Attention has been recently drawn to the difficulty in establishing the actual prevalence of pyruvate kinase (PK) deficiency in the general white population. Beutler and Gelbart1 have estimated this to be 51 cases per million white population by gene frequency studies. We have been centrally registering all patients with PK deficiency within the former Northern Health Region of the United Kingdom since 1974. This is a mainly static population of 3.1 million2 with less than 3% from ethnic minority groups and a minimal apparent incidence of inbreeding.

During the period since 1974, we have had 10 patients with PK deficiency registered (see Table 1). The diagnoses were based on pyruvate kinase activity measurement. No genetic diagnoses were made. Our observed prevalence of 3.3 per million in a mainly white population (more than 97% white) is more than an order of magnitude lower than the 51 per million predicted by Beutler and Gelbart using gene frequency techniques/methods.

View this table:
Table 1.

Prevalence of PK in Northern Health Region, United Kingdom, 1930–99

During the decade 1960-69 (secondary “baby boomers” peak in the United Kingdom), peak prevalence was achieved at a rate of 8.5 per million of population. From our registrations there are very few older patients. We postulate that a possible explanation of a low prevalence of patients more than 60 years old is that the advent of routine blood transfusion and neonatal exchange transfusion did not occur until the post–World War II period. One further potential explanation of the discrepancy between observed and predicted prevalence of PK deficiency is that the condition is associated with a significant intrauterine death rate.


PK deficiency prevalence and the limitations of a population-based survey

There are no other studies in which an attempt has been made to relate the prevalence of pyruvate kinase (PK) deficiency to a population base, and therefore Carey et al's studies are of interest. It must be pointed out, however, that disease prevalence estimates based on the detection of affected patients are bound to be substantially lower values than estimates based on the prevalence of heterozygotes in the population. First, as Carey et al point out, prenatal or neonatal mortality lowers the frequency with which a disease is found in the population at large. A second important consideration is the efficiency of ascertainment. The diagnosis of PK deficiency is made only when the physician considers the possibility and has red cell enzyme assays performed. Even then errors in diagnosis are not infrequent. In 1979 we reported that only 6 of 13 PK-deficient patients had been diagnosed correctly even though PK assays had been performed.1-1 The results were no better in a 1989 study in which we found that only 4 of 17 PK-deficient patients whose red cells had been assayed for pyruvate kinase had been diagnosed correctly.1-2 For these reasons a population-based survey could well give results that are as much as an order of magnitude lower than those predicted from the gene frequency in the population.


  1. 1-1.
  2. 1-2.