Fax +41 61 306 12 34 E-Mail karger@karger.ch www.karger.com Original Paper Hum Hered 2006;62:213–220 DOI: 10.1159/000097305 Sex and Age Specificity of Susceptibility Genes Modulating Survival at Old Age Giuseppe Passarino a Alberto Montesanto a Serena Dato a Sabrina Giordano b Filippo Domma b Vincenzo Mari c Emidio Feraco c Giovanna De Benedictis a Departments of a Cell Biology and b Economy and Statistics, University of Calabria, Rende, and c Italian National Research Center on Aging, Cosenza, Italy Introduction The probability of reaching very advanced ages in good health depends on a complex interplay of genetic, environmental and stochastic factors. As for genes, it is difficult to disentangle the genetic network which affects the quality of aging and survival up to advanced age, con- sequently several strategies have been adopted over the years which utilize familial or population data. In popu- lation studies, the ideal design is based on longitudinal studies, but these studies are difficult to carry out and usually consider only advanced ages [1, 2]. On the other hand, most of the information regarding the genetics of human aging has been obtained by cross-sectional stud- ies, under the hypothesis that unfavorable genotypes should be dropped out of the population by a sort of ‘de- mographic selection’ [3] which finally results in an en- richment of favorable genotypes in the gene pool of very old people. However, cross-sectional studies may suffer from the lack of appropriate control groups, as cohort- specific effects may confound comparisons between very old people (for example centenarians) and younger co- horts [4]. The problem is hindered by the rapid changes which occur in human society that increase the level of population heterogeneity, thus introducing a further complicating factor. To cope with this problem, algo- rithms which integrate genetic and demographic data have been proposed [5] . Key Words Logistic regression  Longevity  Multilocus analysis  Survival phenotype Abstract Objective: We aimed to investigate the influence of the ge- netic variability of candidate genes on survival at old age in good health. Methods: First, on the basis of a synthetic sur- vival curve constructed using historic mortality data taken from the Italian population from 1890 onward, we defined three age classes ranging from 18 to 106 years. Second, we assembled a multinomial logistic regression model to evalu- ate the effect of dichotomous variables (genotypes) on the probability to be assigned to a specific category (age class). Third, we applied the regression model to a cross-sectional dataset (10 genes; 972 subjects selected for healthy status) categorized according to age and sex. Results: We found that genetic factors influence survival at advanced age in good health in a sex- and age-specific way. Furthermore, we found that genetic variability plays a stronger role in males than in females and that, in both genders, its impact is espe- cially important at very old ages. Conclusions: The analyses presented here underline the age-specific effect of the gene network in modulating survival at advanced age in good health. Copyright © 2006 S. Karger AG, Basel Received: May 11, 2006 Accepted after revision: September 14, 2006 Published online: November 22, 2006 G. Passarino Department of Cell Biology University of Calabria IT–87036 Rende (Italy) Tel. +39 0984 492 930, Fax +39 0984 492 911, E-Mail g.passarino@unical.it © 2006 S. Karger AG, Basel 0001–5652/06/0624–0213$23.50/0 Accessible online at: www.karger.com/hhe