The shock conclusion of a study 20 years ago indicated that sperm counts had halved. But a closer look at the evidence then and now paints a much more complex story.
Ask someone what they know about men’s sperm counts, and the chances are you will hear that they are dropping. People will probably have plenty of theories as to why this is happening – perhaps because of herbicides, pesticides, or oestrogens in the water from so many women being on the pill. Or there are theories based on lifestyle – for instance, men are taking less exercise, eating more fast food and getting fatter. Some even suggest that wearing tight pants can cause problems.
But is it really true that sperm counts are in free-fall globally? To understand why this perception is so common, we have to go back to a paper published 20 years ago.
Elisabeth Carlsen and her colleagues reviewed 61 studies of semen quality carried out between 1938 and 1990, and their conclusionspublished in the British Medical Journal in 1992 were shocking. In 50 years sperm counts had halved. The authors were open in saying that the data did not indicate whether or not the decline was continuing. But the mass of publicity their findings received left many with the idea that it had been proven beyond doubt – sperm counts are not only falling everywhere, but will continue to do so.
However, there are a number of issues with drawing conclusions about trends from these 61 original studies, and these are worth going through in some detail. Starting with the participants, studies of the sperm count of the average man are supposed to preclude men with fertility problems. But getting volunteers to provide semen samples is not easy, so some subjects might enrol in studies because they are concerned about their sperm count, yet fail to divulge this to the researchers. Different studies use different methods of ensuring that fertility is proven, but if some volunteers suspect they might have low sperm counts this could skew the sample, if you’ll forgive the pun.
There are also other factors that vary from study to study, such as the length of time there has been between the provision of the tested sample and the previous ejaculation. Also, methods of analysing semen samples have changed between 1938 and 1990. The World Health Organisation (WHO) recommends using laboratory techniques that do not rely on an individual making a judgement, but this wasn’t the case for the earlier studies, making it hard to compare them with measurements taken 50 years later. And it is not just the practical techniques that have changed; statistical methods of analysing data have evolved too.
When you look at the set of studies reviewed in the paper in more detail, it seems that although sperm counts do appear to have dropped in some places, they might have risen in others, even in different regions of the same country. In Paris, sperm counts appear to have declined, while they remained stable in another French city, Toulouse.
Many of the original studies were too small to be considered as representative of a population. A re-analysis of the 20 largest studiesfound that the majority had been conducted before 1970 in New York, where sperm counts seemed particularly high. The later studies were from different places, which meant that sperm counts in one geographical location were compared with counts from somewhere different over a different period of time. Once these studies were removed, the picture looked a little different and sperm quality didn’t seem to be in decline. So perhaps the 1992 review was in fact highlighting differences between different regions, rather than differences over time.
Picking apart the studies in this review paper is one thing, but what does the research post-1992 tell us? The picture is mixed, with some studies still showing evidence for a decline in some places and others showing sperm counts have remained stable. To take two examples, a Finnish study published last year found that men born towards the end of the 1980s had lower sperm counts on average than those born at the beginning of the same decade. But Danish research that has been taking semen samples from young men about to begin military service has found no decline in sperm quality in 5,000 volunteers so far.
Geographical location seems to make a difference, but the lack of studies conducted in low-income countries makes it difficult to establish a full picture. A paper published this year highlights the fact that even studies from middle-income countries have rarely included the poorest people.
So, twenty years on from the BMJ paper we cannot say we have a complete picture about sperm counts globally. We can say that sperm counts are declining in some places, and this needs to taken seriously. But from the best evidence it seems this crisis is not happening everywhere. And without new and carefully controlled studies, we are likely to spend many more years in the dark over this issue. It is only by looking carefully at where counts are and are not falling, using the most accurate methods available, that we might find any clue to the cause.