed in this series, none seems more amenable to lifestyle changes than longevity. In the 18th and part of the 19th centuries, for example, men probably outlived women because of the increased mortality of women during childbirth. In the early 20th century, the hypervirulent flu that ravaged the population mortally affected more men than women. And, by the end of the 20th century, the life expectancy of both sexes started leveling off as more women became addicted to nicotine. "Throughout the last 100 years, women have lived longer than men by various numbers of years," says lead statistician Robert N. Anderson, Mortality Statistics Branch, Division of Vital Statistics, National Center for Health Statistics, Centers for Disease Control and Prevention.2
Since prehistoric times, tremendous variability has existed between the sexes' life expectancies, says Samuel Preston, the Frederick Warren Professor of Demography and dean of the School of Arts and Sciences, University of Pennsylvania. At that point, he says, it was a narrow difference. Since then, the gap has fluctuated. "An important component about identifying what is constant and what is variable about sex differences is that most of the differences at a moment in time and place seem to be a product of circumstances in that time and place," he says.
It is "very unusual," notes Anderson, to see female life expectancy lower than that of men, and when it does happen, it is due to high maternal mortality during childbirth. Details prior to 1900 regarding maternal mortality are sketchy, he says, but the evidence suggests that in the 18th century to early 19th century, life expectancy for men was actually higher.
Which is It?
Smoking as the overriding explanation for sex-based differences in life expectancy is a "fair assessment," but not the whole story, says Preston. As the percentage of male smokers increased, their incidence of heart disease also increased, with concomitant widening of the life-expectancy gap. Then, as women started lighting up, the gap lessened as the incidence of heart disease in men decreased compared with women. "Now smoking patterns are changing—men are quitting at higher rates than women, and this is why the gap is narrowing," says Preston, although women's rates have decreased overall.
But medical treatment also plays a role. "Heart disease has been viewed as a male disease—not a female [one]—and men may be more likely to get the intervention," says Anderson. A decline in the incidence of heart disease also is related to improvements in diet and in medical care, agrees Waldron. The decline is strong for both sexes, and that also reduces the longevity gap. Waldron notes, however, that the largest drops in sex-mortality ratios are seen in association with lung cancer, chronic obstructive-pulmonary diseases like emphysema, and motor-vehicle accidents.3
Two recent studies show that women respond differently than men to cardiac treatments. At Emory University in Atlanta, investigators found that women younger than 50 were about three times more likely to die after coronary-artery bypass compared with men of the same age group.4 Researchers at Albert Einstein College of Medicine in New York City found that women were more than twice as likely than men to die in the hospital after angioplasty.5
A study from the Mayo Clinic in Rochester, Minn., found that the rate of heart attacks for women increased by 26% during the 1980s and early 1990s and declined by 8% for men during the same period.6 The researchers attribute these statistics to lifestyle changes in women and heart-attack prevention efforts for men.
Influence of Influenza
And its impact was striking; the researchers found that in the years following the epidemic, life expectancy of men increased faster than that of women.7 More men died in 1918 because they were unhealthy; flu deaths in the United States totaled 500,000, or 0.5% of the total population at that time. Surviving males were healthier, however, and the sex-based gap did not return until the 1930s. Noymer and Garenne reported that an identical pattern was seen with tuberculosis patients at that time as well.
The flu typically kills those who are youngest and oldest; middle-aged people experience serious symptoms but usually survive. In 1918 the strain was hypervirulent and resulted in extremely high mortality. The epidemic was also "idiosyncratic," says Noymer, in that it killed not only many in the expected age groups, but those in the middle as well.
The researchers then studied statistics regarding tuberculosis (TB), a common chronic ailment of the time. Unlike the flu, TB occurs mostly in middle-aged people. In 1918 an unusual situation occurred: flu cases and TB cases overlapped with respect to patient age. The reason? TB and flu both affect the lungs, so people who had TB were simply more susceptible to the flu. Also, TB is a slow killer, taking many years, so a pool of TB-infected people already existed. The flu greatly depleted this pool, infecting and killing more men than women, because men were at greater risk due to occupational exposure. This discrepancy set up a selection effect; many 1918 flu deaths were among those with TB. Mortality due to TB dropped after 1918, more so among men than women.
The life expectancy gap didn't regain its preepidemic level until the 1930s. Men caught up to women after 1918 largely because of increased mortality due to heart disease, which in turn is related to increases in male smoking. S. Jay Olshansky, a biodemographer at the University of Illinois at Chicago School of Public Health, calls these conclusions "fascinating and unexpected, and deserving of further study."
Despite the research, "differences in life expectancy are something that we just take for granted," says Shiriki Kumanyika, associate dean, Health Promotion and Disease Prevention, Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine and an Institute of Medicine panel member.8 "And then you start wondering, well, why? Is there a biological reason for this?"
Perhaps there is. Researchers have found biological reasons for sex-based differences regarding drug metabolism, autoimmunity, neurological processes, and genetics. Perhaps, they will establish contributing reasons for life expectancy as well.
1. I. Waldron, "Recent trends in sex mortality ratios for adults in developed countries," Social Science & Medicine, 36:451-62, 1993.
2. US Decennial Life Tables for 1989-91 (www.cdc.gov/nchs/data/lifetables/dec89_1_3.pdf); United States Life Tables, 1998 (www.cdc.gov/nchs/data/nvsr/nvsr48/nvs48_18.pdf).
3. I. Waldron, "Trends in gender differences in mortality: Relationships to changing gender differences in behavior and other causal factors." In Gender Inequalities in Health, E. Annandale, K. Hunt, eds., Buckingham, UK: Open University Press, 2000.
4. V. Vaccarino et al., "Sex differences in hospital mortality after coronary artery bypass surgery: Evidence for a higher mortality in younger women," Circulation, 105:1176-81, Feb. 18, 2002.
5. B. Vakili et al., "Sex-based differences in early mortality of patients undergoing angioplasty for first acute myocardial infarction," Circulation, 104:3034-8, Dec. 18, 2001.
6. V.L. Roger et al., "Trends in the incidence and survival of patients with hospitalized myocardial infarction, Olmsted County, Minnesota, 1979 to 1994," Annals of Internal Medicine, 136:341-8, March 5, 2002.
7. A. Noymer, M. Garenne, "The 1918 influenza epidemic's effects on sex differentials in mortality in the United States," Population and Development Review, 26:565-81, 2000.
8. Institute of Medicine, Exploring the Biological Contributions to Human Health: Does Sex Matter? T.M. Wizemann, M. Pardue, eds., Washington, D.C.: National Academy Press, 2001. (www.nap.edu)