Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Effects of Castration on the Life Expectancy of Contemporary Men

16 Fluttershy 08 August 2015 04:37AM

Follow-up to: Lifestyle Interventions to Increase Longevity


A recent review article by David Gems discusses possible mechanisms by which testosterone and dihydrotestosterone could shorten the life expectancies of human males, and examines previous research on the effects of castration on male survival. However, Gems does not examine how age at castration affects how much castration extends one's life by, which this post does. In general, castration after puberty in males prolongs life to a lesser extent than castration before the onset of puberty.

Additionally, Gems' review does not estimate how long modern-day eunuchs might live relative to intact human males. Two of the other three known studies on the effects of castration on human life expectancies found that, historically, castration prolonged life by more than a decade in the median case. However, some of the life expectancy gains from castration are due to the increased ability of eunuchs to fight off infections. The fact that fewer men die from infections in the 21st century than was the case in previous centuries means that modern-day eunuchs gain fewer years of life from castration than eunuchs gained from castration in the past. As seen from comparing Figure 3b and Figure 4, eunuchs castrated just before the onset of puberty extended their (mean) life expectancies by 11 years in Hamilton & Mestler's study, though modern eunuchs castrated at similar ages might expect to extend their life expectancies by 7 years. 


A few relevant studies, such as the study of institutionalized eunuchs by Hamilton & Mestler, the study of Korean eunuchs by Min, Lee, and Park, and the review article by Gems are particularly worth reading or skimming for those interested in this topic. The excel file showing the work behind this post is also available. These documents are supplementary; reading them is not a  prerequisite for reading this post.

This post will examine the proposition that castration of human males (specifically, orchiectomy, the surgical removal of both testicles, but not the penis) either before or after the onset of puberty will extend both their life expectancy, and their lifespan. In light of antagonistic pleiotropy, it a priori makes sense that castration might extend one's life expectancy.

A number of papers have mentioned that the effects of castration on the life expectancy of different types of nonhuman animals don't provide a good model for the effects of castration on the life expectancy of human males. Specifically, "the relationship of gonadal functions to survival seems to involve many variables... individuals, strains, and species may vary in their response to gonadectomy". This leaves only a small number of studies that have much bearing on the question of whether or not orchiectomy extends human life expectancy. While there are some studies on the health effects of chemical and physical castration of (often elderly) modern men with prostate cancer, it seems like having prostate cancer would correlate with having other pathologies. Further, as will be examined later, it seems that orchiectomies performed at early ages have many positive effects on health, whereas orchiectomies performed later in life have fewer positive effects, and may even negatively affect some aspects of health.

After setting aside animal studies and studies of men with prostate cancer, only four papers directly relevant to whether orchiectomy increases the life expectancy of men remain. This is worth stating explicitly, since citing only a fraction of the available research on a given topic can be a fallacy. First, the study by Min, Lee, and Park found that, historically, Korean eunuchs lived 14-19 years longer than intact males from similar social classes in the median case. Secondly, Hamilton and Mestler's study of the effects of orchiectomy on the life expectancies of mentally retarded individuals found that males castrated before puberty lived about 13 years longer than intact men in the median case, and that males castrated after puberty experienced smaller lifespan gains. Thirdly, a letter to Nature by Nieschlag et. al which compared the lifespans of famous castrato singers to the lifespans of other singers from the same era found that orchiectomized male singers lived about as long as intact male singers. Lastly, page four of the review article by David Gems examined all three of these studies, and, after finding methodological issues with the letter to Nature, concluded that the results in these papers were "consistent with the idea that testes are a determinant of the gender gap in human lifespan".

Evidence Regarding Whether or Not Orchiectomy After Puberty Increases Life Expectancy

The only study that examined the effect which age at orchiectomy had on the life expectancy gains from castration in humans was Hamilton and Mestler's study of mentally retarded, institutionalized individuals. Note that the participants in Hamilton & Mestler's study lived shorter lives than non-institutionalized Americans of the same era lived, which could likely be explained by the mentally retarded status of the participants, and the plausibly poor conditions under which participants might have lived. In Figure 4 and Table 10 from Hamilton and Mestler's paper, it is shown that males castrated between 15 and 40 years of age live longer than intact males, but that within this range, earlier castrations added more years to the life expectancy of eunuchs than later castrations did.

It is worth reproducing Figure 4 from Hamilton & Mestler's article, which shows the survival curves (starting at 40 years of age) of intact males and males castrated at various ages:

One thing about this figure that stands out is that the portion of the survival curve for institutionalized non-castrates shown in this figure is nearly linear. In the present day, intellectually disabled populations have survival curves which look quite different from the one for non-castrates shown in the figure above. For reference, the survival curve for castrated females in Figure 5 of this post has a shape which is comparable to the shape of survival curves for modern first-world populations. It is also remarkable that the tail end of the survival curve for non-castrates in the above figure is fatter than the tails of the survival curves for men castrated after 14 years of age-- it isn't obvious whether or not this difference reflects a real phenomenon. Further, the 3.7 % centenarian rate for Korean eunuchs in the study by Min, Lee, and Park suggests that eunuchs should have a longer (maximum) lifespan than non-castrates, which isn't borne out in Figure 4 from Hamilton & Mestler. This having been said, Figure 4 and Table 10 from Hamilton & Mestler's study show that castration at earlier ages prolongs life more than castration at later ages does.

Below, in Figure 1, the attempted linear fit between median life expectancy versus age at castration given on p. 403 of Hamilton and Mestler's paper is shown. The authors used data from Table 10 of their paper to determine this fit, but did not graph the data or determine an R2 value for this linear fit. The estimated median life expectancy of the non-castrates was 64.7 years-- a reasonable value, given their status as institutionalized mentally retarded men in the early 20th century. Thus, Figure 1 can be used to visualize the fact that even men who were castrated at 30-39 years of age lived longer than non-castrates in the study (p = 0.002). Since the data shown in Figure 1 did not follow a linear trend, additional fits were tried below.

Figure 1. Hamilton & Mestler's Regression of Median Life Expectancy v. Age at Castration

Figure 2. Polynomial Fit for Median Life Expectancy v. Age at Castration

Figure 3. Raw Data and Fits For Interpolation of Mean Life Expectancy v. Age at Castration

Data and fits for the median and mean life expectancies of eunuchs are given in Figures 2 and 3, respectively. The data plotted in sections a and b of Figure 3 could not be reasonably fitted to a curve directly, so sections c and d of Figure 3 show the same data as sections a and b, but plotted on an inverted x axis and successfully fitted to a curve. The polynomial data fits given in all Figures are only intended for use in interpolation.

Effects of Orchiectomy on Mortality from Infectious Diseases and Cardiovascular Mortality

Literature has suggested that castration in human males may promote longer lifespans and higher life expectancies by protecting against infections and cardiovascular events. Much of the evidence for the proposition that castration protects against cardiovascular disease (CVD) comes from basic biology rather than from studies of eunuchs, since Hamilton & Mestler's paper is the only study on eunuchs which attempted to collect data on causes of death in castrated men, and only did so from clinical diagnoses of the primary causes of deaths of eunuchs and intact men between 1940-1964. Still, modern men die of cardiovascular events more often than modern women do, so investigating whether or not castration protects against cardiovascular events is worthwhile.

The authors of the study on Korean eunuchs cite this review as evidence that "male sex hormones reduce the lifespan of men because of their antagonistic role in immune function". Gems' review article also suggests that male sex hormones may act as an immune suppressant. Moreover, in Hamilton & Mestler's study, 27% of eunuchs died of infections, compared to 44% of intact men (p = 0.02), and the mean age of eunuchs dying of infections was 44, compared to 35 for intact men (p = 0.03). However, Table 14 of Hamilton & Mestler's study suggests that castration protects more against deaths from certain kinds of infections, such as tuberculosis, than others. In general, it seems like the claim that castration protects against deaths from infections is true.

On the other hand, the data relevant to whether or not eunuchs die more from CVD than intact men do is muddled at best, and it isn't obvious that castration protects males from CVD by much, if at all. One mostly irrelevant data point is men who have undergone chemical or physical castration after being diagnosed with prostate cancer, as well as hypogonadic men in general; many meta-analyses on the relationship between hypogonadism and frequency of adverse cardiovascular events (and on the effects of hormone replacement therapy on the frequency of adverse cardiovascular events) in men have been done. Men castrated after being diagnosed with prostate cancer tend to have more adverse cardiovascular events than other similarly aged men, but this could be because hypogonadism correlates with being unhealthy, rather than because castration at advanced ages decreases life expectancy.

One poorly done study on Danish eunuchs who were predominantly drawn from the lower class found that these eunuchs did not live as long as men in Denmark did on average, and also found that the standardized mortality ratio for cardiovascular disease-related deaths was higher than the all-cause standardized mortality ratio in eunuchs. However, men in this study were often castrated later in life-- all but one man were castrated after the age of 18, and the average age at castration was 35. As suggested by Figure 2 and Figure 3 above, this means that most of the Danish eunuchs gained appreciably fewer years of life from being castrated than they would have gained if the castrations had been carried out much earlier in their lives. These concerns suggest that this study should not change one's credence in the proposition that castration protects against CVD mortality by much.

Lastly, Hamilton & Mestler's study found that eunuchs dying of cardiovascular disease during or after 1940 lived an average of 51.6 years, while intact males dying of that cause lived an average of 51.1 years. This difference was not found to be significant. However, since not all eunuchs included in the study had died by the time of publication, it is still possible that castration early in life protects against late-life cardiovascular mortality, but not early and mid-life cardiovascular mortality.

Effects of Orchiectomy on Modern Lifespans and Life Expectancy

Some common causes of death in both Hamilton & Mestler's study and the study of Korean eunuchs, such as tuberculosis, are no longer common causes of death. Thus, data from Table 14 in Hamilton & Mestler were used alongside modern actuarial data to crudely predict how long eunuchs castrated in the 21st century might live. The details of the analysis are given in this excel file. The results of this analysis are given below.

Figure 4. Life Expectancy Gains for Modern Eunuchs


Table 1. Life Expectancy Gains for Modern Eunuchs

For the most part, the data in Figure 4 and Table 1 are consistent with my holistic understanding of the effects of castration in men. It is hard to say how castration after age 35 would affect life expectancy, as very few eunuchs in Hamilton & Mestler's study were castrated after 35. It's also a shame that about 27% of the eunuchs and intact males who died during 1940-1964 were not listed as having a primary cause of death-- this may have led to an overestimation of the extent to which castration is expected to extend modern eunuch's life expectancies. On the other hand, Min, Lee, and Park found that 3.7% of Korean eunuchs who died between the late 14th to early 20th century were centenarians, "a rate at least 130 times higher than that of present-day developed countries", which suggests that modern eunuchs would likely benefit from increased lifespans.

Effects of Orchiectomy on Health and Physiology

Gems' review article, this article on historical eunuchs, the wikipedia page on castration, and Hamilton & Mestler's study all note certain effects that castration can have on human males.

All castrated males have an increased risk of developing sarcopenia, and becoming overweight. Wilson and Roehrborn note that eunuchs have historically suffered from skeletal problems such as osteoporosis and kyphosis; this is especially true of elderly eunuchs, and eunuchs castrated at earlier ages. Hormone replacement therapy can prevent or deter sarcopenia, osteoporosis and kyphosis. Castration also decreases sex drive, prevents baldness if done early enough in life, and may result in enlarged pituitaries and enlarged breasts. Castration causes the prostate to shrink over time, and if done early enough in life, effectively prevents the development of prostate cancer. Castration also prevents the development of prostatic hyperplasia and testicular cancer.

Men castrated before puberty will develop higher voices, little or no sex drive, and smaller penises.

Effects of Gonadectomy on Human Females

There is very little data relevant to whether or not oophorectomy (castration) of women extends life expectancy or lifespan. Hamilton & Mestler have a small section dedicated to estimating the life expectancy of castrated females based on only 11 female castrates of known fate. They also find the median lifespans of castrated and intact females known to be dead by the end of the study to be equal. Lastly, Hamilton & Mestler find the mean lifespan of institutionalized castrated females known to be dead by the end of the study, 56.2 years, to be significantly greater than the mean lifespan of institutionalized intact females known to be dead by the end of the study, 33.9 years (p < 0.001). The estimated survival curves for all castrated females and all intact females-- not just those known to be dead by the end of the study-- are given in Figure 5.

Figure 5. Survival Curve for Intact and Castrated MR Females

Conclusion and Motivation

Orchiectomy should prolong the lifespans of modern males, especially if done before puberty. While the estimates of life expectancy gains from castration given in Figure 4 and Table 1 aren't perfect, they are my best guesses, and should be interpreted with the correspondingly appropriate level of credence.

My original motivation for writing this post was that I was interested in learning about the different ways in which humans could extend their lifespans and life expectancies. So, while being castrated is one way for males to live longer, quitting smoking and improving one's diet and exercise regimen are better uses of time and energy for people who are just beginning to think about changing their lifestyles in order to live longer.

Thanks to Vaniver, who caught several errors in an earlier draft of this post, and thanks to btrettel for pointing me to a few papers early on. All remaining errors in this post are solely my own.


1. Castration. http://en.wikipedia.org/wiki/Castration

2. Antagonistic Peliotrophy Hypothesis. http://en.wikipedia.org/wiki/Antagonistic_pleiotropy_hypothesis

3. Bittles, A. H.; Petterson, B. A.; Sullivan, S. G.; Hussain, R.; Glasson, E. J.; Montgomery, P. D. The influence of intellectual disability on life expectancy. J. Gerontol. A Biol. Sci. Med. Sci. 2002, 57, M470-2.

4. Corona, G.; Maseroli, E.; Rastrelli, G.; Isidori, A. M.; Sforza, A.; Mannucci, E.; Maggi, M. Cardiovascular risk associated with testosterone-boosting medications: a systematic review and meta-analysis. Expert opinion on drug safety 2014, 13, 1327-1351.

5. Corona, G.; Rastrelli, G.; Monami, M.; Guay, A.; Buvat, J.; Sforza, A.; Forti, G.; Mannucci, E.; Maggi, M. Hypogonadism as a risk factor for cardiovascular mortality in men: a meta-analytic study. Eur. J. Endocrinol. 2011, 165, 687-701.

6. Gems, D. Evolution of sexually dimorphic longevity in humans. Aging (Albany NY) 2014, 6, 84-91.

7. Hamilton, J. In Duration of Life in Lewis Strain of Rats After Gonadectomy at Birth and at Older Ages; Reproduction & Aging; 1974; pp 116-122.

8. HAMILTON, J. B. Relationship of Castration, Spaying, and Sex to Survival and Duration of Life in Domestic Cats. J. Gerontol. 1965, 20, 96-104.

9. Hamilton, J. B.; Mestler, G. E. Mortality and survival: comparison of eunuchs with intact men and women in a mentally retarded population. J. Gerontol. 1969, 24, 395-411.

10. Jones, C. M.; Boelaert, K. The Endocrinology of Ageing: A Mini-Review. Gerontology 2015, 61, 291-300.

11. Mestler, H. In The Role of Testicular Secretions as Indicated by the Effects of Castration in Man and Studies of Pathological Conditions and the Short Lifespan Associated with Maleness; Pincus, G., Ed.; Recent Progress in Hormone Research; Laurentian Hormone Conference: 1948; pp 257.

12. Min, K.; Lee, C.; Park, H. The lifespan of Korean eunuchs. Current Biology 2012, 22, R792-R793.

13. Nieschlag, E.; Nieschlag, S.; Behre, H. M. Lifespan and testosterone. Nature 1993, 366, 215-215.

14. Roberts, M. L.; Buchanan, K. L.; Evans, M. Testing the immunocompetence handicap hypothesis: a review of the evidence. Anim. Behav. 2004, 68, 227-239.

15. Talbert, G. B.; Hamilton, J. B. Duration of life in Lewis strain of rats after gonadectomy at birth and at older ages. Reproduction & Aging 1974, 116.

16. Wilson, J. D.; Roehrborn, C. Long-term consequences of castration in men: lessons from the Skoptzy and the eunuchs of the Chinese and Ottoman courts. The Journal of Clinical Endocrinology & Metabolism 1999, 84, 4324-4331.

Lifestyle interventions to increase longevity

120 RomeoStevens 28 February 2014 06:28AM

There is a lot of bad science and controversy in the realm of how to have a healthy lifestyle. Every week we are bombarded with new studies conflicting older studies telling us X is good or Y is bad. Eventually we reach our psychological limit, throw up our hands, and give up. I used to do this a lot. I knew exercise was good, I knew flossing was good, and I wanted to eat better. But I never acted on any of that knowledge. I would feel guilty when I thought about this stuff and go back to what I was doing. Unsurprisingly, this didn't really cause me to make any positive lifestyle changes.

Instead of vaguely guilt-tripping you with potentially unreliable science news, this post aims to provide an overview of lifestyle interventions that have very strong evidence behind them and concrete ways to implement them.

continue reading »

View more: Next