How much of the decline in sperm counts can be explained by relaxed reproductive selection?

Abstract

To the Editor: Czeizel and Rothman 1 have proposed a hypothesis to explain the possible decline in sperm concentration observed over time in some countries, 2,3 and the simultaneous increase in disorders of male reproductive health. 4 We were interested in their hypothesis based on a relaxed reproductive selection of men with poor reproductive health, attributable to a decreasing fertility of very fecund men, but we had some concern about the data given to support it. We will discuss these concerns, referring to fecundity as the biological ability to conceive and to fertility as the number of children per couple. The authors reported that the fertility of fertile couples (that is, those who conceived a child) in Hungary declined from 11 children per couple at the end of the 19th century down to 1.5 by 1990, 1 and that similar changes had occurred in other countries, quoting a reference regarding fertility trends in France. 5 This reference, however, indicates that the evolution of fertility in France is neither pronounced nor monotonous across the 20th century. The fertility of fertile couples in France was 3.1 in 1900, 2.3 in 1990, 6,7 and in the 1950s and 1960s France (as countries like England, Denmark, and the United States) experienced a baby boom. 8 A decreasing time trend in sperm concentration has, indeed, been reported in France. 9 We conducted a simple simulation study to evaluate to what extent the temporal trends in fertility in France could have induced an increase in the proportion of men with a low sperm concentration. We considered that 10% of the French men born in 1870 had a low sperm count (below 20 million per ml), and 90% had a high sperm count (above 20 million per ml). Based on this, we estimated by simulation the proportion of their sons who had low sperm counts, over three generations. The number of boys born from each generation come from national censuses. 6,7 We supposed that (a) whatever the period, the fertility of fertile men with low sperm count was that observed for all fertile men born in 1930, whereas (b) the proportion of men with a low sperm count remaining childless did vary; more precisely, the proportion of childless men among those with low sperm count born in 1870 was 1.5 times that observed for men with high sperm count, and this proportion had the value observed for men with high sperm count thereafter. The first supposition was motivated by the assertion made by Czeizel and Rothman that the number of offspring from couples with a high and low fecundity has converged in the recent period. 1 Also, a man with a high sperm count had an 8% risk that his son would have a low sperm count, this risk being eight times as high for fathers with low sperm count—a high value likely to favor the relaxed reproductive selection hypothesis (Table 1). Under these assumptions, the proportion of men with low sperm count varied from 12.9% among the sons whose fathers were born in 1870 up to 17.0% among the sons whose fathers were born in 1930 (Table 2). 10 These values would have been observed in cross-sectional studies assessing sperm concentration when those sons were adults, in about 1930 and 1990. Supposing that the mean sperm concentration was 8 million per ml and 100 million per ml among men with sperm count below and above 20 million per ml, respectively, this would correspond to a decrease from 88 to 84 million per ml, a much smoother evolution than the decrease from 113 to 66 million per ml reported between 1940 and 1990 in the industrialized countries. 2TABLE 1: Distribution of the Number of Children of French Men Born in 1870, 1900, and 1930, According to Their Semen Quality. Data for the Whole Population Come from National Censuses 6,7; the Other Values Were Estimated as Explained in the TextTABLE 2: Number of Sons with a High or a Low Sperm Count According to the Sperm Count of Their Fathers, in the Offspring of Three Generations of French Men, Based on a Simulation on 100 Men in Each GenerationWe had to make several assumptions to define our model, and a given fertility might be reached with varying contrasts in fertility between men with high and low sperm concentration. There is still evidence that sperm concentration is a real but limited determinant of fecundity. 11,12 We therefore think that, in the case of France, the hypothesis of a relaxed reproductive selection is unlikely to induce an increase in the proportion of men with low sperm counts of the same magnitude as reported in the literature, and that other major mechanisms are likely to be involved. Rémy Slama Henri Leridon