Evolution in human life-history traits is influenced by environmental factors and, when genetic components underlie the relations, by micro-evolutionary forces. Age at reproduction is largely influenced by the familial cultural context and socioeconomic level, besides the maternal well-being and genetic background. The Sardinian population is characterized by historico-geographical isolation and differentiates from Italian mainland and other European populations in bio-demographic and cultural characteristics, among which the tendency to delay maternity persisting through generations. In our study, we investigated whether, in Sardinia, areas of “reproductive longevity” exist, where a higher-than-average incidence of late maternities combines with a lower-than-average cost in terms of perinatal death. Data from the Italian Central Institute of Statistics regard all 1980–1996 Sardinian births. Using spatial analysis of late maternity (proportion of babies born to mothers aged ≥35 years) and associated perinatal mortality (proportion of babies stillborn and dead within 0–6 days born to mothers aged ≥35 years), we aimed at singling out areas where the indicators run high and low, respectively. The perinatal mortality cost associated with the advanced maternal age [odds ratios (95% CI)] was evaluated through multiple logistic regression models. We identified central inland excess areas qualified by higher incidence of late maternities (27% vs. 22% in nonexcess area) and lower cost in perinatal mortality [OR=1.38 (1.04–1.84) vs. OR=1.74 (1.55–1.96) in nonexcess area]. In these “reproductive longevity” areas, the inbreeding coefficient was 3.7-fold higher than in the nonexcess areas, suggesting possible population homozygosity in genetic factors affecting the trait. Further and deeper investigations on biological and environmental determinants could focus on these target areas.
Cultural, socioeconomic and biological factors influence the evolution of human life-history traits and can affect individual fitness in terms of fertility as well as progeny health and survival. In addition, if genetic factors underlie the relation between the above determinants and life-history traits, population structure and micro-evolutionary forces, such as mutation, genetic drift and selection, can lead to evolutionary changes and, consequently, to intra-specific heterogeneities. In terms of evolutionary fitness, natural selection is expected to favour high fertility, early age at first reproduction and late age at last reproduction. However, its role is largely influenced by several factors, such as maternal well-being and health status, biological and genetic characteristics, familial wealth and socioeconomic level. Among the cultural factors which contribute to modifying maternal life history, the tendency to delay reproduction has been widely spreading in the Western populations during the last 20 years, driven by the improvements in education and reproductive autonomy of the women as well as by their pursuit of social and economic success.
Because of the concurrence of biological, cultural and socioeconomic determinants, evaluation of the genetic components and of the strength of natural selection on female life history is an exacting task. As for age at last reproduction, a study on Australian twins reported heritability estimates for age at menopause (Kirk et al., 2001), and investigations on preindustrial Finns estimated significant heritability and additive genetic variation for female life-history traits. In particular, differences were highlighted in the selective pressure on the end of the reproductive period, according to the wealth classes (Pettay et al., 2005 and Pettay et al., 2007).
The postponement of reproduction, in spite of its relevant costs in terms of offspring survival, child and mother health (Fretts et al., 1995, Jolly et al., 2000, Joseph et al., 2005, Tarin et al., 1998 and van Katwijk and Peeters, 1998), is a widespread characteristic in the Sardinian population. Several peculiar traits, resulting from Sardinia's geographical and historical isolation, make the island population different from continental Italy as well as from other European countries and have been reported to cause demographic, biological and genetic differentiation among municipalities (Cavalli-Sforza et al., 1994 and Zei et al., 2003). Among these distinctive traits, we can include the socioeconomic context based on the patriarchal structure of the family and the sheep-rearing economy, the high endogamy and low immigration rate (Golini, 1967), the consequent genetic makeup that is also partially determined by the past malaria endemia (Caglia et al., 1997 and Modiano et al., 1986), the historical demographic characteristics of high and late fertility, the still present tendency to postpone marriage and childbearing (Astolfi et al., 2002, Golini, 1967, ISTAT, 1990-1997, Livi Bacci, 1977 and Zei et al., 1990), and the rates of high and exceptional longevity (Caselli and Lipsi, 2006, Gatti and Salaris, 2004 and Poulain et al., 2004).
These considerations led us to investigate whether in Sardinia local differentiations can be evidenced in the reproductive behaviour and associated outcome. More precisely, we aimed at exploring whether the female tendency to delay childbearing is heterogeneously distributed in the island, and whether areas qualified for “reproductive longevity” exist where a higher incidence of elderly mothers combines with a lower cost in terms of perinatal deaths, due to a reduced impact of late maternity on the chance of delivering a healthy child.
Using 1980–1996 Sardinian birth records, we first defined two indicators of late childbearing and perinatal death, and computed smoothed estimates to reduce their spatial instability (Cressie, 1993 and Lawson, 2001). Second, through isopleth maps, we represented the spatial course of the late maternity indicator as a continuum, not limited by administrative boundaries, and singled out critical areas of high incidence. Finally, through a qualitative graphical matching procedure and a quantitative evaluation of the risk of perinatal death associated with late maternal age, we explored whether such areas could be qualified for late and successful childbearing.
