A polymorphism in the alpha2a-adrenoceptor gene and endurance athlete status.

PURPOSE: In a case control study, we examined the allelic frequencies and genotype distributions of two restricted fragment length polymorphisms (RFLP) in the alpha-2A-adrenoceptor gene (ADRA2A) and beta-2-adrenoceptor gene (ADRB2) among elite endurance athletes (EEA) and sedentary controls (SC). METHODS: The EEA group included 148 Caucasian male subjects recruited on the basis that they had a VO2max > 74 mL O2 x kg(-1) x min(-1). The SC group comprised 149 unrelated sedentary male subjects, all Caucasians, from the Quebec Family Study. After digestion with the restriction enzymes Dra I (ADRA2A) and Ban I (ADRB2), Southern blotting and hybridization techniques were used to detect the mutations in the two ADR genes, which are encoded on chromosomes 10 (q24-26) and 5 (q31-32), respectively. RESULTS: For the Dra I ADRA2A RFLP, we observed a significant difference in genotype distributions between the two groups (P = 0.037). A higher frequency of the 6.7-kb allele was observed in the EEA group compared with the SC group (P = 0.013). No statistically significant difference was found between groups for the Ban I ADRB2 polymorphic site. Genotype frequencies for both genes in both groups were in Hardy-Weinberg equilibrium. CONCLUSIONS: In summary, we found evidence that ADRA2A gene variability detected with Dra I is weakly associated with elite endurance athlete status, and we conclude that genetic variation in the ADRA2A gene or a locus in close proximity may play a role in being able to sustain the endurance training regimen necessary to attain a high level of maximal aerobic power.

A mitochondrial DNA D-loop polymorphism and obesity in three cohorts of women.

OBJECTIVE: To examine the hypothesis of an association between a mtDNA D-loop Kpn I restriction site polymorphism (RSP) at base pair (bp) 16,133 (morph-1) and obesity in women. DESIGN: Comparisons of carriers and noncarriers of the mutation for BMI (Body Mass Index) levels and of the frequency of the mutation in obese and normal weight women. SUBJECTS: 567 unrelated adult Caucasian non-diabetic women from the HERITAGE Family Study (n = 63; BMI: 15-47 kg/m2), Quebec Family Study (QFS; 77 controls, BMI: 19-26 kg/m2 and 38 obese, BMI: 27-56 kg/m2) and Swedish Obese Subjects (SOS) Study (81 controls, BMI: 18-26 kg/m2 and 308 obese, BMI: 33-58 kg/m2). MEASUREMENTS: BMI was calculated from weight and height (kg/m2). mtDNA was amplified between base pair 15,928 and 16,500 by polymerase chain reaction (PCR) and digested with the restriction endonuclease Kpn I. RESULTS: No significant differences in the age-adjusted BMI for the mtDNA D-loop Kpn I RSP at base pair (bp) 16,133 (morph-1) between carriers and non-carriers in the HERITAGE cohort. No significant association was found between BMI and the Kpn I RSP carrier status in the SOS and QFS cohorts. The observed frequencies for the Kpn I RSP were not significantly (P > 0.05) different between the SOS controls and SOS obese irrespective of the degree of severity of obesity (BMI > 40, > 45 or > 50 kg/m2). CONCLUSION: We conclude that the mtDNA D-loop Kpn I RSP at bp 16,133 (morph-1) is not a determinant of human obesity.

Linkage between a muscle-specific CK gene marker and VO2max in the HERITAGE Family Study.

PURPOSE: We have reported a significant association between VO2max in the sedentary state and its response (delta VO2max) to an endurance training program with a muscle-specific creatine kinase (CKMM) gene polymorphism. The purpose of this study was to test the hypothesis of genetic linkage between the same CKMM marker and VO2max in the sedentary state as well as delta VO2max. METHODS: Sib-pair linkage analysis was performed on 277 full sib-pairs from 98 Caucasian nuclear families of the HERITAGE Family Study. VO2max was measured during cycle ergometry tests before and after 20 wk of endurance training. The CKMM polymorphism was detected by the polymerase chain reaction and digestion with the Ncol restriction enzyme. RESULTS: Frequencies for the rare (1170 base pairs) and common (985 + 185 base pairs) alleles were 0.32 and 0.68, respectively. No significant linkage (t = -0.02, P = 0.49) was detected between the CKMM marker and the age and sex adjusted VO2max (mL x kg(-1) x min(-1)) in the sedentary state. However, after adjustment of delta VO2max for the effects of age, sex, initial VO2max, and body mass, evidence for linkage between the CKMM locus and delta VO2max was suggestive (P = 0.04). CONCLUSION: The present results provide further support for the notion that the CKMM gene, or some gene in close linkage disequilibrium with it, may contribute to individual differences in the VO2max response to endurance training.

Three mitochondrial DNA restriction polymorphisms in elite endurance athletes and sedentary controls.

This study examined the associations between elite endurance athlete (EEA) status and three mitochondrial DNA (mtDNA) restriction fragment length polymorphisms (RFLPs) in the subunit 5 of the NADH dehydrogenase (MTND5) locus and one in the D-loop region. A group of 125 Caucasian male EEA well endowed with the phenotypic expression of VO2max (78.9 +/- 3.8 mL x kg(-1) x min(-1), mean +/- SD) and 65 sedentary controls (SCON: VO2max = 39.8 +/- 8.2 mL x kg(-1) x min(-1)) participated in the study. VO2max was determined during an incremental exercise test on a cycle ergometer or a motor-driven treadmill. mtDNA was extracted from white blood cells or lymphoblastoid cell lines and specific regions were amplified by the polymerase chain reaction. The Pearson Chi-square statistic test and Fisher exact test revealed no significant association (P > 0.05) between any of the three mtDNA RFLPs and EEA status. The MTND5-BamHI RFLP at bp 13,470 (morph 3) was found in 12.8% of the EEA and 12.3% of the SCON (chi2 = 0.009, P = 0.92). The prevalence of the MTND5-Ncil RFLP at bp 13,364 (morph 2) was 12.9% and 14% for the EEA and SCON, respectively (chi2 = 0.043, P = 0.83). The D-loop-KpnI RFLP at bp 16,133 (morph 1) was found in 5.8% of the EEA and in 1.6% of the SCON (Fisher exact test = 1.80, P = 0.18). The MTND5-HincII RFLP at bp 12,406 (morph 1) was not present in this study sample. These results indicate no evidence for a difference in the frequency of two polymorphic restriction sites in the subunit 5 of the NADH dehydrogenase gene of mtDNA and one in the D-loop region between elite endurance athletes and sedentary controls.

Muscle-specific creatine kinase gene polymorphisms in elite endurance athletes and sedentary controls.

The purpose of this study was to investigate the association between elite endurance athlete (EEA) status and two restriction fragment length polymorphisms (RFLPs) at the muscle-specific creatine kinase (CKMM) gene locus. Genomic DNA was extracted from white blood cells or lymphoblastoid cell lines of 124 unrelated Caucasian male EEA (VO2max > 73 mL.kg-1.min-1) and 115 unrelated Caucasian sedentary male controls (SCON). The genetic polymorphism at the CKMM locus was detected by the polymerase chain reaction and DNA digestion with the NcoI and TaqI restriction endonucleases. The allelic frequencies for the NcoI and TaqI RFLPs were not different (P > 0.05) between EEA and SCON subjects. The three expected genotypes for CKMM-NcoI (1170/1170 bp, 1170/985 + 185 bp, and 985 + 185/985 + 185 bp) and CKMM-TaqI (1170/1170 bp, 1170/1020 + 150 bp, and 1020 + 50/1020 + 150 bp) were observed in the EEA and SCON groups. These genotype frequencies were in Hardy-Weinberg equilibrium, but they were not significantly (P > 0.05) different between the EEA and SCON. A strong (P < 0.001) linkage disequilibrium was detected among the NcoI and TaqI RFLPs in both EEA and SCON. These findings indicate that the skeletal muscle CK-NcoI and CK-TaqI gene polymorphisms are not associated with the elite endurance athlete status.

Muscle-specific creatine kinase gene polymorphism and VO2max in the HERITAGE Family Study.

This study examined the association between a DNA polymorphism in the muscle-specific creatine kinase (CKMM) gene and VO2max in the sedentary state, as well as its response (deltaVO2max) to a standardized 20-wk endurance training program. The subjects were 160 biologically unrelated Caucasian parents (80 women, 80 men) and 80 biologically unrelated adult offspring of the HERITAGE Family Study. The CKMM polymorphism was detected by PCR and digestion with the NcoI restriction enzyme. VO2max was measured during maximal cycle ergometer tests. VO2max was 2119 +/- 45 mL x min(-1) (mean +/- SE) or 26 +/- 0.4 mL x kg(-1) x min(-1). Both sexes had a significant (P < 0.05) increase in the deltaVO2max (women = 283 +/- 20 mL x min[-1] and men = 363 +/- 25 mL x min[-1]). Allele and genotype frequencies were not significantly different (P > 0.05) between sexes. Age and sex adjusted VO2max was significantly (P = 0.007) associated with the CKMM genotype in the parents, whereas no association (P > 0.05) was observed in the offspring. DeltaVO2max values adjusted for age, sex, VO2max, and body mass were characterized by genotype differences in both parents (P = 0.0004) and offspring (P = 0.0025). A significantly (P < 0.05) lower deltaVO2max to endurance training was detected in both parents and offspring homozygotes for the rare allele. The genotype accounted for at least 9% of the variance in deltaVO2max. These results indicate that the NcoI polymorphism in the 3' untranslated region of the muscle-specific creatine kinase gene is associated with the deltaVO2max to endurance training.

Identification of an obesity quantitative trait locus on mouse chromosome 2 and evidence of linkage to body fat and insulin on the human homologous region 20q.

Chromosomal synteny between the mouse model and humans was used to map a gene for the complex trait of obesity. Analysis of NZB/BINJ x SM/J intercross mice located a quantitative trait locus (QTL) for obesity on distal mouse chromosome 2, in a region syntenic with a large region of human chromosome 20, showing linkage to percent body fat (likelihood of the odds [LOD] score 3.6) and fat mass (LOD score 4.3). The QTL was confirmed in a congenic mouse strain. To test whether the QTL contributes to human obesity, we studied linkage between markers located within a 52-cM region extending from 20p12 to 20q13.3 and measures of obesity in 650 French Canadian subjects from 152 pedigrees participating in the Quebec Family Study. Sib-pair analysis based on a maximum of 258 sib pairs revealed suggestive linkages between the percentage of body fat (P < 0.004), body mass index (P < 0.008), and fasting insulin (P < 0.0005) and a locus extending approximately from ADA (the adenosine deaminase gene) to MC3R (the melanocortin 3 receptor gene). These data provide evidence that a locus on human chromosome 20q contributes to body fat and insulin in a human population, and demonstrate the utility of using interspecies syntenic relationships to find relevant disease loci in humans.

Suggestive linkages between markers on human 1p32-p22 and body fat and insulin levels in the Quebec Family Study.

A single-gene rodent mutation (diabetes) and a quantitative trait locus (dietary obese 1) mapped to the mid portion of mouse chromosome 4 have been related to obesity and/or insulin levels. Synteny relationships place their putative human homologs on 1p31 and 1p35-p31, respectively. In 137 sibships of adult brothers and sisters from the Québec Family Study, genetic linkages between seven microsatellite markers from 1p32-p22 and various obesity- and diabetes-related quantitative phenotypes were examined using single locus sibpair linkage analysis. Suggestive linkages were observed between markers D1S476 and body mass index (p = 0.05), fat mass (p = 0.02), the sum of six skinfolds (p = 0.02), the insulin area after an oral glucose tolerance test (p = 0.02), and between the neighboring marker D1S200 and body mass index (p = 0.03), and fat mass (p = 0.009). Suggestive linkages were also observed between the more telomeric markers D1S193 and body mass index (p = 0.03), and between the neighboring marker D1S197 and fasting insulin level (p = 0.05). No linkage was observed with the trunk to extremity skinfolds ratio. These linkages suggest that human homologs of the mouse diabetes or dietary obese 1 and/or other genes in this interval on chromosome 1 play a role in the regulation of body mass, body composition, and insulin levels, but not of subcutaneous fat distribution.

The human obesity gene map: the 1996 update.

An update of the human obesity gene map up to October 1996 is presented. Evidence from Mendelian disorders exhibiting obesity as a clinical feature, single-gene mutation rodent models, quantitative trait loci uncovered in crossbreeding experiments with mouse, rat, and pig models, association and case-control studies with candidate genes, and linkage studies with genes and other markers is reviewed. All chromosomal locations of the animal loci are converted into human genome locations based on syntenic relationships between the genomes. A complete listing of all these loci reveals that only 4 of the 24 human chromosomes are not yet represented, i.e., 9, 18, 21, and Y. Several chromosome arms are characterized by the presence of several putative loci. The following arms include at least three such loci: 1p, 1q, 3p, 4q, 6p, 7q, 8p, 8q, 11p, 11q, 15q, 20q, and Xq. Studies with negative association and linkage results are also reviewed.

The Trp64Arg mutation of the beta3 adrenergic receptor gene has no effect on obesity phenotypes in the Québec Family Study and Swedish Obese Subjects cohorts.

The beta adrenergic system plays a key role in regulating energy balance through the stimulation of both thermogenesis and lipid mobilization in brown and white adipose tissues in human and various animal models. Recent studies have suggested that a missense Trp64Arg mutation in the beta3 adrenergic receptor (ADRB3) gene was involved in obesity and insulin resistance. We have investigated the effect of this mutation on obesity-related phenotypes in two cohorts: the Québec Family Study (QFS) and the Swedish Obese Subjects (SOS). In QFS, no association was found between this mutation and body mass index (BMI), body fat including abdominal visceral fat, resting metabolic rate, various diabetes and cardiovascular risk factors, and changes in body weight and body fat over a 12-yr period. With the exception of RMR (P = 0.04), no evidence of linkage was detected between the mutation and phenotypes of QFS based on sib-pair data. In SOS, the frequency of the Trp64Arg allele was not significantly different between nonobese and obese female subjects and no association was found between the mutation and body weight gain over time. These findings do not support the view that there is an association between the Trp64Arg mutation in the ADRB3 gene and obesity.

The apoB-100 gene EcoRI polymorphism influences the relationship between features of the insulin resistance syndrome and the hyper-apoB and dense LDL phenotype in men.

The aim of this study was to investigate whether the EcoRI polymorphism of the apolipoprotein B (apoB) gene influences the relationships between features of the insulin resistance syndrome and the dense LDL phenotype and apoB concentrations. A sample of 65 men was divided into two groups on the basis of the EcoRI genotype. Forty-four subjects were (+/+) homozygotes for the presence of the EcoRI restriction site that is associated with a glutamic acid at codon 4154. Twenty-one men were (+/-) heterozygotes for the absence of the restriction site resulting from a glutamic acid to a lysine substitution at codon 4154. In the (+/-) group, fasting plasma FFA levels were positively correlated with plasma apoB, LDL-apoB, and the LDL particle score that was calculated from the migration distances of LDL subspecies and their relative band intensities, reflecting the proportion of small dense LDL particles. However, these associations were not found among (+/+) subjects. The two genotypic groups were further divided into two subgroups on the basis of fasting FFA concentrations, and the LDL particle score and the LDL-apoB levels were compared. High FFA levels were associated with a higher proportion of small dense LDL particles, as reflected by a higher mean LDL particle score, irrespective of the genotype. However, the apoB-EcoRI polymorphism appeared to influence the association between high FFA levels and LDL-apoB concentrations because (+/-) heterozygotes with high FFA levels had higher LDL-apoB concentrations than (+/-) heterozygotes with low FFA levels. In addition, the integrated area under the curve of plasma insulin concentrations, measured in response to a 75-g oral glucose challenge, and the amount of visceral adipose tissue, measured by computed tomography, were positively associated with the LDL particle score only in (+/-) heterozygotes. When subjects were divided on the basis of insulin area (low vs. high) or visceral adipose tissue (low vs. high), (+/-) heterozygotes with high insulin area or with high levels of visceral adipose tissue had a higher mean LDL particle score than (+/-) heterozygotes with low insulin area or low visceral adipose tissue. However, among (+/+) homozygotes, low or high levels of insulin or visceral adipose tissue could not discriminate between men with large or small LDL particles. Therefore, (+/-) heterozygotes may be more susceptible to develop the dense LDL phenotype in presence of hyperinsulinemia and visceral obesity. Results of the present study suggest that the apoB-EcoRI polymorphism may exacerbate the alterations in the LDL particle (size and concentration) found among visceral obese-hyperinsulinemic men.

The Bcl I polymorphism of the human uncoupling protein (ucp) gene is due to a point mutation in the 5'-flanking region.

The polymorphic Bcl I site in the human ucp gene associated to percentage fat gain over time in the Québec Family Study cohort (Oppert et al. Int J Obesity 1994; 18: 526-531) has been positioned to the 5'-flanking region. This polymorphism results from a unique A/G mutation. Oligonucleotides used to amplify the polymorphic region, and allowing future studies of any cohort of patients, are described.

DNA polymorphisms in the alpha 2- and beta 2-adrenoceptor genes and regional fat distribution in humans: association and linkage studies.

The aim of this study was to investigate the relationships between DNA restriction fragment length polymorphisms (RFLP) in the alpha 2- and beta 2-adrenoceptor genes and body fat distribution in humans. Skinfold thickness measurements and genetic analyses (Southern blot) were performed on 280 individuals (142 parents and 138 offsprings) from the Québec Family Study. Using the association study design in unrelated adults, women but not men carrying the 6.3-kb allele of an alpha 2A-adrenoceptor/DraI RFLP had a significantly higher trunk to extremity skinfold ratio (= sum of subscapular+suprailiac+abdominal skinfolds/sum of biceps+triceps+medial calf skinfolds) compared to women without the allele (1.44 +/- 0.52 vs. 1.12 +/- 0.33; p < 0.005 after adjustment for age, p < 0.002 after adjustment for age and body mass index or for age and subcutaneous fat). Using the sib-pair linkage procedure, a significant inverse relationship was found between the proportion of alleles identical by descent shared by sibs at the alpha 2A RFLP marker locus and the squared differences of the trunk to extremity skinfold ratio (p = 0.02 after adjustment for age or for age and body mass index or for age and subcutaneous fat). For a beta 2-adrenoceptor/BanI RFLP, no significant association or linkage was found between fat distribution indicators and the marker. These results suggest that alpha 2A-adrenoceptor gene variability detected with DraI is associated with a relative subcutaneous fat pattern favoring accumulation of truncal-abdominal fat in women, and that the alpha 2A-adrenoceptor gene, or a locus in close proximity, may be linked to body fat distribution in humans independently of the overall level of fatness.

Relation between BglII polymorphism in 3beta-hydroxysteroid dehydrogenase gene and adipose tissue distribution in humans.

The aim of this study was to investigate the association between a restriction fragment length polymorphism (RFLP) at the 3beta-hydroxysteroid dehydrogenase locus and adipose tissue distribution phenotypes. A total of 132 unrelated individuals from the Quebec Family Study were followed prospectively for an average period of 11.3 years. The BglII polymorphism in exon 4 of the 3beta-HSD gene was detected by PCR. Body mass, body fat, and regional fat distribution indicators were adjusted for age and age2 within each gender. Associations were assessed in unrelated adults with ANOVA across three genotypes. No association was found for the indicators of body mass, body fat, and regional distribution of adipose tissue measured in 1992. In women, the changes (difference between data collected in 1992 and at entry) in the sum of six skinfolds (p=0.04), abdominal skinfold (p=0.01), and abdominal skinfold adjusted (p=0.03) for the sum of six skinfolds at entry were related to the BglII polymorphism at the 3beta-HSD locus. These relations were not found in men, but they gained less body mass and body fat over the 11.3-year period. This suggests that sequence variation at the 3beta-HSD locus or in neighboring genes on chromosome 1 may contribute to individual differences in body fat content and adipose tissue distribution in adult women, particularly in abdominal adipose tissue deposition as they grow older and gain body fat.

DNA polymorphism in the uncoupling protein (UCP) gene and human body fat.

The objective of this study was to identify DNA sequence variation in the UCP gene and to investigate its relationship with some obesity phenotypes. Two studies were carried out: (1) association study in unrelated subjects, and (2) sib-pair linkage analysis study in brothers and sisters. The subjects were 261 individuals from the Québec Family Study (123 parents and 138 offsprings from 64 families). The following were measured: Body mass index, percent body fat (measured by hydrostatic weighing), and subcutaneous fat (estimated by the sum of 6 skinfolds) were measured in 1978-81 and again 12 years later. Resting metabolic rate (RMR) was measured only in 1989-93. Genetic analyses were performed using Southern blotting technique and a human UCP genomic probe. (1) A BcII restriction fragment length polymorphism was identified with two alleles of 8.3 and 4.5 kb in length, and respective frequencies of 0.28 and 0.72. (2) In unrelated adults from the parental generation, a cross-sectional analysis of the 1989-93 data showed no difference in body fat and RMR between the UCP genotypes. No significant difference for the absolute changes in body fat over the 12-year period among the UCP genotypes was observed. However, a higher frequency (P < 0.05) of the 8.3-kb allele was found in high gainers compared to low gainers (i.e., above and below the median value) for percent body fat over the 12-year period. (3) No evidence of linkage between any of the obesity phenotypes and the UCP BcII marker was found. For the first time, the presence of DNA polymorphism in the human UCP gene is reported. Although, no significant association and linkage were found between the UCP BcII gene marker and body fat in the cohort of the Quebec Family Study, a higher frequency of the 8.3-kb allele was found in individuals who gained more body fat over time.

ApoB-100 gene EcoRI polymorphism. Relations to plasma lipoprotein changes associated with abdominal visceral obesity.

The aim of this study was to investigate whether the EcoRI restriction fragment length polymorphism (RFLP) of the apolipoprotein (apo) B-100 gene influences the associations described among obesity, regional adipose tissue distribution, and plasma lipoprotein levels. For this purpose, blood samples were collected from 56 healthy men for whom we had extensive measurements of regional adipose tissue distribution (both anthropometric and computed tomography-derived measurements) and data on the plasma lipoprotein-lipid profile. DNA was extracted from white blood cells, and RFLP analysis was performed. Subjects were classified into two groups on the basis of their apoB-100 EcoRI genotype: subjects homozygous for the major 11-kb allele, the 11/11 group (n = 40), and subjects carrying the minor 13-kb allele, the 13/11 group (n = 16). Subjects carrying the 13-kb allele had lower percent body fat and abdominal adipose tissue accumulation than subjects homozygous for the 11-kb allele (P < .05). Although leaner, the 13/11 group did not show a more favorable plasma lipoprotein-lipid profile than the group homozygous for the 11-kb allele. In fact, after statistical control for the difference in percent body fat between the two genotype groups, the 13/11 group showed significantly higher plasma cholesterol levels (P < .05) and nearly significantly higher apoB levels than the 11/11 group (P = .06). The association patterns between indices of regional adiposity and plasma cholesterol and apoB levels were also different between the two EcoRI genotype groups. Only in the 13/11 group was the abdominal visceral adipose tissue area significantly associated with these plasma variables.(ABSTRACT TRUNCATED AT 250 WORDS)

Nuclear-encoded subunits of human cytochrome c oxidase: SstI restriction fragment length polymorphism.

A DNA polymorphism of the nuclear-encoded subunit Va of the human cytochrome c oxidase (COX), a mitochondrial respiratory enzyme, is reported. No polymorphism was detected in genes for the subunits IV and Vb of the same enzyme.

Detection of a MspI restriction fragment length polymorphism for the human sex hormone-binding globulin (SHBG) gene.

A rare polymorphism in the human sex hormone-binding globulin (SHBG) gene was detected using a human SHBG cDNA probe. It is the first DNA sequence variation reported in this gene.

Mitochondrial DNA sequence polymorphism, VO2max, and response to endurance training.

Mitochondrial DNA sequence variation was determined in 46 sedentary young adult males who took part in ergocycle endurance training programs in two laboratories to assess whether mitochondrial DNA variants were associated with individual differences in maximal oxygen uptake (VO2max) and its response to training. VO2max was obtained from a progressive ergocycle test to exhaustion. White blood cell mitochondrial DNA was characterized with the restriction fragment length polymorphism (RFLP) technique using 22 restriction enzymes and human mitochondrial DNA as a probe for hybridization. Multiple mitochondrial DNA variants were detected with 15 of the enzymes. Some subjects exhibited many RFLPs, while others showed no variation. These RFLPs (morphs) were generated by base substitutions located in gene regions coding for mitochondrial proteins as well as in the noncoding regions. Carriers of three mitochondrial DNA morphs, two in the subunit 5 of the NADH dehydrogenase gene and one in the tRNA for threonine, had a VO2max (ml.kg-1.min-1) in the untrained state significantly higher than noncarriers, while carriers of one mitochondrial DNA morph in subunit 2 of NADH dehydrogenase had a lower initial VO2max. Endurance training increased VO2max by a mean of 0.5 l of O2, with individual differences ranging from gains of 0.06 to 1.03. After adjustment for training site and initial VO2max, a lower response was observed for three carriers of a variant in subunit 5 of the NADH dehydrogenase detected with HincII (mean gain of 0.28 l; P < 0.05). These results suggest that sequence variation in mitochondrial DNA may contribute to individual difference in VO2max and its response to training.