Dear Dr. Throckmorton,
As sadly anticipated, Drs. Paul and Kirk Cameron were not objective when writing their report â€™Federal Distortion of Homosexual Footprint (Ignoring Early Gay Death?)â€™ (1). The mission statement of their professional affiliation, the Family Research Institute, concludes as follows: â€We welcome all who would join in the fight to restore a world…where homosexuality is not taught and accepted, but instead is discouraged and rejected at every level.â€ As a consequence, any report on human sexuality originating from this institution will by definition be devoid of objectivity and of questionable scientific value.
As a statistical researcher, Dr. Kirk Cameron must know well the inferential problems that prevail when comparing the average age at death in two study groups with vastly different age distributions. Elementary textbooks in epidemiology warn against such undue comparisons because they lead to apparently common-sense, but overtly wrong, conclusions (2). Assume for the purpose of illustration that Cameron and Cameron had restricted their study to all newly-married and all newly-partnered people in Denmark during the study window 1990-2002 (Norway 1997-2002) with the aim to make the studied groups of homosexuals and heterosexuals more comparable. In Danish men, the median age at first homosexual partnership was 4 years higher (32.6 years) than the median age at first heterosexual marriage (28.6 years) in the period 1989-2001 and, in women, the difference was about 6 years, being 32.6 years for first homosexual partnership vs. 26.5 years for first heterosexual marriage (3). Among those relatively few newly-married and newly-partnered people who actually died in the short observation period, the average age at death would likely be higher in the homosexually partnered group than in the heterosexually married group, simply because of the older age distribution of the homosexually partnered group. Using the Cameronsâ€™ flawed logic of inference such a modification of their study design would lead to the opposite conclusion; i.e., that heterosexual marriages shorten peoplesâ€™ life span. Obviously, this conclusion would be as unsubstantiated as the one reached by the authors.
Working to promote their anti-homosexual agenda, the Camerons presumably have plenty of time and resources to discuss these issues at length. As a researcher continuously fighting hard to obtain the required funding and time for my projects and those of my students I will have to stop here, leave the Camerons with their tragic parody of science, and focus on true scientific questions instead. I have previously published studies in peer-reviewed scientific journals that were warmly applauded by gay advocacy groups (4) as well as studies that made me a persona non grata in the same circles (3). I donâ€™t have an agenda or a political mission for my scientific work, but I certainly have a different starting point than the one expressed in the mission statement of the Family Research Institute. Unlike what Dr. Kirk Cameron believes, I donâ€™t object to the theoretical possibility that homosexual persons may have somewhat shorter life spans than heterosexuals or, for that matter, the other way around. We just donâ€™t have meaningful prospective data available to inform us yet.
Although the Cameronsâ€™ report has no objective scientific value, the authors should be acknowledged for providing teachers with a humorous example of agenda-driven, pseudo-scientific gobbledygook that will make lessons in elementary study design and scientific inference much more amusing for future epidemiology students.
Morten Frisch, MD, PhD, DSc(Med)
1. Cameron, P. and Cameron, K. Federal Distortion Of Homosexual Footprint (Ignoring Early Gay Death?). 1-27. 2007. Family Research Institute.
2. Rothman KJ. Introduction to Epidemiologic Thinking. In: Rothman KJ (ed), Epidemiology – An Introduction. New York: Oxford University Press, 2002, 1-7
3. Frisch M, Hviid A. Childhood family correlates of heterosexual and homosexual marriages: a national cohort study of two million Danes. Arch Sex Behav 2006;35:533-47
4. Frisch M, Smith E, Grulich A, Johansen C. Cancer in a population-based cohort of men and women in registered homosexual partnerships. Am J Epidemiol 2003;157:966-72