Editor—The ACE gene has a 287 bp Alu insertion in intron 16.1 The presence (I) or absence (D) of this insertion produces three population genotypes, II, ID, and DD. The D allele has been proposed as an indicator of cardiovascular risk in several studies,2-4 although this was not supported in a large study on US physicians.5There is a codominant relationship betweenACE ID genotype and serum ACE levels in white populations, with the D allele associated with increased levels.1 It is not clear whether a similar relationship exists in black populations. One report showed no difference in serum ACE levels between the different I/D genotypic groups in American blacks.6 Two others, both on the Jamaican population,7 8 suggested an important impact of the D allele. It is possible that the black populations reported could have a genetic contribution from other ethnic groups. We therefore examined a Ghanaian population, where African descent was known back to the grandparental generation, to see if any relationship existed between ACE polymorphism status and circulating serum ACE levels.

There were 97 subjects, 70 males and 27 females. The mean age (SD, range) was 26.2 (4.9, 22-45) for males and 28 (7.2, 22-51) for females. None were on ACE inhibitors or any other medication, and all were normotensive, healthy volunteers from the Kumasi area and could be considered representative of that region. All subjects had Ghanaian parents and grandparents. No females were pregnant.

Blood for ACE levels was collected in serum tubes. These were centrifuged and the serum transferred to sterile plastic tubes and frozen at −20°C. For DNA extraction, blood taken into EDTA tubes was kept at −20°C. All specimens were transported frozen on dry ice and were only thawed before analysis.

PCR was performed exactly as described previously.9 Owing to the possibility of mistyping ID subjects as DD,10 all DD genotypes were confirmed with insertion specific primers, again as previously described.8 ACE serum levels were measured using “ACE Reagent” kit from Sigma in the diagnostic Chemical Pathology laboratory where the test is used routinely and has been quality control checked. Serum samples are known to dilute as expected in this test, but this was not performed with this set of samples.

The Kruskal-Wallis one way analysis of variance was used as the data were not normal even after log transformation.

The genotypes and serum ACE levels are given in table 1. The data are not normal and positively skewed and with greater variability in the ID and DD groups than in the II group (fig 1). Using the Kruskal-Wallis one way analysis of variance by ranks gives a p value of 0.03 for a tendency for members of some groups to exceed members of others. Table 1 gives the median and range for each group, from which it is evident that the members of the II group tend to have lower serum ACE levels than the ID or DD groups. There is a considerable spread of ACE enzyme activity in all groups, but more especially in the ID and DD groups. The ages in each genotypic group were similar at 26.8, 26.6, and 26.4 (means) years for II, ID, and DD respectively. There were more females in the II group than in the ID or DD groups, at 42%, 24%, and 27% respectively, but the serum ACE levels for males and females in the II group were 35.8 and 32.8 U/l, suggesting that gender was not an important confounder. The prevalence of the I allele in the population was 0.4, similar to that reported in the Jamaican black population.7 8 The sample population is in Hardy-Weinberg equilibrium.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Serum ACE in U/l plotted against ACE genotype.

1Three reports on black populations have been published, one on American blacks which showed an absence of any association between genotype and ACE serum levels,6 while the others showed a significant effect of the D allele.7 8 These last two studies, both on Jamaican populations, suggested the same general trend but with slightly differing results. Forresteret al 7 showed a significant difference between serum ACE in all three genotypic groups with the same codominant effect seen in whites. McKenzie et al 8 also showed significance between all groups, but this was less pronounced between II and ID than for any of the other cross comparisons. The ACE levels reported by McKenzieet al 8 showed considerable overlap between groups, and this was especially prominent between the ID and DD groups. We also found considerable variation in the ACE serum levels within genotypic groups, with the greatest scatter in the ID and DD groups. There were significantly lower serum ACE levels in the II group compared to the ID or DD groups in the Ghanaian population, but no difference between the ID or DD groups. This trend is seen in the data of McKenzie et al,8 but not at all in that of Forrester et al.7 Since these two reports studied the same population, it is possible that the difference is simply a statistical artefact.

Our data show that in a black African population the trend in McKenzieet al 8 is increased to produce a dominant effect of the D allele on ACE serum levels rather than codominant. It may be that there has been genetic input from white gene pools in the Jamaican population which has produced a less dominant relationship between the ACE D allele and serum ACE levels than we have shown. The fact that Forresteret al 7 attempted to show black ethnicity by having “three or four grandparents of predominantly African origin” shows the problems with such a population. Unlike Blom et al,6 we do find a relationship between ACE I/D polymorphism and ACE serum levels in the black population, but one where the D allele shows dominance rather that codominance. The numbers in this study are not large and the data could be influenced by this, but the sample is larger than that used by Rigat et al 1 to show the codominant influence of the I and D alleles in whites. Nonetheless, a much larger study in this or another black African population would be useful to confirm these data, with a matched white population as a comparison.