🧬 Aging Wiki

Mortality and survival: comparison of eunuchs with intact men and women in a mentally retarded population

6 min read

⚠️ Permanently unverifiable against primary PDF — this paper (doi:10.1093/geronj/24.4.395) is closed-access with no PMC deposit and no OA version available (not_oa per archive; no PMC full-text). Full-PDF verification of the original tables and survival curves cannot be performed. gap/no-fulltext-access

What HAS been cross-checked (2026-06-04): The primary quantitative result — mean age at death 69.3 years (castrated men) vs. 55.7 years (intact men), difference = 13.6 years — is confirmed via Min et al. 2012 (Current Biology 22:R792–R793, doi:10.1016/j.cub.2012.06.036), which cites Hamilton & Mestler 1969 directly and quotes these figures verbatim in the results section. That paper was read in full. The cohort n (number of subjects per group) was NOT reported in Min 2012 and remains unconfirmed from any OA source.

Mortality and survival: comparison of eunuchs with intact men and women in a mentally retarded population

Hamilton JB, Mestler GE · Journal of Gerontology 24(4):395–411 · 1969 · DOI: 10.1093/geronj/24.4.395

253 citations (Crossref). Research article (full paper; 17 pages). Closed access.

TL;DR

Hamilton and Mestler analyzed mortality records for a cohort of men who had been castrated (surgically or by disease) and were residents of a single institution for the intellectually disabled, compared against intact male and female residents of the same institution. Castrated men reportedly survived substantially longer than intact men in the same institutional setting — a finding widely cited as early human evidence that testosterone deprivation extends male lifespan. The institutional setting is both a strength (controlling for gross socioeconomic variables) and a confound (not generalizable to typical male populations; intellectually disabled populations have distinct health profiles).

Study Design

Population

Residents of an institution for the intellectually disabled in the United States (institution not specified in secondary accounts). Included men who had been castrated (via orchiectomy, secondary to trauma, or from other medical causes) and compared them to intact men and women from the same residential setting.

Rationale for the design

The institutional setting was chosen to control for major confounds that plague most comparisons of eunuchs to intact men in historical records:

  • Diet is equalized within the institution
  • Healthcare access is equalized
  • Physical activity and occupation are similar
  • Social determinants of health are broadly controlled

This makes it more plausible that observed survival differences reflect biological effects of castration rather than lifestyle or socioeconomic differences.

Outcomes

Mean and maximum age at death; survival curves across age categories from adolescence through old age.

Reported findings (from verified secondary citation; full text not OA)

The following figures are cross-checked against Min et al. 2012 1, which cites Hamilton & Mestler 1969 directly and quotes their figures in the results section of the peer-reviewed paper (full text read 2026-06-04):

  • Mean age at death, castrated men: 69.3 years
  • Mean age at death, intact men: 55.7 years
  • Lifespan difference: 13.6 years (Min 2012 rounds this to “14 years longer”)
  • Castrated men in the institution survived substantially longer than intact men.

The following claims appear in tertiary/review accounts of this paper but were not confirmed in any peer-reviewed secondary source read for this page and therefore cannot be stated with confidence:

  • Women’s survival being intermediate between castrated and intact men.
  • Effect being most pronounced at older ages.
  • Age at castration modifying the longevity advantage.
  • The total cohort n (number of subjects per group) — not reported in Min 2012 or any other OA source reviewed. gap/no-fulltext-access

Verification status: The 13.6-year mean lifespan difference (69.3 vs. 55.7 years) is as reliable as secondary sourcing allows for a paywalled 1969 paper. All other quantitative details (n, survival curves, women’s data, age-at-castration effects) cannot be confirmed without primary PDF access.

Critical Appraisal

Strengths

  • Institutional control — controlling for socioeconomic status, diet, and healthcare access within a single facility is methodologically superior to cross-population historical comparisons (contrast: Min 2012’s Joseon court records comparison).
  • Full research article (not a letter) — presumably contains statistical analysis and survival curves with uncertainty estimates.
  • High citation count — 253 citations; widely regarded as one of the foundational human natural-experiment datasets for the androgen-longevity hypothesis.
  • Directional consistency with animal data — effect direction matches castration studies in cats, dogs, and mice.

Weaknesses

  • Institutionalized population with intellectual disability — this is not a generalizable sample. Residents of such institutions may have had unusually poor health baselines, specific comorbidities, and unusual mortality patterns that do not generalize to the broader male population. Institutional care quality in the mid-20th century US was highly variable.
  • Castration heterogeneity — surgical orchiectomy (clean androgen removal) and disease-related gonadal loss have distinct health contexts. If the castrated cohort included many men who lost testicular function due to trauma or infection, this introduces selection on pre-existing health.
  • No serum testosterone confirmation — androgen deprivation is inferred from castration status; partial androgen production from adrenal androgens (DHEA, androstenedione) is not accounted for.
  • No replication — this cohort has not been re-analyzed with modern survival statistics or confirmed in an independent institutionalized castrated cohort.
  • Era — medical and nutritional care in this era is substantially different from modern environments, limiting direct quantitative translation.

Extrapolation Assessment

DimensionStatus
Population: human?Yes
Effect replicated in independent cohort?Directionally replicated in Min 2012 (Korean eunuchs; also small-n historical)
Mechanism known?No — correlation with androgen absence; mechanism unknown
Animal model consistency?Partial — castration studies in cats, dogs, and mice show similar directional effects

Significance in the Wiki

Hamilton & Mestler 1969 is the earlier and more methodologically detailed of the two main human natural-experiment datasets. Together with Min 2012 2, it forms the human evidence base for the androgen-deprivation-longevity hypothesis — a base that is directionally consistent but small-n, unverified by modern standards, and not yet replicated in contemporary cohorts. See androgen-deprivation-longevity-hypothesis for the full synthesis and an accounting of the limitations.

Cross-References

Footnotes

Footnotes

  1. Min KJ, Lee CK, Park HN. “The lifespan of Korean eunuchs.” Current Biology 22(18):R792–R793, 2012. doi:10.1016/j.cub.2012.06.036. Full text read 2026-06-04. In the results section, Min 2012 explicitly states: “One study [Hamilton & Mestler 1969] reported that castrated men residing in a mental hospital lived 14 years longer (69.3 vs. 55.7 years) than intact men in the same hospital.” These figures are taken as the best available OA confirmation of Hamilton & Mestler’s primary numbers. Min 2012 does not report the n per group for the Hamilton & Mestler cohort. ↩

  2. min-2012-korean-eunuchs-lifespan · n=81 eunuchs · observational, letter · Korean court genealogical records · eunuch mean lifespan 70.0 ± 1.76 years vs. intact controls 50.9–55.6 years (3 comparison families) · disputed by Le Bourg 2015 (doi:10.1159/000435854) on statistical grounds · doi:10.1016/j.cub.2012.06.036 ↩

Graph View

Backlinks