The familial risk of 18,199 women with invasive and/or in situ cervix cancers when compared with that of 72,796 women free of cervical tumors was analyzed. The risk of cervical tumors was significantly higher for women with a mother and/or sister(s) with cervical tumors (odds ratio (OR) 5 1.79, 95% CI: 1.71-1.88) than with an affected grandmother and/or aunt(s) (OR 5 1.28, 1.22-1.35). This risk did not differ according to the family side of the affected relative. Taking the familial risk of 1.28 for second-degree relatives as the best estimate of a true heritable effect for 25% of gene sharing, the calculated heritable effect for full siblings, sharing 50% of their genes, would correspond to an OR of 1.56. This model suggests, as a best estimate, a heritable component of 71% and an environmental component of 29% in young familial cervical tumors. The data imply that familial risks for cervical tumors are best explained by complex multifactorial mechanisms. ' 2006 Wiley-Liss, Inc.Key words: family history; heredity; genetics; gene sharing; environmental effects Cervical cancer is the second most common cancer among women worldwide, and the leading neoplasm in developing countries.1,2 Infection with human papilloma virus (HPV) is the main causal factor of cervical cancer and the risks have been estimated to be over 60. 3,4 Familial aggregation for cervical cancer has been observed in most but not all studies, [5][6][7][8][9] and also the heritability estimates differ extensively.10-12 Sexual behavior is a major attribute of HPV infections, and this is related to many life-style and socioeconomic factors, which family members have in common. [13][14][15] We use here the nation-wide Swedish Family-Cancer Database, by far the largest ever used to study familial cancer. 16,17 This database has previously been used to study the effect of first-degree family history in invasive and in situ cervical cancers risk, both showing a familial risk of about 1.8-2.0.6,18 Here we investigate familial cervical tumors considering family histories of these tumors both in first-and second-degree relatives in order to distinguish the possible genetic component in familial cervical cancer, influenced by an assumed strong environmental factor.
Material and methodsThe Swedish Family-Cancer Database was described previously. 16,17 The information was gathered from the Multigeneration Register, censuses, Swedish Cancer Registry and national death notifications. Histories of invasive and in situ cervical tumors among mothers, sisters, grandmothers and aunts were extracted from the Multigeneration registry. Information from 1960, 1970, 1980 and 1990 censuses was considered. The information provided by the different registries was linked together using the national 10-digit personal number. Information on both maternal and paternal grandmothers was available for 1,550,374 women. Among those, 18,199 women, born between 1950 and 1983, had been diagnosed with invasive (N 5 597) and/or in situ (N 5 17,650) cervical cancer. Each case was matche...