Telomere dysfunction seems to underlie the development of a range of human genetic, degenerative, aging diseases and cancer 
. Our results demonstrated that short telomeres in peripheral blood cells were a feature of hereditary breast cancer patients and that telomere shortening frequently occurred with successive generations in these families, suggesting that telomere shortening could be the mechanism to explain the phenomenon of age anticipation in this disease.
Decreasing age of onset in families with hereditary breast cancer has been observed before 
. Similar to what we have found, Paterson et al. 
reported earlier age of diagnosis, of between 6 and 9 years, in successive generations of breast cancer families. Another study performed in a smaller number of BRCA1, BRCA2 and BRCAX families demonstrated that the mean maternal age at diagnosis in the BRCA1 group was significantly lower comparing to the BRCA2 or BRCAX groups, and no significant difference was found in the mean age at diagnosis between mothers and daughters in BRCA1 families 
. In our study we detected lower anticipation effect in BRCA1 families comparing to BRCA2 or BRCAX. Since BRCA1 mutations predispose to breast cancer at an earlier age, it would be more difficult to have large differences between generations in BRCA1 families. However, there are not definitive conclusions about whether there exists a real anticipation phenomenon. We cannot exclude that there would be alternate explanations for the earlier age of diagnosis in daughters. This observation could be due to ascertainment bias of subjects and lead-time bias, as a result of early detection of cancers by extensive screening or surveillance programs in high risk families 
. More sophisticated statistical approaches and mechanistic studies are warranted to answer this complicated problem. To date there are studies for 
and against 
a real phenomenon of anticipation in breast cancer, but a final conclusion regarding anticipation is still an open matter that is complicated by the fact that there is no molecular mechanism to explain it. Importantly we provide data indicating that telomere shortening is a possible biological explanation for the complex phenomenon of anticipation in breast cancer.
There are several studies trying to find association between short telomeres and risk of breast cancer that have reported contradictory results 
. The fact that sporadic breast cancer showed normal telomere length distribution in our study agrees with previous results from case-control studies, focused on sporadic cases, indicating lack of association between telomere lengths in blood leukocytes with risk of breast cancer 
. Interestingly, one association study reported significant association with breast cancer risk in women under 50 years of age, but no association between telomere length and breast cancer in women 50 years of age or older 
. Hereditary breast cancer, which we found characterized by short telomeres, typically occurs in women under 50 years. Then, it seems that hereditary breast cancer is associated with short telomeres, especially in BRCA1
mutation carriers, as well as in a subgroup of BRCAX.
It has been reported that telomere attrition may be affected by factors such as smoking, oxidative stress or obesity 
. Recently a prospective case-control study in breast cancer suggested that telomere shortening mainly occurs after diagnosis, as an effect of chemotherapy or other aspects of disease progression 
. However, the observed telomere shortening in unaffected BRCA mutation carrier daughters, as well as shorter telomere length in hereditary versus sporadic cases, although they followed similar therapeutic strategies, indicate that telomere shortening is largely influenced by genetic events. Nevertheless, the results are limited by the absence of data regarding timing of the sample draw, and the application of any chemotherapy. Our results showing that telomeres of affected women from BRCA1
families were significantly shorter than normal controls suggest that these genes might be involved in telomere regulation. Similarly, we can speculate that other genes involved in telomere maintenance could also be susceptibility genes that explain at least part of the BRCAX families with short telomeres.
Telomeres were shorter in affected mutation carriers versus non carriers and, moreover, we found that telomeres shortened in successive generations in affected families comparing with controls, suggesting that this is a plausible mechanism explaining the observed anticipation effect in familial breast cancer. We can speculate that mutations in BRCA1/2 or other genes induce faster telomere attrition during life time increasing the probability of genomic instability and the risk of developing breast cancer. Breast cancer would develop, first as a consequence of the inherited mutation, as this is the critical risk factor, and then the telomere length could modify the age at which the cancer would appear, as shorter telomeres at birth would reach dramatic genetic instability earlier that longer ones. On the other hand, telomere shortening could be a process mainly associated with the presence of mutations in BRCA1/2, or some other genes, regardless of the disease pathophysiology or age of presentation. Interestingly, we found four families in which daughters showed shorter telomeres than their mothers although the age of breast cancer was not anticipated with respect their mothers (), indicating that other genetic modifiers or environmental factors would be also determining the age of onset of breast cancer.
Telomere length variations as well as different telomere erosion rates were observed within the different lymphocyte subsets 
. In our study we used unselected peripheral blood leukocytes to estimate telomere lengths and then it could be affected by cell-to-cell variations. Selection of specific lymphocyte subpopulation would provide a more refined knowledge of the role that telomere length could have in cancer risk assessment.
The fact that telomere shortening has been associated with anticipation in other diseases, such as dyskeratosis congenita 
, as well as Li-Fraumeni Syndrome 
, suggest that telomere shortening may be responsible for genetic anticipation in a wide spectrum of genetic diseases.
Our findings indicate that the study of telomere length would be of relevance in the clinical surveillance and design of appropriate screening tests for patients with familial breast cancer.