Main

MCAD catalyzes the first reaction of the β-oxidation cycles for 4-10-carbon fatty acids. Genetic MCAD deficiency was first identified by Kølvraa et al. in 1982(1), immediately followed by two other research groups(2, 3). A large number of patients with this disease have since been identified mostly in Caucasian populations of northwestern European origin, demonstrating that MCAD deficiency is one of the most frequent inborn metabolic disorders among these ethnic groups. Initially, there were no case reports from eastern and southern Europe, with an exception of a single case from Spain(4). Infants and young children with MCAD deficiency appear to be healthy, but may suddenly become ill after prolonged fasting with repeated vomiting and hypoglycemic coma. Such an episode could be fatal, mimicking sudden infant death syndrome in some patients. Some other patients may have a recurrence of episodes(5). For these reasons, MCAD deficiency is one of the important diseases to be considered in the diagnosis of young infants with drowsiness, floppiness, and unresponsiveness with hypoglycemia of unknown origin.

The recent progress in the study of the molecular basis of MCAD deficiency revealed an unusually high prevalence of a single mutation within the MCAD gene locus(69). This mutation, an adenine 985 → guanine transition, causes glutamate substitution for lysine 329 in the MCAD precursor (lysine 304 in the mature MCAD protein). It has since been shown that the chaperonin-mediated folding of the mature form of this variant peptide and its assembly into the native tetrameric form in the mitochondria are perturbed(1012), resulting in the instability(10, 12) and disappearance of the variant protein from the patients' cells(1315).

To facilitate the detection of this important variant allele, a simple convenient method, that can be applied to dried blood samples on filter paper, was developed(6, 7, 9, 16). Two extensive retrospective studies using this method(17, 18) indicated that approximately 80% of MCAD-deficient patients were homozygous for the G985 variant allele. In the 1992 collaborative compilation of the data from the retrospective studies of 172 unrelated MCAD-deficient patients worldwide, each representing an independent pedigree, 306 of 344 variant alleles were found to be G985, indicating that its prevalence was 88.9%(19). The subsequent studies of non-G985 variant alleles in the compound heterozygotes and those in the non-G985 patients resulted in the identification of more than 20 other mutations(17, 1922), but the incidence of each of the non-G985 variant alleles was found to be extremely low.

Two groups of investigators then studied the geographical distribution of MCAD-deficient patients. In the study of 55 MCAD-deficient patients mainly from the United States, Yokota et al.(17) sent a questionnaire to the caring physicians, and reported that, among 29 who responded, 19 and 5 were descendants from the British Isles and Germany, respectively, with these two areas representing 82% of the totalG985 alleles involved. In their study of 110 MCAD-deficient patients in Europe, Gregersen et al.(18) found that the number of MCAD-deficient patients due to the G985 allele was particularly high in the United Kingdom and the Netherlands, followed by Germany, with these three countries accounting for 78% of the total. Furthermore, these two groups of investigators demonstrated that theG985 allele, in all the cases that they studied, were linked to a single haplotype (H-1 or 112) among the four known haplotypes among a general Caucasian population, defined using restriction polymorphism(17, 22). These results appeared consistent with a notion that a adenine 985 → guanine mutation occurred in a person, possibly in an ancient Germanic tribe, and that the current G985 alleles all derived from this single source by inheritance(17, 22).

Direct delineation of such opportunistic clinical observations as above are subject to various biases. For instance, the effectiveness of recognizing patients may greatly differs from community to community, or from country to country, depending on the level of awareness of MCAD deficiency among physicians and in the general population. The difference in the ability to make a diagnosis at local medical institutions is also another important source of bias. It should be noted that the history of MCAD deficiency is relatively short. The confirmatory diagnosis requires cumbersome enzyme assays or metabolite analysis using special mass spectrometric methods that make diagnoses difficult. It is possible that patients with MCAD deficiency may not have been detected in some areas in which the awareness of the disease is low and the advanced diagnostic techniques are unavailable.

To estimate the incidence of chromosomes carrying the G985 allele, two groups of investigators independently carried out preliminary surveys of less than 500 newborns each in England, resulting in the incidence ranging from 1 in 40 to 1 in 68 newborns(23, 24). Furthermore, one of the groups surveyed the G985 allele incidence in two states in the United States, as well as in Australia and Japan. They found in each of the states and Australia a slightly lower incidence than in England and detected none in Japan. A larger scale, systematic multination survey was needed for more accurate clinical significance and evolutionary information of this important variant allele. In 1991, we made a plan for a collaborative multination study of the incidence of chromosomes carrying the G985 allele. In this study, we report the results from the survey of the incidence of the G985 variant allele in the general newborn populations in Belgium, Germany, Poland, Czech Republic, Hungary, Bulgaria, Spain, Turkey, and Japan.

METHODS

Sources of the blood sample. Samples of blood were soaked and dried on filter paper and sent to a central laboratory where the analysis was performed. Approximately 1000 samples each were collected from newborns in Belgium (Belgian Limburg and Flemish Brabant), each of three areas of Germany(Bavaria, Westfalen, and Schleswig-Holstein), Poland (central and western regions), Czech Republic (Prague and central Bohemia), Hungary (southeastern region), Bulgaria (entire country), Spain (Catalonia), Turkey (Ankara), and Japan (Kagawa Prefecture). All samples, except for those from two areas of Germany (Bavaria and Westfalen), Spain, and Japan, were first sent toÅrhus, Denmark. At Århus, the samples from each country were divided into two groups of equal numbers (500 samples). One group of the samples were analyzed at Århus, and those of the second group were analyzed at New Haven, CT. The samples from Turkey and Japan were entirely, or almost entirely, sent to and analyzed at New Haven. All samples from Bavaria and Westfalen and Spain were analyzed in Münster, Germany, and Barcelona, Spain, respectively.

Method for G985 allele survey. Analyses of samples for theG985 allele were performed using the previous method based on the restriction digestion of the PCR-amplified fragments with NcoI orSty I(7, 16). When the screening of all 500 samples from a country was completed either at Århus or New Haven, the samples that exhibited a G985 allele were sent to the other group for confirmation.

Haplotype analysis. All haplotype studies were carried out atÅrhus. Haplotypes were defined with regard to the three two-allelic restriction enzyme recognition site polymorphisms(22) and to the six-allelic GA/GT microsatellite polymorphism(GT-repeat)(25). The BamII andPst I sites were determined by restriction fragment length polymorphism as previously described(22). For theTaq I site, a previously published PCR method(25) was optimized for blood-spot analysis(16). The GT-repeat was defined and determined in isolated DNA as follows. The MCAD gene contains a putative polymorphicGT-repeat in intron 5 (the structure was kindly provided by Arnold Strauss, St. Louis, MO). We defined a set of PCR primers flanking thisGT-repeat. The sense primer was biotinylated with a sequence of Bio 5′-CAACAGTTTCAGAATAGAGC. The sequence of the antisense primer was 5′-TAAACCTCTGAAAGCAACTC. PCR was performed using a Perkin-Elmer DNA Thermal Cycler model 480 in standard buffer (Perkin-Elmer), with 35-45 denaturation cycles at 94°C for 1 min, annealing at 55°C for 1 min, and extension at 74°C for 2 min. After PCR, the sense strand containing biotin was purified by capturing on strepavidin/magnetic beads (DYNA-beads, Dynal, Norway) using the method recommended by the manufacturer. The purified product was sequenced using the antisense PCR primer as sequencing primer and the Sequenase/fluorescence labeled dideoxynucleotides kit (Perkin-Elmer). Sequences were read using an ABI 377A Automated DNA sequencer.

For the determination of the haplotype frequency, the PCR assay for theGT-repeat was optimized for blood-spot analysis by including 8% DMSO in the buffer (10 × buffer: 670 mmol/L Tris·HCl (pH 8.8), 166 mmol/L ammonium sulfate, 100 mmol/L mercaptoethanol), and using32 P-labeled dCTP for detection. The PCR conditions were the same as above. The labeled PCR products were electrophoresed on a PAGE gel, and bands were determined by autoradiography.

RESULTS

Survey for chromosomes carrying the G985 allele. We carried out surveys for chromosomes carrying the G985 allele of newborns from Belgium, three areas of Germany, Poland, Czech Republic, Hungary, Spain, Bulgaria, Turkey, and Japan. The results from the present study are shown in Table 1. The total number of samples analyzed and the numbers of G985 homozygote and heterozygotes in each country are shown in the second, third, and fourth columns, respectively. Carrier frequency, shown in the fifth column, was calculated from these figures, considering that all the G985 alleles were distributed in the heterozygous form. The 95% confidence limits(26) of the carrier frequency are shown in parentheses. For each of the European countries and Turkey, approximately 1000 samples were analyzed, and 5 to 13G985 heterozygotes were detected. A single homozygote was identified in both Spain and Bulgaria, but no homozygotes were found in other countries. We found no G985 allele at all among 500 samples from Kagawa Prefecture in southwestern Japan.

Table 1 Survey for children carrying 985G allele in various countries

Among the countries surveyed in the present study, Belgium had the highestG985-carrier frequency, 1 in 77, which was similar to, or slightly lower than, that found in England. The frequencies found in three areas in Germany, namely Bavaria, Schleswig-Holstein, and Westfalen, were almost identical, ranging from 1 in 107 to 1 in 128. They were somewhat lower than that in Belgium. Unexpectedly, G985 allele-carrying chromosomes were also detected in Slavic countries, such as Poland (1 in 98) and Bulgaria (1 in 91), in frequencies that were only slightly lower than that in Belgium but somewhat higher than those in Germany. The frequencies in Spain (1 in 143), Hungary (1 in 168), and Czech Republic (1 in 240) were considerably lower, only half to one third as high as that in Belgium. The variant allele was also detected in Turkey in a frequency (1 in 216) that is slightly higher than that in the Czech Republic, but the confidence limits of the data for these two countries are low.

Haplotype analysis. We analyzed the haplotype of the chromosomes from G985 heterozygotes that were detected in various countries. In the haplotype determination, the analysis for CA/GT microsatellite(25) was included, in addition to the determination of three two-allelic polymorphisms using BanII,Pst I, and TaqI, that are known to define four haplotypes in the general Caucasian population(17, 22).

Preliminary results of sequencing DNA samples (22 alleles) isolated from 11 control individuals with a variety of haplotypes, defined by the three restriction sites (BanII, PstI, and TaqI), indicated that the microsatellite structure was in all cases(GT)nATGTCTGTGTGTGT, with n ranging from 16 to 20. We then examined DNAs from seven patients from six European countries and the United States, who are homozygous for G985. The structure, comprised of a variable and an invariable part, was also present, but in contrast to controls, the GT-repeat number, n, was 19 in all sevenG985 homozygotes, suggesting that the G985 mutation, without exception, is located on an allele with the 1 1 2(GT)19 haplotype.

We then analyzed, for the GT-repeat polymorphism, a total of 50 control chromosomes from 12 independent Danish families, who were previously used to determine the population frequencies of the four haplotypes defined by the three two-allelic polymorphisms(22). The results indicated that six different (GT)n alleles exist, with numbers varying from 16 to 21 as observed in the preliminary survey(Table 2). When the CA/GT microsatellite determination and the three two-allelic polymorphisms were combined, a total of nine haplotypes were identified in the general Danish controls, and their frequencies are shown in Table 2.

Table 2 Haplotypes in 50 independent Danish control chromosomes defined by three two-allelic and one six-allelic polymorphisms at the MCAD gene locus

The blood spots from 57 newborns showing heterozygosity for G985 and one Bulgarian homozygous sample were then subjected to PCR assays for theTaq I and GT-repeat polymorphism. The numbers of heterozygote samples from each country were: Belgium, 8; Germany, 9; Denmark, 6; Italy, 3; Hungary, 4; Czech Republic, 5; Bulgaria, 7; Poland, 10; and Turkey, 5. Such determination would provide information regarding the time and location of the occurrence of adenine 985 → guanine mutation. The results are summarized in Table 3.

Table 3 Haplotypes of hetero- and homozygotes detected as defined by RFLP for TaqI site and GT-repeats at the MCAD gene

In the Bulgarian homozygous baby, the haplotype of both G985 alleles could be assigned as 1 1 2 (GT)19. In 57 heterozygous cases, only a diplotype could be determined in the single analyses. In all cases, the diplotype contained at least one TaqI2 and one(GT)19 allele, however (Table 3). This observation is consistent with the notion that one allele possesses bothTaq I2 and (GT)19, and therefore, the G985 allele in heterozygotes is also most probably with the 1 1 2(GT)19 haplotype, as in the case of the Bulgarian homozygote.

After taking out the G985 allele with TaqI2(GT)19 haplotype from each diplotype, the haplotype frequencies of the non-G985 alleles in the panEuropean heterozygous samples (Table 3, bottom half) are considerably different from those seen in the general population in Denmark (Table 2). Especially (GT)18 was much more prevalent in the panEuropean neonatal materials than in the Danish materials, adding one new haplotype (type 5) that was not observed in the latter.

Compilation of data by us and others on the distribution of chromosomes carrying the G985 allele. In the course of our study,G985 allele incidences in a number of other countries have been reported. Thus, combining the results from the present study with those by others, a comprehensive compilation of the incidences of chromosomes carrying the G985 allele in various countries was established as shown in Table 4. Listing was done in the approximate order of northern/western to southern/eastern countries in Europe, with data shown in the form of gene frequency (frequency of G985 allele per 1000), with 95% confidence limits(26). The data on the United States and Australian populations are listed in the lower section of the table.

Table 4 Gene frequency of 985G allele in various countries

The numbers of newborns surveyed in the two recent surveys, one in England(29) and the other in the Netherlands(30), are large, with the total amounting to 10 171 and 6 195, respectively. They detected 158 and 99 G985 heterozygotes, respectively, achieving narrow confidence of limits. The total number of samples surveyed in Germany and also that in three areas of France, excluding the Rhone-Alps region, were 3015 and 4504, detecting 26 and 43 ofG985 allele-carrying chromosomes, respectively, each giving a narrow range of confidence limits. In contrast, in the two early surveys in England(23, 24) (not listed in Table 4) and those in Scotland(28) and Russia(31), the sample sizes were 500 each or less, and only 200 samples were analyzed in the Finnish survey(27). Thus, the ranges of confidence limits in these studies are broad. To illustrate the variability of the data based on the survey of 500 samples or less, our observations at the two laboratories at Århus and New Haven, each having analyzed one half of the 1000 samples from each country, are given in the following, with the numbers of G985 heterozygotes found in each set shown in brackets: Belgium [8 versus 5], northern Germany(Schleswig-Holstein) [2 versus 7], Poland [7 versus 3], Czech [4 versus 1], Hungary [3 versus 3], and Bulgaria [3versus 7]. With the survey of 1000 samples, the confidence limit is smaller and the variability is considerably less than with the survey of 500 samples or less, but it is still far from being comfortably reliable.

Table 4 offers several lines of interesting information within the confines of this qualification. First, it is important to note that the G985 allele was found all over Europe, and in Turkey, the United States, and Australia, but was completely absent in 1000 samples from Japan. It has been well established that although Finns/Hungarians speak Uralic language, genetically, 87 and 90% of the genes of modern Hungarians and Finns, respectively, are European genes(38a). It is also known that although Turkish speak Anatolian language, they originally were mainly IndoEuropean-speaking people, who later started to use Anatolian languages, and that genetically, there is little difference between the Turkish and the people of the surrounding countries(38b). With regard to the G985 allele incidence among Africans, the only available information at present comes from the study of nonCaucasian residents of North Carolina, which are presumed to be mostly of African descent. TwoG985 heterozygotes were found among 984 individuals surveyed in this group. This frequency is approximately 17% of that found among Caucasian residents in the same area. Because it has been well established that the admixture of Caucasian genes in African-Americans can be as high as 25%(39), the G985 alleles found among African-Americans have been attributed to admixing Caucasian genes, implying that this variant allele was originally absent among Africans(18). Thus, the available data suggest that theG985 allele is distributed exclusively among Caucasoids.

DISCUSSION

The inclusion in this study of the analysis of the CA/GT microsatellite, a six-allelic polymorphism, made the characterization of theG985 allele more accurate by further splitting the four previously restriction-defined haplotypes in the normal population. Also, there wasa priori a possibility that this microsatellite may have mutated on the G985 allele, a feature useful for estimating the age for the occurrence of the mutation. The inclusion of CA/GT microsatellite analysis resulted in the definition of nine haplotypes in 50 independent control chromosomes from 12 Danish families (Table 2), and an additional haplotype found in some blood samples heterozygous forG985 from various European countries (Table 3). Our preliminary analyses of 13 G985 homozygotes from a variety of European countries and the United States showed that the allele carryingG985, without exception, was the 1 1 2 (GT)19 allele. The identification of this haplotype can be carried out using a PCR assay for TaqI2 and that for (GT)19, a convenient feature for mass survey using blood spots.

Using the combination of GT microsatellite and TaqI analyses, we established that, in the single Bulgarian homozygous sample,G985 was located on a set of two alleles, each havingTaq I2 and (GT)19, and therefore was defined of the 1 1 2 (GT)19 haplotype. For each of the 57 G985 heterozygous babies, only diplotypes could be defined using blood spots(Table 3). In each of the heterozygous babies, the diplotypes were compatible with the conclusion that G985 was located on the 1 1 2 (GT)19 allele, on the following evidence(Table 3). 1) In 12 heterozygous babies, the diplotype was TaqI2/TaqI2,(GT)19/(GT)19, thus unambiguously assigning the haplotype of both normal and G985 MCAD alleles to be 1 1 2(GT)19. 2) Ten other heterozygous babies possess the diplotype 2/2, (GT)18/(GT)19, but none are homozygous 2/2 (GT)18/(GT)18, strongly indicating that no G985 mutation is located on the TaqI2,(GT)18 allele. Thus, the haplotype of the G985 allele in the 10 babies of this group may also be assigned 1 1 2(GT)19. 3) In the remaining 35 heterozygous babies, although all of them possessed both 1 1 2 and (GT)19, the possibility that the G985 mutation is located on aTaq I1-defined haplotype was considered for the sake of theoretical possibility, but was excluded for the following reason. If we assume such a possibility, we need to speculate a bifocal occurrence of the mutation in such distant areas as Denmark, Italy, Bulgaria, Belgium, and Germany, where the homozygous 2/2, (GT)19/(GT)19 was found. This is improbable because of the slow mutation rate for point mutations (the TaqI site)(25). Thus, the haplotype study data indicate that the G985 alleles in all the panEuropean heterozygous and homozygous samples, including those from the three Slavic countries, Hungary, and Turkey, originated from a single ancestral origin. These data also indicate that the adenine 985 → guanine mutation occurred after the 10 known haplotypes had already been established in the MCAD gene, suggesting that the adenine 985→ guanine mutation occurred relatively recently in the prehistory of modern humans.

In the compilation of the frequency of the chromosomes carryingG985 allele in various countries, shown in Table 4, the variant allele is distributed over all of Europe and Turkey, plus the United States and Australia, where the main inhabitants are European descendants, but none was found in Japan. In Europe, incidences are high in the northern countries, from England to Russia, and also in Bulgaria in southeastern Europe, but are markedly lower in the southwestern countries. The distribution of G985 appears to be often discontinuous. For instance in England, G985 incidence in the West Midland region was 38% higher than that in the Trent region(29). A much greater difference was observed in France: although the G985 incidences in Aquitaine, Normandy, and Paris are similar, that in the Rhones-Alps area was 5 times lower than the combined incidence in the three other regions(3437). The incidences in Italy(Piedmont area), Scotland, Czech Republic, and Hungary are markedly lower than those elsewhere. It should be noted that the Piedmont area of Italy is contiguous to the Rhones-Alps area of France, suggesting that theG985 incidences in these two contiguous areas are indeed low. In Hungary, two surveys, one in the southeastern region and the other in Budapest, both resulted in similarly low incidences. The combined incidence in Hungary is comparable to that in the Czech Republic, indicating that the incidences in these two neighboring countries are indeed lower than those in the surrounding countries, such as Switzerland, Germany, Poland, and Bulgaria. The sample size of the Scottish study is too small to ascertain the apparent low incidence. The irregular distribution of the G985 allele is consistent with the notion that the adenine 985 → guanine mutation occurred relatively recently, as suggested by the haplotype study.

In considering the source of the G985 allele, it is important to note that the G985 incidence among ethnic Basques in Pays Basque, that stretches from northeastern Spain to southwestern France, may be low. Basques are genetic isolates and are considered to be remnants of the ancient European population(38c). Ninety percent of ethnic Basques live in the Spanish Pays Basque with the remaining 10% residing in the French part. In the on-going survey of Spanish Basques (A. Ribes, unpublished observation) so far analyzed samples from 290 ethnic Basque babies and have found no G985 heterozygotes. In France, as a part of their survey in the Aquitaine area, Ged et al.(35) analyzed 309 samples from the French Pays Basque, and found three G985 heterozygotes, a frequency similar to those in five other neighboring areas. Approximately 30% of the residents in the French Pays Basque are ethnic Basques, whereas less than 5% are in the neighboring regions. It is unknown whether any of the three G985 heterozygotes in the French Pays Basque were ethnic Basques or not (C. Ged, personal communication). The apparent low incidence, or absence, of G985 among Basques is consistent with the notion that G985 occurred outside of Europe and was brought into Europe by IndoEuropean-speaking people in the Neolithic or later period from their homeland, which is generally considered to be in the region encompassing southern Ukraine and southern Russia(40). The high incidences of G985 in Russia, Poland, and Bulgaria are consistent with this view.

Further studies are necessary in the future to confirm the exclusive distribution of G985 allele among Europeans and descendants, and to test the above hypothesis on its evolution. These include more extensive surveys in various European, Asian, and African countries, as well as surveys among Jews and Arabs. Surveys among Indians, Pakistani, and Iranians should be of particular interest for exploring whether the G985 allele occurred before or after the European and the IndoIranian families parted in their migration from their common homeland to each respective destiny.

The observations on G985 are in contrast to those onΔF508 mutation of the transmembrane conductance regulator(CFTR) gene causing cystic fibrosis, another common mutation with a panEuropean distribution. Unlike G985, ΔF508 is characterized by its diffuse, continuous distribution within Europe with a west-to-east and a north-to-south gradient(41, 42), a high incidence among Basques(43), and polymorphic haplotypes within the variant allele(44). Based on these data, the ΔF508 mutation is considered to have occurred in an indigenous European population in the Paleolithic or Mesolithic period, at least 10 000-50 000 y ago(38d, 44).

In Table 5, the incidence of the G985 homozygote and that of MCAD deficiency in each country was calculated for clinical consideration. The frequency is shown in the number of newborns from which one case is expected to be detected. MCAD deficiency incidence was calculated on the assumption that the combined frequency of mutations other than G985 was equal for each country, at the level of one eighth of the G985 frequency in England (combined). This figure was used based on the fact that in the earlier retrospective studies, in which the samples were from predominantly Anglo-Dutch/American patients, the frequency of theG985 allele was found to be 88.9% of the total variant alleles identified, with all the other rare mutations combined representing 11.1%(1719).

Table 5 Frequency of G985 homozygosity and MCAD deficiency in various countries

Recently, Ziadeh et al.(45) presented the results of a study in which 80 371 newborn blood samples were prospectively screened for increased acylcarnitine in Pennsylvania using tandem mass spectrometry. Increased acylcarnitine was found in nine babies. Upon molecular analysis, four of them were homozygous, whereas the remaining five were heterozygous for G985, with the G985 prevalence among these nine patients being 72%. From these data, they suggested that the previous retrospective studies of clinically ascertained cases may have led to preferential identification of the G985 allele. However, their contention is too premature at present. First, a group of only nine patients is too small for determining the prevalence of G985 among various MCAD variant alleles. The lower G985 prevalence in their series may simply be due to greater variance of the data caused by the small number of patients involved. Second, in the four presumed compound heterozygotes in which the second variant allele has not been identified, the diagnosis of MCAD deficiency needs to be ascertained by molecular identification of the second mutations, or by other confirmatory procedures such as enzyme assay.

Table 5 shows that in England, the Netherlands, Belgium, and Switzerland the frequency of MCAD deficiency ranges from 1 in 10 000 to 1 in 20 000 newborns, whereas those in Denmark, Germany, France(excluding the Rhone-Alps region), Poland, Bulgaria, and Russia were in the ranges of 1 in 25 000-40 000 newborns. The incidences in Spain, Italy, Scotland, Czech Republic, Hungary, and Turkey are considerably lower, with incidences of 1 in 56 000-200 000 newborns, representing only 1/4 to 1/14 of that found in England. As awareness and medical technologies improve in these countries, more patients are now being reported. For instance, six additional cases have recently been reported in Spain(46), and two G985 homozygotes were found among Gypsies in Portugal(47). In Italy, the firstG985 homozygote has recently been identified(48). As far as we are aware, no MCAD-deficient patients have ever been reported from the eastern European countries, with the exception of four patients from the Czech Republic last year (S. Kmoch, unpublished results). In Poland, in spite of an extensive nationwide metabolic disease detection program involving gas chromatography-mass spectrometry and a 3-phenylpropionic acid loading test in the last 11 y, no patients with MCAD deficiency have so far been found (E. Pronicka, personal observation).

The results of this and other recent surveys for G985 allele emphasize the need for raising the level of awareness of this disease, not only in the countries in the eastern and southern European countries, but also in northwestern European countries and in the United States as well. For instance, the recent study in England by Seddon et al.(29) showed that, although in Trent MCAD-deficient patients were detected in the number almost equal to that expected of gene frequency, the number detected in West Midland was only one sixth of that expected of gene frequency, probably leaving many patients undetected in the latter. They attributed this difference in the patients identification efficiency in the two areas to the differences in awareness of MCAD deficiency and diagnostic ability. Their observations in England strongly underscore the importance of further raising the level of awareness of this disease and improving the effectiveness of diagnosis of this disease in the advanced countries as well.