How Rare Is Rare? Carrier Frequencies for Fanconi Anemia In the United States and Israel

Although carrier frequencies for Fanconi anemia (FA) have been estimated for several founder populations, carrier frequencies in different countries remain unclear. One exception is Israel. In Jewish populations, founder mutations have been identified, e.g. the FANCC IVS+4 A>T mutation in Ashkenazi Jews, and the FANCA 2172–2173insG mutation in Moroccan Jews. The ability to test for specific mutations has enabled screening studies: the carrier frequency in Israeli Ashkenazi Jews is around 1:85; limited available data suggest that carrier frequencies in other Israeli Jewish subgroups may be similar (Tamary et al. BJH 2000). FA also occurs in Israeli Arab populations but carrier frequencies have not been determined. In contrast, there are much less data for the United States (US). Swift (Nature 1971) estimated the US FA carrier frequency as 1:300, but this estimate was based on surprisingly limited data - in total, the 12 persons with FA born in New York State from 1956 until 1967 who were known to the author among the corresponding total of 4.2 million live births. Nonetheless, this figure remains widely cited and has not been updated even after 40 years. We sought to update this estimate given the biological importance of the FA pathway.

We applied Swift's approach (knowledge of the number of FA cases born during a given period with a known overall birth rate) to contemporary data for the United States and Israel. Specifically, we used the Hardy-Weinberg Law and demographic data from the Fanconi Anemia Research Fund (FARF, 488 FA) and the Israeli Fanconi Anemia Registry (ISFAR, 66 FA).

On average during the 1990s, 15 persons with FA were born each year in the US who eventually became known to the FARF, amongst the 4.0 million persons born each year in the US during that period. The corresponding Hardy-Weinberg carrier frequency is 1:257 (95% Confidence Interval: 1:240 – 1:277). This range describes a lower bound because ascertainment in FARF must be less than 100%. Even so, the lower confidence limit is significantly greater that 1:300. The true ascertainment in FARF is unknown. We propose that values between 40% – 60% are plausible. Using this range to adjust the observed birth incidence upwards, we obtained a plausible range for the carrier frequency of 1:156 – 1:209 [midpoint 1:182]. We applied the same approach to the ISFAR where 2.6 births per year were observed. We assumed 50% – 100% ascertainment by ISFAR since cases were identified through a country-wide hospital network. For the entire country of Israel (Jews and non-Jews combined), we obtained a plausible range for the carrier frequency of 1:66 – 1:128 [midpoint 1:93]. Hence, the range for Israel derived using Swift's indirect approach (estimated from birth rates) is broadly consistent with direct surveys.

The FA carrier frequency in the US may be higher than previously thought, around 1:200 or perhaps even higher. From the perspective of population genetics, our results suggest there is less difference between the average carrier frequency in the US and higher carrier frequencies of around 1:100 reported for a number of ethnic groups including Ashkenazi Jews. This is consistent with the facts that the general US population is heterogeneous mixture of descendents of many ancestral groups, and FA is found world-wide. Our results also suggest that some European populations may have higher carrier frequencies than currently recognized. Our findings are sensitive to a number of assumptions. Going forwards, large scale re-sequencing studies could more precisely determine how many persons in the general population carry causative alleles for FA and other rare recessive syndromes.

No relevant conflicts of interest to declare.