Objective. Hypospadias is a common congenital anomaly, the cause of which is unknown. Unexplained increases in the rates of hypospadias occurred in five European countries in the 1970s and 1980s. We examined data from two birth defects surveillance systems in the United States for evidence of similar trends.
Methodology. The Metropolitan Atlanta Congenital Defects Program (MACDP) provided birth prevalence rates from 1968 to 1993. The nationwide Birth Defects Monitoring Program (BDMP) provided rates from 1970 to 1993. MACDP data are population-based and could be categorized by the severity of the hypospadias. BDMP data allowed analysis of rate trends for the four census regions of the United States.
Results. Data from both surveillance systems showed an approximate doubling of hypospadias rates in the 1970s and 1980s. MACDP data showed that the rate of severe cases increased while the ratio of mild to severe cases decreased. BDMP data showed that hypospadias rates increased markedly in all four regions of the United States.
Conclusions. The observed increases are unlikely to be attributable to increased sensitivity of the surveillance systems or the identification of more mild cases by physicians over time, because either trend would have increased rather than decreased the ratio of mild to severe cases. If real, these trends represent the largest number of cases and the first report of an increase in hypospadias rates outside of Europe. Additional investigation of a possible increase in hypospadias rates is warranted.
Embryologic formation of the penile urethra takes place between 9 and 12 weeks after ovulation and depends on the secretion of testosterone by the fetal testes.1 Incomplete fusion of the urethral folds results in a urethral opening on the ventral surface of the penis or on the scrotum, a condition known as hypospadias.
Hypospadias is one of the most common congenital anomalies, in spite of the fact that it occurs almost exclusively in males. The cause of hypospadias is unknown, although epidemiologic studies have identified some associated factors. Rates are highest among whites, lowest among Hispanics, and intermediate among African-Americans.2 Some studies have found associations between hypospadias and maternal factors such as primiparity and advanced age.3,4 Familial aggregation with a likely multifactorial mode of inheritance has also been noted.5 Clinical studies have shown that some affected boys have defects in testosterone metabolism or testosterone receptors,6,7 suggesting that hypospadias can result from an abnormality in endocrine factors that influence the development of the male genitalia.
In the 1970s and 1980s, congenital anomaly surveillance systems reported upward trends in the birth prevalence of hypospadias in Norway,8 Sweden,9 Denmark,4England,10 and Hungary.11 Some of the authors attempted to link these trends to fetal exposure to progestins that their mothers had taken as contraceptives, as part of a pregnancy test, or for maintenance of pregnancy. Several studies during the 1970s indicated an association between such exposures and hypospadias.12-14 In addition, a study of fetal rats showed that exposure to synthetic progestins could produce hypospadias in male offspring.15 However, given the relative risk associated with such exposure and the prevalence of progestin use, women's increased use of progestins could account for only a fraction of the increase observed in hypospadias rates.10 Moreover, a recent metaanalysis of human studies found no association between such exposure and external genital malformations.16 To date, the increases remain unexplained. The most recent hypothesis suggests that widespread pollution of the environment by synthetic compounds with estrogenic or antiandrogenic activity is to blame.17,18
To determine whether an increase in the birth prevalence of hypospadias has occurred in the United States, we examined data from two well-established surveillance systems, the Metropolitan Atlanta Congenital Defects Program (MACDP) and the Birth Defects Monitoring Program (BDMP).
The MACDP and the BDMP have been described in detail elsewhere.19 The MACDP is a population-based registry that began in 1968 and uses active case ascertainment in 22 hospitals and clinics in the Atlanta, GA, area. The BDMP began in 1970 and relies on diagnoses recorded on newborn discharge summaries from a sample of hospitals nationwide. Neither system depends on the filing of reports of defects by health care providers.
Both systems use coding schemes based on the ninth revision of theInternational Classification of Diseases, Clinical Modification (ICD9-CM). The MACDP, however, bases its nosology on the 1979 British Pediatric Association modification of the ICD9-CM. MACDP subdivides the ICD9-CM category for hypospadias (752.6) by using fifth and sixth digits. In the MACDP coding scheme, 752.60 and 752.62 defined hypospadias. Epispadias, 752.61, was excluded because its embryologic origin differs from that of hypospadias. For the BDMP, only four-digit ICD9-CM codes were available; thus, it was not possible to exclude epispadias. Epispadias is, however, about 50 times less common than hypospadias.
The MACDP uses a sixth digit to subdivide hypospadias into first, second, and third degrees of severity. In first-degree cases, the urethral opening is on the ventral surface of the glans penis; in second-degree cases, the opening is on the ventral surface of the shaft; and in third-degree cases, the opening is scrotal or perineal.20 Second- and third-degree cases are often accompanied by a ventral bowing of the penis known as chordee. Second- and third-degree cases also are associated with difficulties in forming a normal urine stream and in normal reproductive function. Together, second- and third-degree cases are classified as severe hypospadias in this analysis. Unclassified cases are included in the not-otherwise-specified (NOS) category.
The MACDP calculated rates using as denominators the total number of live births to mothers resident in the metropolitan Atlanta area, an average of 29 700 per year from 1968 to 1993. The BDMP calculated rates using as denominators the total number of newborns discharged from hospitals in the sample. An annual average of 763 000 births occurred in BDMP hospitals between 1970 and 1993. BDMP data can be broken into the four regions of the United States: West, Central, Southeast, and Northeast, as defined by the US Census Bureau. Although calculating hypospadias rates per 10 000 male births might be more appropriate, we chose to calculate rates per 10 000 total births to facilitate comparison with earlier published rates, almost all of which have used total births.
The annual percent of increase in rates was calculated from the slope of a regression line drawn through the observed values. The significance of the trends in rates was tested using the Mantel–Haenszel extension test.21
A total of 2 372 hypospadias cases occurred in metropolitan Atlanta between 1968 and 1993. Overall, 59.9% of the cases fell in the NOS category, 25.7% fell in the first-degree category, and 8.8% fell in the second- or third-degree category. Chordee alone was diagnosed in 5.6% of cases. The ratio of first-degree to second- or third-degree cases was 4.2 from 1968 to 1982 and 2.6 from 1983 to 1993. The percentage of NOS cases declined from 80% of the total in the early 1970s to ∼50% during 1990 to 1993.
The total hypospadias rate nearly doubled in the MACDP between 1968 and 1993 (Fig 1). The trend was statistically significant (P < 10−6). A regression line drawn through the observed rates indicated that the annual rate of increase was 2.9%. The overall increase occurred at a rate of 1.4% per year among whites and 5.7% per year among nonwhites. Inspection of rates by hospital indicated that the increase could not be attributed to a single hospital or group of hospitals in the Atlanta area. The rate of severe hypospadias increased three- to fivefold, from 1.1 per 10 000 births in 1968 to between 2.7 and 5.5 per 10 000 births per year from 1990 to 1993. This trend was also statistically significant (P < 10−6).
A total of 49 162 cases of hypospadias were recorded by the BDMP between 1970 and 1993. As shown in Fig 2, hypospadias rates measured by the BDMP nearly doubled from 1970 to 1993 (from 20.2 to 39.7 per 10 000). The trend was statistically significant (P < 10−6). All four regions of the United States showed increases. The increase was greatest in the Southeast region and least in the West region (Fig3).
Hypospadias rates measured by both independent surveillance systems have increased. This is the first report of an increase in the birth prevalence of hypospadias outside of Europe. It is also the first published report of an increase in the rate of severe hypospadias. During the 1985 to 1988 period, MACDP and BDMP rates for hypospadias exceeded those reported by all 25 surveillance systems participating in the International Clearinghouse for Birth Defects Monitoring Systems.22
In the mid-1980s, European registries reported upward trends in hypospadias. The hypospadias rate in England and Wales increased from 7.5 to 18.3 per 10 000 births between 1964 and 1983,10 the rate in Hungary increased from 12 to 24 per 10 000 births between 1971 and 1983,11 and the rate in Sweden increased from 8 to 12 per 10 000 births from 1965 to 1979.9 However, almost all of the Swedish increase occurred from 1969 to 1973, and rates were stable thereafter. The author of the Swedish study noted no change in the criteria used in reporting this defect to the Swedish registry during this period. A similar increase occurred in Norway between 1967 and 1973 and in Denmark between 1970 and 1981.4 More recent data are available in the annual reports of the International Clearinghouse for Birth Defects Monitoring Systems.23 These reports indicate that the upward trends have continued through 1993 in Denmark and Norway but have leveled off in England, Sweden, and Hungary. The only previous longitudinal study of hypospadias in the United States used Mayo Clinic birth and pediatric records for all children born in Rochester, Minnesota, from 1940 to 1970.24It noted an exceptionally high rate of 41 per 10 000 births, probably as a result of access to the findings of multiple examinations through childhood. The Mayo Clinic rates did not vary significantly over time.
The doubling of rates observed here is unlikely to be attributable to increased sensitivity of the surveillance systems. A system would have to initially record fewer than half of all diagnosed defects for a doubling of rates to be attributable to sensitivity improvement alone. Such a low sensitivity is unlikely. Furthermore, no major changes in the methods of either the MACDP or the BDMP occurred during the 1970s or 1980s.
Although to our knowledge, no change in diagnostic practices for hypospadias has occurred, an alternative explanation is that the doubling of rates is attributable to the increasing identification of numbers of mild cases by physicians. If this had occurred, however, one would expect an increasing proportion of mild cases in MACDP data. Instead, the ratio of first-degree to severe cases in the MACDP data decreased as the rates rose. However, the shrinking proportion of NOS cases makes these changes difficult to interpret. Better identification of mild cases by physicians, therefore, cannot be ruled out as at least a partial cause of the increase in the hypospadias rate.
Although these results are not sufficient in themselves to prove that the rate of hypospadias is increasing, the possibility of such an increase merits additional investigation, given that this study involved more cases, more recent data, and a longer period than any of the reports from Europe. And, if real, the increased rate means that large numbers of newborn males are being affected, some of them severely. If the increased hypospadias rate can be confirmed, more study of risk factors for this anomaly will be needed, with attention to those risk factors that may have become more prevalent during the period of rate increase.
We thank the Metropolitan Atlanta Congenital Defects Program abstracters Jo Anne Croghan, Joann Donaldson, Joan Garcia, Debbie Nurmi, Charlie Peters, and Connie Thompson. Their conscientious and skilled data collection efforts provide the foundation on which the MACDP is built.
- Received November 7, 1996.
- Accepted March 13, 1997.
Reprint requests to (L.J.P.) Birth Defects and Genetic Diseases Branch, Division of Birth Defects and Developmental Disabilities, National Center for Environmental Health, Centers for Disease Control and Prevention, Mailstop F-45, 4770 Buford Hwy, NE, Atlanta, GA 30341-3724.
- MACDP =
- Metropolitan Atlanta Congenital Defects Program •
- BDMP =
- Birth Defects Monitoring Program •
- ICD9-CM =
- International Classification of Diseases, 9th ed, Clinical Modification •
- NOS =
- not otherwise specified
- ↵O'Rahilly R, Muller F. Human Embryology and Teratology. New York, NY: John Wiley and Sons; 1992
- Angerpointer TA
- Bjerkedal T,
- Bakketeig LS
- Källén B,
- Winberg J
- Aarskog D
- Czeizel A,
- Toth J,
- Erodi E
- ↵Giwercman A, Carlsen E, Keiding N, Skakkebaek NE. Evidence for increasing incidence of abnormalities of the human testis: a review. Environ Health Perspect Suppl, 1993;101(suppl 2):65–71
- Edmonds LD,
- Layde PM,
- James LM,
- Flynt JW,
- Erickson JD,
- Oakley GP
- ↵Anonymous. Congenital Malformations Worldwide: A Report From The International Clearinghouse for Birth Defects Monitoring Systems. Amsterdam, The Netherlands: Elsevier Science Publishers; 1991
- ↵Anonymous. International Clearinghouse for Birth Defects Monitoring Systems Annual Reports, 1980–1993. Rome, Italy: International Centre for Birth Defects; 1982–1995
- Copyright © 1997 American Academy of Pediatrics