The Māori population of New Zealand (NZ) represents the final link in a long chain of island-hopping voyages beginning in Taiwan and stretching across the South Pacific—the last of the great human migrations. There is evidence that this population originated from restricted groups of common ancestors around 800 years ago and underwent rapid growth within the geographic isolation of NZ (Marshall et al. 2005). Considering this unusual history we envisage that the Māori population will have formed distinctive genomic signatures, with haplotype blocks likely to extend over larger distances and allele frequencies likely to differ markedly from those of other human populations. It is plausible that genomic variation peculiar to Māori may partially explain differences in disease susceptibility in this indigenous population compared to Caucasians.

Variation in ADH enzyme activity influences metabolic response to ingested alcohol and susceptibility to abuse behaviour. This variation is due in part to single nucleotide polymorphisms (SNPs) within the genes that encode the alcohol dehydrogenase (ADH) enzymes (Thomasson et al. 1993). One commonly studied exonic SNP allele in ADH1B (variously referred to as Arg47His, ADH1B*2, or ADH2*2, but now called ADH1B*47His) has been associated with protection against alcohol dependence in numerous studies (Chambers et al. 2002a; Chen et al. 1997; Osier et al. 2002), and is common in Asian and Pacific populations (Chambers et al. 2002a; Chen et al. 1997). Previously, Chambers et al. (2002b) compared ADH genotype frequencies between alcoholics and non-alcoholics in a sample of young male native NZ Polynesians (Māori). They found that the ADH1B*47His frequency in diagnosed Māori alcoholic patients (0.15) was much lower than that in Māori non-alcoholic volunteer controls (0.42, P < 0.01), suggestive of a protective effect. Studies have also implicated other ADH gene variants in alcoholism traits and shown highly variable patterns of linkage disequilibrium (LD) across this genomic region among different ethnic populations (Osier et al. 2002; Edenberg et al. 2006).

Here we examine SNPs, and estimate LD, across the ADH region in a sample of Māori in an attempt to characterise the patterns shaped by migrations and that may have relevance to future genetic studies of alcohol response in people with Polynesian ancestry. We concentrate on comparisons between Māori and Caucasian populations because of the hypothesis that the differences in alcohol consumption patterns between these two major NZ ethnic groups may be partially due to genetic differences at the ADH loci.


Study population

We analysed a sample of 47 unrelated male and female individuals drawn from the general population of Wellington, NZ. The subjects reported having eight Māori great grandparents and as such might be representative of the ancestral Māori population (i.e. they have very little Caucasian admixture). All individuals provided informed consent prior to enrolling in the study. Approval for this phase of our research programme was granted by the Central Region Ethics Committee (New Zealand) in 2004.

SNP information

The nine SNPs were chosen to span the entire set of ADH genes on chromosome 4, and include both intronic and exonic polymorphisms. SNPs were chosen based on existing knowledge of mutations that were well characterised in the literature. No SNPs were chosen within the ADH1A gene because, at the time of SNP selection, we were not aware of any significant mutations within that gene. Genotyping was done via a service contract with the Australian Genome Research Facility (, which uses the Sequenom MassArray Genotyping System (Buetow et al. 2001). Figure 1 provides a visual overview of the ADH gene region, together with an indication of the relative positions of the nine SNPs.

Fig. 1
figure 1

An overview of a region of human chromosome 4, showing the relative positioning of the alcohol dehydrogenase (ADH) genes and single nucleotide polymorphisms (SNPs). The minor allele of the rs1229984 SNP (boxed) is the frequently studied mutation ADH1B*47His

Statistical analyses

Allele frequencies at SNP sites were calculated for the Māori sample and compared with Caucasian frequencies derived from the ALFRED (Rajeevan et al. 2003) and HapMap (The International HapMap Consortium 2005) databases. χ 2 values (1 df) were calculated and their associated probabilities determined using the CHIDIST function in Calc ( 2006). The GOLD program was used to identify LD (haplotype) patterns among SNPs within the Māori sample.

Results and discussion

SNP allele frequencies

Table 1 shows a comparison of the minor allele frequencies (MAFs) in the Māori and Caucasian samples. One SNP, rs1789882, was excluded due to the observed allele frequencies deviating from Hardy–Weinberg Equilibrium (HWE, P < 0.01). We suspect that the deviation is due to a failure in the genotyping assay, and has no biological significance. Allele frequency differences between these two populations were statistically significant for all remaining SNPs except for rs4699733 and rs971074.

Table 1 Minor allele frequency (MAF) statistics for single nucleotide polymorphisms (SNPs) spanning the alcohol dehydrogenase (ADH) region on chromosome 4

The largest observed difference in allele frequency was at the key SNP rs1229984 (ADH1B*47His), as reported previously by Chambers et al. (2002b). The minor allele was present at a frequency of 0.41 in the Māori population but was only observed at a frequency of 0.04 in the Caucasian population.

Figure 2 shows LD (haplotype) analysis results for ADH SNPs in the Māori sample. The figure indicates a region of relatively high LD between SNPs rs4699733 and rs971074. In this region there were two SNPs that were in complete LD (D’ = 1, P < 0.01) with each other (rs1229984/rs698).

Fig. 2
figure 2

LD patterns of SNPs spanning the ADH gene region in Maori. D’ values > 0.50 are statistically significant at α = 0.05

The haplotype frequencies for this major LD block spanning SNP rs1229984 and rs698 (Fig. 2) were markedly different between Māori and Caucasian populations (P < 0.05). Haplotype AA had a much lower frequency in the Caucasian population (0.02) compared to Māori (0.45). The GA and GG haplotypes had lower frequencies in the Māori (0.26 and 0.30 respectively) compared to Caucasian (0.54 and 0.44 respectively). The AG haplotype was not observed.

In this study we observed marked differences in the allele frequency between Māori and Caucasian groups at six SNPs spanning the ADH gene region. In Māori, a region of apparent high LD including the well-known ADH1B variant was identified, which is perhaps indicative of a large (≈200 kb) haplotype block of Polynesian origin.

In conclusion, we have identified very different haplotype signatures at the alcohol-metabolising genes in Māori compared to Caucasians. These findings probably reflect the unique genetic history of this population and provide important information for designing association (including admixture mapping) studies of the ADH genomic region in alcohol-related traits in Polynesians.