Testing for bimodality in frequency distributions of data suggesting polymorphisms of drug metabolism--histograms and probit plots

3D Characterization of the Aortic Valve and Aortic Arch in Bicuspid Aortic Valve Patients

Patients with bicuspid aortic valve (BAV) commonly have associated aortic stenosis and aortopathy. The geometry of the aortic arch and BAV is not well defined quantitatively, which makes clinical classifications subjective or reliant on limited 2D measurements. The goal of this study was to characterize the 3D geometry of the aortic arch and BAV using objective and quantitative techniques. Pre-TAVR computed tomography angiogram (CTA) in patients with BAV and aortic stenosis (AS) were analyzed (n = 59) by assessing valve commissural angle, presence of a fused region, percent of fusion, and calcium volume. The ascending aorta and aortic arch were reconstructed from patient-specific imaging segmentation to generate a centerline and calculate maximum curvature and maximum area change for the ascending aorta and the descending aorta. Aortic valve commissural angle signified a bimodal distribution suggesting tricuspid-like (≤ 150°, 52.5% of patients) and bicuspid-like (> 150°, 47.5%) morphologies. Tricuspid like was further classified by partial (10.2%) or full (42.4%) fusion, and bicuspid like was further classified into valves with fused region (27.1%) or no fused region (20.3%). Qualitatively, the aortic arch was found to have complex patient-specific variations in its 3D shape with some showing extreme diameter changes and kinks. Quantitatively, subgroups were established using maximum curvature threshold of 0.04 and maximum area change of 30% independently for the ascending and descending aorta. These findings provide insight into the geometric structure of the aortic valve and aortic arch in patients presenting with BAV and AS where 3D characterization allows for quantitative classification of these complex anatomic structures.

Functional phenotyping of the CYP2D6 probe drug codeine in the horse

BACKGROUND

In humans, the drug metabolizing enzyme CYP2D6 is highly polymorphic resulting in substantial differences in the metabolism of drugs including anti-arrhythmics, neuroleptics, and opioids. The objective of this study was to phenotype a population of 100 horses from five different breeds and assess differences in the metabolic activity of the equine CYP2D6 homolog using codeine as a probe drug. Administration of a probe drug is a common method used for patient phenotyping in human medicine, whereby the ratio of parent drug to metabolite (metabolic ratio, MR) can be used to compare relative enzyme function between individuals. A single oral dose of codeine (0.6 mg/kg) was administered and plasma concentrations of codeine and its metabolites were determined using liquid chromatography mass spectrometry. The MR of codeine O-demethylation [(codeine)/(morphine + morphine-3-glucuronide + morphine-6-glucuronide)] was determined using the area under the plasma concentration-time curve extrapolated from time zero to infinity (AUC0-∞) for each analyte and used to group horses into predicted phenotypes (high-, moderate-, and low-MR).

RESULTS

The MR of codeine O-demethylation ranged from 0.002 to 0.147 (median 0.018) among all horses. No significant difference in MR was observed between breeds, age, or sex. Of the 100 horses, 11 were classified as high-MR, 72 moderate-MR, and 17 low-MR. Codeine AUC0-∞ and O-demethylation MR were significantly different (p < 0.05) between all three groups. The mean ± SD MR was 0.089 ± 0.027, 0.022 ± 0.011, and 0.0095 ± 0.001 for high-, moderate-, and low-MR groups, respectively. The AUC for the morphine metabolites morphine-3-glucuronide and morphine-6-glucuronide were significantly different between high-and low-MR groups (p < 0.004 and p < 0.006).

CONCLUSIONS

The MR calculated from plasma following codeine administration allowed for classification of horses into metabolic phenotypes within a large population. The range of codeine metabolism observed among horses suggests the presence of genetic polymorphisms in CYP2D82 of which codeine is a known substrate. Additional studies including CYP2D82 genotyping of high- and low-MR individuals are necessary to determine the presence of CYP2D polymorphisms and their functional implications with respect to the metabolism of therapeutics.

Evaluation of CYP2D6 phenotype in the Yoruba Nigerian population

BACKGROUND

There is a lack of information on CYP2D6, a major metabolizing enzyme, in Africa ethnic nationalities. The objective was to determine CYP2D6 phenotype in Yoruba Nigerians using dextromethorphan (DEX).

METHOD

A total of 89 healthy volunteers received 30 mg of DEX orally followed by blood and urine sample collection at 3-hour and over 8 h post-dose, respectively. DEX and dextrorphan (DOR) concentrations were determined using High Performance Liquid Chromatography (HPLC). The metabolic ratio (MR, DEX/DOR) were plotted for the phenotype determination.

RESULTS

The log MR that separated poor (PMs) from normal metabolizers (NMs) was 0.28 and 0.75 for urine and plasma, respectively. Two subjects (2.3%) identified as PMs had a mean MR of 17 and 3.2 in plasma and urine, significantly higher than that of NMs (p < .0001). A positive correlation between urine and plasma MR was noted.

CONCLUSION

The prevalence of PMs in the Yoruba Nigerians was similar to that reported among blacks.

Dextromethorphan and debrisoquine metabolism and polymorphism of the gene for cytochrome P450 isozyme 2D50 in Thoroughbreds

OBJECTIVE To characterize polymorphisms of the gene for cytochrome P450 isozyme 2D50 (CYP2D50) and the disposition of 2 CYP2D50 probe drugs, dextromethorphan and debrisoquine, in horses. ANIMALS 23 healthy horses (22 Thoroughbreds and 1 Standardbred). PROCEDURES Single-nucleotide polymorphisms (SNPs) in CYP2D50 were identified. Disposition of dextromethorphan (2 mg/kg) and debrisoquine (0.2 mg/kg) were determined after oral (dextromethorphan) or nasogastric (debrisoquine) administration to the horses. Metabolic ratios of plasma dextromethorphan and total dextrorphan (dextrorphan plus dextrorphan-O-β-glucuronide) and 4-hydroxydebrisoquine concentrations were calculated on the basis of the area under the plasma concentration-versus-time curve extrapolated to infinity for the parent drug divided by that for the corresponding metabolite. Pharmacokinetic data were used to categorize horses into the phenotypic drug-metabolism categories poor, extensive, and ultrarapid. Disposition patterns were compared among categories, and relationships between SNPs and metabolism categories were explored. RESULTS Gene sequencing identified 51 SNPs, including 27 nonsynonymous SNPs. Debrisoquine was minimally detected after oral administration. Disposition of dextromethorphan varied markedly among horses. Metabolic ratios for dextromethorphan ranged from 0.03 to 0.46 (mean, 0.12). On the basis of these data, 1 horse was characterized as a poor metabolizer, 18 were characterized as extensive metabolizers, and 3 were characterized as ultrarapid metabolizers. CONCLUSIONS AND CLINICAL RELEVANCE Findings suggested that CYP2D50 is polymorphic and that the disposition of the probe drug varies markedly in horses. The polymorphisms may be related to rates of drug metabolism. Additional research involving more horses of various breeds is needed to fully explore the functional implication of polymorphisms in CYP2D50.

Interindividual Variations in Enzymes Controlling Organophosphate Toxicity in Man

1 Interindividual variations in an unexposed population have been defined for five enzymes involved in organophosphate (OP) toxicity. The enzymes measured were: red blood cell acetylcholinesterase (AChE), lymphocyte neuropathy target esterase (NTE), serum cholinesterase (ChE), serum paraoxonase and serum arylesterase.

2 AChE and arylesterase were normally distributed in the population whilst the distribution of NTE, ChE and paraoxonase deviated significantly from normal.

3 Assay precision and intra-individual variability were measured for each of the enzymes; the effect on interindividual variation was assessed.

4 Variations in enzyme activities between individuals could have profound effects on susceptibility to OP toxicity. Prior determination of these enzymes may be predictive of susceptibility.

5 Lymphocyte NTE has some limitations as an indicator of exposure to neurotoxic OPs.

Evidence for polymorphism in the cytochrome P450 2D50 gene in horses

Metabolism is an essential factor in the clearance of many drugs and as such plays a major role in the establishment of dosage regimens and withdrawal times. CYP2D6, the human orthologue to equine CYP2D50, is a drug-metabolizing enzyme that is highly polymorphic in humans leading to widely differing levels of metabolic activity. As CYP2D6 is highly polymorphic, in this study it was hypothesized that the gene coding for the equine orthologue, CYP2D50, may also be prone to polymorphism. Blood samples were collected from 150 horses, the CYP2D50 gene was cloned and sequenced; and full-length sequences were analyzed for single nucleotide polymorphisms (SNPs), deletions, or insertions. Pharmacokinetic data were collected from a subset of horses following the administration of a single oral dose of tramadol and probit analysis used to calculate metabolic ratios. Prior to drug administration, the ability of recombinant CYP2D50 to metabolize tramadol to O-desmethyltramadol was confirmed. Sequencing of CYP2D50 identified 126 exonic SNPs, with 31 of those appearing in multiple horses. Oral administration of tramadol to a subset of these horses revealed variable metabolic ratios (tramadol: O-desmethyltramadol) in individual horses and separation into three metabolic groups. While a limited number of horses of primarily a single breed were studied, the variability in tramadol metabolism to O-desmethyltramadol between horses and preliminary evidence of what appears to be poor, extensive, and ultra-rapid metabolizers supports further study of the potential for genetic polymorphisms in the CYP2D50 gene in horses.

Age at onset in bipolar I disorder: two may be better than three subgroups

INTRODUCTION AND OBJECTIVE

Age at onset in bipolar disorder is related to prognosis and to treatment response. However, it is not clear if there are three or two subgroups in relation to age at onset. The objective of this study is to analyze the number of subgroups in relation to age at the beginning of the disease in a representative sample of bipolar I patients and to compare the subgroups in relation to clinical variables.

METHOD

We included 169 patients diagnosed with bipolar I disorder. Normal mixture analysis was performed. The subgroups of patients formed above were compared regarding clinical characteristics. Patients were followed-up during six years.

RESULTS

We found three ages at onset subgroups. The early onset group (18.2±2 years) included 34% of the patients. The second group (26.1±5.5 years) included 44% of the patients. The third group (50.9±9.1 years) included 22% of the patients. Early and intermediate onset groups were not significantly different, and had more family history of affective disorders, more psychotic symptoms, more history of suicide attempts and more history of drug abuse history than the late onset group.

CONCLUSIONS

Our results suggest that there are three groups of age at onset but early and intermediate groups are similar in relation to clinical variables. The late onset group includes almost a quartile of patients and has different clinical profile.

An indirect approach to imply trade-off shapes: population level patterns in resistance suggest a decreasingly costly resistance mechanism in a model insect system

Trade-offs between life history and other traits play a key role in shaping the evolution of individuals. It is well established theoretically that the shapes of trade-off curves are as crucial to the evolutionary outcome as their strengths. However, measuring the shape of these relationships directly is often impractical. Here we use an indirect approach that examines the patterns seen within a population and then use theory to infer the shape of the trade-off curve. Using a bioassay we found that most individuals had either high susceptibility or relatively high resistance to a microparasite in a lepidopteran host population. According to general theory, this type of pattern in resistance would be most likely with a deceleratingly costly impact on fitness of increasing resistance. The implications and generality of the approach are discussed, along with the implications of the results to our understanding of the nature of innate resistance to parasites.

Nonparametric Expectation Maximisation (NPEM) Population Pharmacokinetic Analysis of Caffeine Disposition from Sparse Data in Adult Caucasians

OBJECTIVE

To explore the ability of the nonparametric expectation maximisation (NPEM) method of population pharmacokinetic modelling to deal with sparse data in estimating systemic caffeine clearance for monitoring and evaluation of cytochrome P450 (CYP) 1A2 activity.

DESIGN AND PARTICIPANTS

Nonblind, single-dose clinical investigation in 34 non-related adult Bulgarian Caucasians (18 women and 16 men, aged between 18 and 62 years) with normal and reduced renal function.

METHODS

Each participant received oral caffeine 3 mg/kg. Two blood samples per individual were taken according to the protocol for measuring caffeine plasma concentrations. A total of 67 measured concentrations were used to obtain NPEM estimates of caffeine clearance. Paraxanthine/caffeine plasma ratios were calculated and correlated with clearance estimates. Graphical methods and tests for normality were applied and parametric and nonparametric statistical tests were used for comparison.

RESULTS

NPEM median estimates of caffeine absorption and elimination rate constants, k(a) = 4.54 h(-1) and k(el) = 0.139 h(-1), as well as of fractional volume of distribution and plasma clearance, V(S1) = 0.58 L/kg and CL(S1) = 0.057 L/h/kg, agreed well with reported values from more 'data rich' studies. Significant correlations were observed between paraxanthine/caffeine ratios at 3, 8 and 10 hours and clearance (Spearman rank correlation coefficients, r(s), >0.74, p </= 0.04). Sex or renal function caused no significant differences in clearance. Heavy smokers and drinkers showed 2-fold higher CYP1A2 activity. Normality tests and graphical methods of analysing caffeine clearance supported a non-Gaussian and multicomponent distribution of CYP1A2 activity.

CONCLUSIONS

Collectively, the results show that the NPEM method is suitable and relevant for large-scale epidemiological studies of population phenotyping for cancer susceptibility and for abnormal liver function by monitoring CYP1A2 activity based on sparse caffeine data.

Phenotypic polymorphism and gender-related differences of CYP1A2 activity in a Chinese population

AIMS

To investigate the distribution characteristics of CYP1A2 in a Chinese population, and to examine gender-related differences in CYP1A2 activity.

METHODS

Two hundred and twenty-nine healthy subjects, 120 men and 109 women, were enrolled in this study. CYP1A2 activity was measured by plasma paraxanthine/caffeine (1,7X/1,3,7X) ratio 6 h after administration of 300 mg caffeine. The concentrations of paraxanthine and caffeine in plasma were detected by h.p.l.c.

RESULTS

A 16-fold variation of CYP1A2 activity (range 0. 09 to 1.46) was shown in this study. The coefficient of variation (CV %) of CYP1A2 activity was 62.9%. Non-normal distribution of CYP1A2 activity was indicated by the Shapiro-Wilk test (P<0.001). Probit plots of CYP1A2 activity revealed a bimodal distribution with breakpoint of 1,7X/1,3,7X ratio of 0.12. The percentage of poor metabolizers (PMs) was 5.24% (95% CI: 2.35% approximately 8.13%) in this Chinese population. Residual analysis of the data also supported bimodality (P<0.01). The CYP1A2 activity of men was higher than that of women (median: 0.33 vs 0.23, P<0.001). A probit plot of CYP1A2 activity in men was shifted to the left compared with that in women. Based on phenotype, the gender-related difference was observed in extensive metabolizers (EMs) (P<0.001), but not in PMs (P >0.1). In addition, there was no sex-related difference in the incidence of PMs (P >0.1).

CONCLUSIONS

There is a phenotypic polymorphism in CYP1A2 activity in this Chinese population, and CYP1A2 activity is higher in men than that in women.

Fluoxetine bioequivalence study: quantification of fluoxetine and norfluoxetine by liquid chromatography coupled to mass spectrometry

In this study, the authors assessed the bioequivalence of two fluoxetine tablet formulations in 24 healthy volunteers of both sexes who received a single 20 mg dose of each fluoxetine formulation, and a new sensitive method for the quantification of fluoxetine and norfluoxetine in human plasma was developed. The study was conducted using an open, randomized, two-period crossover design with a 4-week washout interval. Plasma samples were obtained over a 672-hour period. Plasma fluoxetine and norfluoxetine concentrations were analyzed by combined liquid chromatography coupled to mass spectrometry (LC-MS) with positive ion electrospray ionization using selected ion recording (SIR). Kolmogorov-Smirnov's test, histograms, probit plots, and the correlation between norfluoxetine AUC(0-infinity) and fluoxetine AUC(0-infinity) were used to analyze the population distribution. The limit of quantification was 0.15 ng.ml-1 and 0.50 ng.ml-1 for both fluoxetine and norfluoxetine, respectively. Within- and between-run imprecision was less than 13% and 17%, respectively. The pharmacokinetic parameters obtained for fluoxetine and norfluoxetine after the administration of each formulation included AUC(0-672 h), AUC(0-infinity), Cmax, Cmax/AUC(0-672 h), tmax, t1/2, and Ke. The AUC values for fluoxetine were not consistent with a normal distribution, reflecting the existence of two different populations (poor and extensive metabolizers). The mean pharmacokinetic parameters for extensive fluoxetine metabolizers were 27.0 ng ml-1 for Cmax, 2064.0 ng h ml-1 for AUC(0-infinity), and 85.4 h t1/2. The mean pharmacokinetic parameters for norfluoxetine (in extensive metabolizers only) were 2532.0 ng h ml-1 for AUC(0-infinity) and 8.4 ng ml-1 for Cmax. For fluoxetine bioequivalence, the 90% CI of the individual ratio geometric mean for Psiquial/Prozac (including both extensive and poor metabolizers) was 101.6% to 121.1% for AUC(0-672 h) and 86.1% to 102.6% for Cmax. For norfluoxetine, the 90% CI of the individual ratio geometric mean for Psiquial/Prozac (including both extensive and poor metabolizers) was 90.3% to 108.3% for AUC(0-672 h) and 84.5% to 106.3% for Cmax. The new method developed (LC-MS) presented high sensitivity, specificity, and short chromatographic run for the quantification of both fluoxetine and norfluoxetine in human plasma. Since both 90% CI for AUC and Cmax geometric mean ratios were included in the 80% to 125% interval proposed by the U.S. Food and Drug Administration, Psiquial was considered bioequivalent to Prozac according to both the rate and extent of absorption. The finding that there were no significant differences in the bioequivalence assessed by either fluoxetine or norfluoxetine pharmacokinetic parameters indicates that future bioequivalence trials may be performed by quantifying fluoxetine only.

Heterocyclic amines, cytochrome P4501A2, and N-acetyltransferase: issues involved in incorporating putative genetic susceptibility markers into epidemiological studies

PURPOSE

Heterocyclic amines (HCAs), which are found mainly in well-cooked meat, require metabolic activation to function as mutagens and animal carcinogens. Enzymes such as cytochrome P4501A2 (CYP1A2) and N-acetyltransferase (NAT2) perform this task and are subject to interindividual variation. The source of this variation may be genetic, as in the case of NAT2, or both genetic and environmental as with CYP1A2. The present study examined the effect of HCAs on the NAT2 and CYP1A2 phenotypes in 33 males and 33 females.

METHODS

The subjects consumed a low HCA-containing diet for 1 week followed by a high HCA diet for the subsequent week. The subjects were phenotyped for CYP1A2 and NAT2 at the time of entry into the study (free-living), 1 week later (end of low-HCA or low-induction diet) and 2 weeks later (end of high-HCA or high-induction diet).

RESULTS

Consistent with genetic sources of variability, NAT2 showed little effect of a high-HCA diet and exhibited high intraindividual correlation. CYP1A2, in contrast, was induced by a high-HCA diet and exhibited a more modest intraindividual correlation.

CONCLUSIONS

Incorporating putative genetic susceptibility makers in population studies requires consideration of issues of induction and inhibition of metabolizing enzymes, and effects of covariates.

Erythrocyte thiopurine methyltransferase activity in a Korean population

Thiopurine methyltransferase (TPMT) is the enzyme responsible for the S-methylation of thiopurine drugs. The enzyme, present in human red blood cells (RBC), is known to exhibit genetic polymorphism and interethnic differences in its activity have been demonstrated. We have studied the role of TPMT polymorphism in Koreans and compared enzyme activity between this and other ethnic groups. In a population of 360 unrelated healthy Korean subjects TPMT activity showed a large interindividual variation ranging from 3.2 to 22.9 nmol ml-1 packed RBC h-1 with a median value of 12.0 and mode of 11.0 nmol ml-1 packed RBC h-1. The enzyme activity was higher in male subjects than that in female (median values; 12.2 vs 11.2, 95% confidence interval of the difference; -2.1, 4.0 nmol ml-1 packed RBC h-1). All subjects had detectable TPMT activity, but contrary to previous reports in other ethnic groups, this was distributed unimodally. The median RBC TPMT activity was very similar to values found in Caucasian populations, higher than in Floridian blacks and lower than that of a Norwegian Saami population.

Is there a gender difference in red blood cell thiopurine methyltransferase activity in healthy children?

We have examined red blood cell (RBC) thiopurine methyltransferase (TPMT) activity in a healthy population sample of Norwegian children, age 1-10 years. Boys had mean RBC TPMT activity of 11.1 +/- 2.0 U (n = 87) vs. 10.6 +/- 2.2 U (n = 71) in girls, the difference was not significant (P = 0.3). Age was negatively correlated to RBC TPMT activity (rs = -0.2, P = 0.01). As boys with acute lymphoblastic leukemia (ALL) tolerate more 6-mercaptopurine (6-MP) than girls and have a higher risk of relapse, we have searched for pharmacokinetic causes of these gender differences. The gender difference in 6-MP tolerance and clinical outcome in children with ALL cannot be explained by the minor and nonsignificant higher RBC TPMT activity in boys compared to girls.

Phenotypic debrisoquine 4-hydroxylase activity among extensive metabolizers is unrelated to genotype as determined by the Xba-I restriction fragment length polymorphism

1. The major pathway for 4-hydroxylation of debrisoquine in man is polymorphic and under genetic control. More than 90% of subjects (extensive metabolizers, EMs) have active debrisoquine 4-hydroxylase (cytochrome P450IID6) while in the remainder (poor metabolizers, PMs), cytochrome P450IID6 activity is greatly impaired. 2. Within the EM group, cytochrome P450IID6-mediated metabolism of a range of substrates varies widely. Some of this intra-phenotype non-uniformity may be explained by the presence of two subsets of subjects with different genotypes (heterozygotes and homozygotes). 3. Cytochrome P450IID6 substrates have not differentiated between these two genotypes. However, a restriction fragment length polymorphism (RFLP) which identifies mutant alleles of cytochrome P450IID6 locus has been described and can definitively assign genotype in some heterozygous EM subjects. 4. In this study, we used RFLP analysis and encainide as a model substrate to determine if non-uniformity in cytochrome P450IID6 activity among EMs is related to genotype. We tested the hypothesis that heterozygotes exhibit intermediate metabolic activity and that homozygous dominants exhibit the highest activity. We proposed encainide as a useful substrate for this purpose since cytochrome P450IID6 catalyzes not only its biotransformation to O-desmethyl encainide (ODE) but also the subsequent metabolism of ODE to 3-methoxy-O-desmethyl encainide (MODE). 5. A single 50 mg oral dose of encainide was administered to 139 normal volunteers and 14 PMs were identified. Urinary ratios among encainide, ODE and MODE in the remaining 125 EM subjects revealed a wide range of cytochrome P450IID6 activity. However, Southern blotting of genomic DNA digested with XbaI identified obligate heterozygotes in both extremes of all ratio distributions.(ABSTRACT TRUNCATED AT 250 WORDS)

A new, sensitive graphical method for detecting deviations from the normal distribution of drug responses: the NTV plot

1. A new graphical method was developed for the detection of deviations from the normal distribution. The approach took advantage of the similarity of graphical features of a graded dose-response relationship and a cumulative normal distribution. 2. The behaviour of the new normal test variable (NTV) plot was evaluated, in comparison with that of the probit plot and probability density functions (the generalization of histograms), for various assumed distributions. These included skewed distributions and composites of normal distributions with a variety of separations, ratios of peak sizes and widths. 3. The NTV approach generally detected deviations from the normal distribution more sensitively than the probit plot. 4. The NTV and probit plots may be able to identify biomadality by complementary approaches. 5. The characteristics of the three graphical representations were illustrated by a simulated sample from a composite of normal distributions and by an example of sparteine metabolism in 142 Cuna Amerindians.

Genetic polymorphism of sparteine/debrisoquine oxidation: a reappraisal

Polymorphic oxidation of the sparteine/debrisoquine-type has been shown to account for much of the interindividual variation in the metabolism, pharmacokinetics and pharmacodynamics of an increasing number of drugs, including some antiarrhythmic, antidepressant and beta-adrenoceptor antagonist agents. Impaired hydroxylation of these drugs results from the absence of the enzyme cytochrome P450IID6 in the livers of poor metabolisers, who constitute 6% to 10% of Caucasian populations. The clinical importance of the phenomenon has to be explored further and for most sparteine/debrisoquine-related substrates there is a need for controlled prospective studies to define the consequences to the patient of impaired or enhanced drug oxidation.

Unimodal distribution of the metabolic ratio for debrisoquine in Cuna Amerindians of Panama

1. The metabolic oxidation of debrisoquine (DB) was studied in 89 non-related Cuna Amerindian subjects. 2. Means and standard deviations for urinary recoveries of the intact drug and its 4-hydroxy metabolite (4-HD) were: %DB: 6.8 +/- 4.5; %4-HD: 16.0 +/- 9.1; %sum: 22.8 +/- 12.0. The log10 metabolic ratios for DB (LMRDB) were distributed within a single mode of insignificant skewness (-0.01, P greater than 0.10), which was unimodal (log kernel density and maximum likelihood methods) and normal (chi 2 = 22.5; d.f. = 15; P greater than 0.09; power of the test greater than 80%). 3. Therefore, no poor metabolizers (95% C.I.: 0.1%, 5.2%) were detected in the population sample studied.

Testing for bimodality in frequency distributions of data suggesting polymorphisms of drug metabolism--hypothesis testing

1. The theory of methods of hypothesis testing in relation to the detection of bimodality in density distributions is discussed. 2. Practical problems arising from these methods are outlined. 3. The power of three methods of hypothesis testing was compared using simulated data from bimodal distributions with varying separation between components. None of the methods could determine bimodality until the separation between components was 2 standard deviation units and could only do so reliably (greater than 90%) when the separation was as great as 4-6 standard deviation units. 4. The robustness of a parametric and a non-parametric method of hypothesis testing was compared using simulated unimodal distributions known to deviate markedly from normality. Both methods had a high frequency of falsely indicating bimodality with distributions where the components had markedly differing variances. 5. A further test of robustness using power transformation of data from a normal distribution showed that the algorithms could accurately determine unimodality only when the skew of the distribution was in the range 0-1.45.