NIH News Release
NATIONAL INSTITUTES OF HEALTH
National Institute of Environmental Health Sciences

EMBARGOED FOR RELEASE
Monday, November 24, 1997
7:00 AM Eastern Time
John Peterson
(919) 541-7860

Reanalysis of International Data Finds Sharp Decline in Sperm Density

After an extensive review of data from 61 published studies, three California researchers have concluded that a decline in average sperm density reported in the U.S. and other Western countries may be even greater than previously estimated.

Their analysis of data collected from 1938 to 1990 indicates that sperm densities in the United States have exhibited an average annual decrease of 1.5 million sperm per milliliter of collected sample, or about 1.5 percent per year, while those in European countries have declined at about twice that rate (3.1 percent per year).

The study was conducted by epidemiologists Shanna Swan, Eric Elkin and Laura Fenster of the California Department of Health Services. It appears in the November issue of Environmental Health Perspectives, the monthly scientific journal published by the National Institute of Environmental Health Sciences.

Since the early 1930's, there has been considerable interest in declining semen quality as a key predictor of male reproductive dysfunction. The vast majority of studies designed to answer this question have focused on sperm density - the number of sperm contained in one milliliter of sample. (One milliliter is approximately 1/30th of an ounce.)

Despite enormous differences in data collection methods, study population and time period, most studies have come to the same conclusion - that sperm density has declined. In fact, a 1992 review of 61 such studies (Evidence for decreasing quality of semen during past 50 years, E. Carlsen, A. Giwercman, and N. Skakkebaek, British Medical Journal, vol. 305, page 609) revealed a steady decline, from 113 million sperm per milliliter in 1938 to 66 million in 1990, or about 1 million sperm per milliliter per year.

However, these studies did not take into account such factors as the age of the subjects, the length of abstinence prior to sample collection, and method of sample collection, each of which can influence the observed trend. Swan said, "Most of the critics have suggested ways in which the data analysis might be skewed, but no one has ever looked at the data from these earlier studies to see whether these hypothetical biases are actually present."

Using a statistical model that corrects for individual differences in these key variables as well as geographic area, Swan and her colleagues reanalyzed the data from 56 of the studies cited in the 1992 paper. The investigators excluded three non-English language studies and two others that included men who had conceived only after an infertility workup.

While the results of their analyses also showed a significant decline in sperm density, it was the rate of the decline, particularly in Western countries, that was most surprising. "We observed a decrease of about 1.5 million sperm per milliliter per year in the United States, and a corresponding decrease of about 3 million sperm per year in Europe," reports Swan.

For non-Western countries, a group that included Brazil, India, Israel, Hong Kong, Kuwait, Nigeria and Thailand, the trend was slightly positive. However, because these data were taken from only 13 studies, all of which were published after 1978, this trend was not statistically significant.

Since no mathematical model can ever fit the data perfectly, there is always a certain discrepancy between what the model predicts and what actually happens - this is often referred to as variability. "A perfect model would explain 100 percent of this variability," says Swan. "Our model accounts for 80 percent - that is the best fit of any model that has been proposed."

Although the authors do not address the specific causes of this phenomenon, some recent studies have focused on the relationship between environmental exposures and declining sperm quality. In one such study, researchers reported a significant correlation between lowered sperm densities and increased levels of organochlorine compounds in the subjects' seminal fluid. In another, investigators found that a general decline in sperm concentration during the years 1949 to 1981 was statistically linked to an overall increase in several environmental exposures.

While there is no evidence that this apparent decline in sperm density has led to reduced fertility, the authors say sperm count may be a surrogate indicator of effects on the male reproductive system. They say, for example, that in countries such as Denmark, England and the United States, where sperm counts have fallen, the incidence of testicular cancer has increased dramatically over the last 25 years, while in Finland, where sperm counts are still relatively high, testicular cancer rates have remained low.

Researcher Shanna H. Swan, Ph.D., can be reached at (510) 450-3818.