SYPHILIS HAS PROVIDED valuable lessons for clinicians since the days of William Osler1 and offers equally useful instruction to epidemiologists and practitioners of public health, especially those interested in sexually transmitted diseases (STDs).2 The distribution and trends of syphilis in the population are influenced by several determinants including: biologic factors (e.g., characteristics intrinsic to the Treponema pallidum organism, host responses to infection); sexual behaviors; biomedical factors (e.g., availability of diagnostic tests and effective treatment); availability of and access to healthcare; healthcare seeking behaviors; public health efforts to prevent and control syphilis (e.g., screening programs, case finding and partner notification activities); population factors (e.g., urbanization, poverty, gender ratios); and sociocultural factors (e.g., disintegration of family, stigmatization of those with the disease, patterns of illicit drug use). The focus of this article is to describe past and current trends in the distribution of syphilis in different populations using nationally collected surveillance data.
Sources of Data and Calculation of Rates
The Venereal Disease Division of the U.S. Public Health Service was created in 1918 and collection of surveillance data for syphilis by state health departments began as early as the 1930s. Consistently compiled national data were available dating back to 1941.3 Data on reported cases of syphilis in the primary and secondary (P&S) stages were used for trend calculations because those cases best represent incident cases (i.e., newly acquired infections within the evaluated time period) of syphilis. Rates of P&S syphilis per 100,000 persons were calculated using population denominators from the Bureau of the Census.3 Rates of congenital syphilis per 100,000 live births were calculated using natality data provided by the National Center for Health Statistics.3
Long-Term Trends in the United States and Other Industrialized Countries
In 1947, 93,545 cases of P&S syphilis were reported in the United States (66.4 cases per 100,000 persons) (Figure 1). By 1956, rates had declined to 3.9 cases per 100,000 persons. Several factors contributed to this decline. First, penicillin for treatment of syphilis became widely available after World War II.4,5 Following the recommendations of Surgeon General Thomas Parran,6 the Public Health Service developed a nationwide system of rapid treatment centers to curb the spread of syphilis in the military.7 In 1947, the rapid treatment centers admitted 185,000 persons who were mostly military service personnel returning from overseas. Second, public health programs began a variety of activities to enhance case finding that included widespread serologic testing in many settings (e.g., premarital, prenatal, pre-employment, hospital admission, and military enlistment and separation) and interviewing of cases to identify and treat sex partners. Third, changes in sexual behavior and in cultural norms (e.g., more conservative sexual behavior patterns before the 1960s and 1970s and wider acceptance of government's public health measures, such as contact tracing) may have contributed to the effectiveness of those, mostly biomedically based, interventions during this era.
From the nadir in 1956 to the present, P&S syphilis rates have shown recurrent peaks and troughs in approximately 10-year cycles (Figure 1). The overall trend during that period was toward increasing rates. The most recent epidemic peaked in 1990 at a rate of 20.3 cases per 100,000 persons, the highest rate since 1949. No single reason can explain this progressively increasing trend. Different risk groups were affected during different epidemic cycles. The increases from 1960 to 1983 were mostly in men, and the increase between 1985 and 1990 affected both men and women.
Published data from other industrialized countries showed similar declines after the advent of penicillin therapy. In contrast to the United States, rates have remained very low throughout the 1980s and 1990s in those countries (Figure 2).8–10.
Although syphilis rates are highest in the southern United States, similar rates were seen in many states at an earlier time. In 1946, the majority of states reported rates of P&S syphilis exceeding 50 cases per 100,000 persons (Figure 3). However, 11 of 12 states with rates exceeding 100 cases per 100,000 persons were located in the South.‡ By 1958, rates of P&S syphilis had declined below 10 cases per 100,000 persons for all states except Georgia and South Carolina (Figure 3). Between 1985 and 1992 all regions experienced an epidemic of syphilis; the South peaked in 1990 at a rate of 33.7 cases per 100,000 persons, the highest rate for any region during this period (Figure 4). Between 1963 and 1993, the South had higher rates of P&S syphilis than any other region (Figure 4), despite epidemics that occurred at different times in different regions. The reasons for these regional differences may include differential availability of healthcare services, socioeconomic differences, and differences in demographic composition.
When smaller geographic areas (e.g., counties) are examined, the focal distribution of syphilis in the 1990s is apparent. In 1993, 2,006 (64%) of 3,116 counties in the United States reported no cases of P&S syphilis. Only 461 (including 424 counties located in the South) of the remaining 1,110 counties, had rates higher than 10 cases per 100,000 persons.3 Those counties accounted for 81% of all reported P&S syphilis cases in the United States.
Syphilis in recent years is mostly concentrated in urban areas in most regions of the country. The overall rate of P&S syphilis for 63 large cities (with > 200,000 population) has remained almost twice the U.S. total rate throughout the 1980s and 1990s.3 However, in the South, high rates of P&S syphilis occurred in urban and rural areas.11 In a study in North Carolina, rural rates surpassed urban rates.12
Distribution of Syphilis by Racial and Ethnic Groups
In the United States, syphilis has disproportionately affected minority populations, and the magnitude of this differential effect has tended to increase. Rates in non-Hispanic white men declined steadily between 1983 and 1993; low rates in non-Hispanic white women changed little or declined during the same period (Figure 5). In the 1990s, rates for whites were comparable to the rates in other industrialized countries.
The most recent epidemic of syphilis mostly affected heterosexual black Americans. Between 1985 and 1990, rates in blacks increased 165% from 51.1 to 142.6 cases per 100,000 persons. During the peak of the epidemic, blacks accounted for more than 80% of reported cases of P&S syphilis. The postulated causes for this explosive epidemic in blacks included a dramatic increase in the use of crack cocaine accompanied by an increase in exchange of sex for money and drugs; a decreased effectiveness of STD prevention and control programs; increasing poverty and disenfranchisement of minorities; and urban decay, resulting in an environment of increasing crime, violence, and dysfunctional families.13–15 Since 1990, syphilis has continued to disproportionately affect blacks, despite declining rates for all other racial and ethnic groups. Hispanics rates have been between rates for whites and blacks and have declined since 1988.
Before 1980, data on race and ethnicity were collected only in white and non-white categories. Non-whites have always had higher rates than whites in the United States, but the minority-to-white rate ratio has changed (Table 1). In the 1950s and 1960s the minority-to-white rate ratio was about 15:1. In the 1970s and early 1980s, the rate ratio declined and was 8:1 in 1981. This decline is believed to have been largely due to increasing rates in gay white men rather than declining rates in minorities. In response to public health messages to slow the spread of human immunodeficiency virus infection, rates of sexually transmitted diseases, including syphilis, in gay white men have declined throughout the 1980s and 1990s. As a result, the minority-to-white syphilis rate ratio has increased steadily, reflecting the increasing concentration of syphilis in minority populations.
The high rates of P&S syphilis recognized in the South may be due in part to the differential distribution of minority populations in the country. When rates were adjusted for the racial and ethnic distribution of the population, the differences in rates between states in the South and those in other regions decreased (Figure 6). In 1993, the race-adjusted rates per 100,000 population of P&S syphilis by region were: 3.7 in the West, 11.7 in the Midwest, 5.3 in the Northeast, and 12.6 in the South. However, high rates persisted in many southern states, indicating that racial and ethnic distribution of the population does not alone explain the high rates in this region.
Higher rates in southern blacks have been postulated as a cause of regional differences. However, between 1981 and 1993, rates of P&S syphilis for blacks in different regions peaked at similar levels, although peaks occurred in different years, depending on when epidemics occurred in different regions (Figure 7). The reasons for peaking of syphilis rates in different geographic areas at different times remain unclear. Epidemiological modeling suggests that syphilis epidemics may be initiated by introduction of infection into a small core of susceptible persons with multiple sexual partners.16 The core members tend to transmit disease to large numbers of individuals who are outside the core creating the epidemic. However, these epidemics are difficult to sustain because in time the core members will themselves become noninfectious, either by treatment or by progression to latent stage disease, and recruitment of additional susceptible core members occurs relatively slowly. Thus, temporal movement of epidemics into different regions of the United States may be due to differential development of susceptible core populations (e.g., introduction of crack cocaine and behaviors surrounding its introduction) and differential introduction of syphilis into these populations.
Distribution of Syphilis by Gender and Age
Men consistently have had higher rates than women and have cycled upward during the past 30 years (Figure 8). Women cycled at low levels until the epidemic of the 1980s, when rates in women, primarily black women, increased. The overall male-to-female rate ratio peaked at 3.5:1 in 1980 during the height of syphilis transmission in gay men and has plummeted since then to approximately 1:1 (Figure 9). In 1993, six states (Arkansas, Delaware, Louisiana, Mississippi, Ohio, and South Carolina) reported higher rates of P&S syphilis for women than men.3 Although the declining male-to-female rate ratio suggests a disproportionate increase in rates of syphilis among women, the causes for this trend remain unclear. Prostitution related to the epidemic of crack cocaine use may be continuing to play a role; however, other factors must be considered. Female cases may be identified and reported more efficiently by STD programs, especially if programs prioritize women over men for diagnosis and treatment to prevent congenital syphilis. In addition, other broad changes in sexual behaviors may have contributed to this trend. Since the 1980s, the available pool of low-risk (e.g., those having one sexual partner at a time) heterosexual male partners has declined disproportionately in some groups. Thus, women in those groups have a higher risk of infection, although the number of sexual partners per woman has remained stable. Declines in long-term monogamous relationships may also have occurred and resulted in increased risk of infection.17,18
The age distribution of syphilis in the 1950s showed a similar pattern to other STDs in that much of the disease was in adolescents and young adults in their 20s for both genders (Figure 10). Adolescents and young adults are traditionally believed to be at higher risk for STDs during their premarital years because they are more likely to have multiple sexual partners, engage in high-risk behaviors, and select higher risk partners. As they get older and marry or establish long-term monogamous relationships, the risk for STDs declines.
By the end of the 1970s, the age distribution of syphilis had changed, especially for men; rates were then highest for men from their early 20's to their early 30's (Figure 10). This may have been due in part to a larger proportion of cases in gay men who continued to have high risk behaviors into their 30's. For women, the age distribution did not change much until the epidemic of the 1980s. At the peak of the epidemic in 1990, high rates for men and women persisted into their early 30's. This pattern contrasts sharply with other bacterial STDs that continue to primarily affect teens and young adults less than 25 years of age. Since the epidemic subsided, the older age distribution for both genders has persisted. Because blacks account for the majority of recent cases, the current age profile primarily reflects the pattern in this group. A persistence of high rates into mature adulthood may indicate that a larger proportion of the affected groups have not developed long-term relationships with one partner. These persons continue to maintain behavior patterns of much younger persons, keeping them at risk for syphilis. Core populations at highest risk for syphilis may be composed of persons less likely to form long-term relationships as they enter middle adulthood.
Congenital syphilis is the most serious outcome of syphilis in women. Trends in congenital syphilis cases closely parallel trends in rates of P&S syphilis for women (Figure 11). In 1993, blacks accounted for the majority of congenital syphilis cases; 2,300 (72%) of 3,173 reported cases were of black persons. The rate of congenital syphilis was higher for blacks (344.9 cases per 100,000 live births) and Hispanics (96.3 cases per 100,000 live births) than for non-Hispanic whites (6.1 cases per 100,000 live births).3 Although the currently used surveillance case definition19 includes some cases that may not have been infected, the public health impact of mothers arriving at delivery with evidence of untreated or inadequately treated syphilis is apparent: more than 70% of these infants were treated with 10 days of penicillin therapy, usually in a hospital (Centers for Disease Control and Prevention, unpublished data).
Other Ecologic Factors Associated With the Distribution of Syphilis
There do not appear to be any biologic reasons to explain the disparities in syphilis rates between different racial and ethnic groups. Race and ethnicity are more likely to be easily measured markers that correlate with other, complex risk factors or risk markers. Ecologic correlations between rates of P&S syphilis and indicators of socioeconomic status (e.g., poverty levels, percentage of single head-of-household families); indicators of community well-being, (e.g., violent crime levels, drug-related arrests); availability of and accessibility to health care (e.g., percentage of population with health insurance, number of physicians); and levels of urbanization (e.g., population density) may provide clues to the reasons for the disparities in syphilis rates among different communities other than racial and ethnic differences. Preliminary results from a county-based ecologic study showed significant correlations in the South between rates of P&S syphilis and female head-of-household and violent crime, as well as black race (Table 2). In addition to those factors, high levels of urbanization correlated with syphilis rates in non-Southern areas.
Implications for Future Prevention of Syphilis
A great deal of attention has been paid to biomedical approaches to syphilis control and prevention. Results attributed to this approach have been mixed. In the preface to his book, No Magic Bullet, Allan Brandt states the that the biomedical model as the basis for syphilis control is too restrictive; rather, this disease is defined by social conditions, environmental phenomena, and other variables that also must be addressed.20 The trends in syphilis since the advent of penicillin demonstrate the truth of that premise. The Public Health Service maximized implementation of the biomedical model during the 1950s, and eradicated the disease from much of the United States, but failed to eliminate it entirely.6 Since then, syphilis has either remained or become a significant public health problem for certain groups. These groups have changed over time, reflecting the changes in social, economic, behavioral, and cultural norms. During the 1970s, the sexual revolution and gay liberation movements created groups at risk for both syphilis and for human immunodeficiency virus infection. In the 1980s, crack cocaine and crack-related prostitution, increasing urbanization and poverty, and disintegration of marriages and families contributed to the spread of syphilis in minority populations. It is likely that new groups who are at risk for syphilis will continue to emerge. Prevention efforts based mostly on the biomedical model have been unable to prevent epidemics of syphilis in the past and probably will be unable to do so in the future. Rather, public health efforts must address the broader issues in communities that result in the emergence of high-risk groups.
For more than 50 years, syphilis control and prevention in the United States has been largely carried out through categorical STD programs based in state health departments and funded in part by the federal government. Although reliance on categorical programs has produced important successes, the public health infrastructure to address health-related factors affecting the entire community has been insufficient. Today, STDs occur in populations having multiple complex socioeconomic, behavioral, and cultural risk factors for adverse health outcomes. Sexually transmitted diseases cannot be prevented in these populations without also solving some of these problems. The challenge for STD programs in the future will be to develop interventions that are community-based, and to address broader public health issues while keeping scarce resources focused on STDs.
1. Osler W. Internal Medicine as a Vocation. In: Aequanimitas: With Other Addresses to Medical Students, Nurses and Practitioners of Medicine, 3rd ed., Philadelphia, 1932.
2. Rolfs RT, Cates W. The perpetual lessons of syphilis. Arch Dermatol 1989; 125:107–109.
3. Division of STD/HIV Prevention, Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance, 1993. US Department of Health and Human Services, Public Health Service. Atlanta: Centers for Disease Control and Prevention, 1994.
4. Mahoney JF, Arnold RD, Harris A. Penicillin treatment in early syphilis: a preliminary report. J Vener Dis Inform 1943; 24:355–357.
5. Brown WJ, Donohue JF, Axnick NW, Blount JH, Jones OG, Ewen NH. Syphilis and Other Venereal Diseases. Cambridge, Massachusetts: Harvard University Press, 1970.
6. Parran T. Shadow on the Land: Syphilis. New York: Reynal & Hitchcock, 1937:246–287.
7. U.S. Department of Health, Education, and Welfare. The eradication of syphilis: a task force report to the Surgeon General, Public Health Service, on syphilis control in the United States. Public Health Service Publication No. 918. Washington DC: U.S. Government Printing Office, 1962.
8. Piot P, Islam MQ. Sexually transmitted diseases in the 1990s: global epidemiology and challenges for control. Sex Transm Dis 1994; 21(suppl 2)S7–S13.
9. Catchpole MA. Sexually transmitted diseases in England and Wales: 1981–1990. Communicable Disease Report Review 1992; 2:R1–R7.
10. Kohl PK. Epidemiology of sexually transmitted diseases: what does it tell us? Sex Transm Dis 1994; 21(suppl 2)S81–S83.
11. Nakashima AK, Rolfs RT, Ladan AI. Trends in the urban and rural distribution of primary and secondary syphilis, United States, 1986–1990. Presented at the 9th International Society for Sexually Transmitted Diseases Research, Banff, Alberta, Canada, October 6–9, 1991 (Abstract).
12. Thomas JC, Kulik AL, Schoenbach VJ. Syphilis in the South: rural rates surpass urban rates in North Carolina. Am J Public Health 1995; 85: 1119–1122.
13. Marx R, Aral So, Rolfs RT, Sterk CE, Kahn JG. Crack, sex, and STD. Sex Transm Dis 1991; 18:92–101.
14. Aral SO, Holmes KK. Sexually transmitted diseases in the AIDS era. Sci Am 1991; 264:62–69.
15. Rolfs RT, Nakashima AK. Epidemiology of primary and secondary syphilis in the United States, 1981 through 1989. JAMA 1990; 164:1432–1437.
16. Oxman GL, Smolkowski K, Noell J. Mathematical modeling of epidemic syphilis transmission: implications for syphilis control programs. Sex Transm Dis 1996; 23:30–39.
17. Aral SO. The social context of syphilis persistence in the southeastern United States. Sex Transm Dis 1996; 23:9–15.
18. Kilmarx PH, St. Louis ME. The evolving epidemiology of syphilis. Am J Public Health 1995; 85:1053–1054.
19. Centers for Disease Control and Prevention. Guidelines for prevention and control of congenital syphilis. MMWR 1988; 37:S1.
20. Brandt AM. No Magic Bullet. New York: Oxford University Press, 1985:4.
‡The “South” is defined by the Bureau of the Census and includes Alabama, Arkansas, Delaware, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, Oklahoma, North Carolina, South Carolina, Tennessee, Texas, Virginia, and West Virginia.