Complex adaptive systems and human health: the influence of common genotypes of the apolipoprotein E (ApoE) gene polymorphism and age on the relational order within a field of lipid metabolism traits.

We analyzed the influence of age, apolipoprotein E (ApoE) genotype, and their interaction on the variation of each of all possible pairwise correlations among plasma levels of ApoE, ApoB, total cholesterol, triglyceride, and HDL cholesterol. Our cross-sectional study sample included 1,876 individuals (979 females and 897 males) from the Rochester, MN population, unselected for health, with a common ApoE genotype of epsilon32, epsilon33, or epsilon43, and ranging in age from 5 to 90 years. We conducted analyses on data from female and male subjects separately, using a hierarchical set of generalized additive models. The age changes in the correlations were estimated using a 30-year sliding window across the age range. There were qualitative differences between genders in the age at which the peaks in the correlations occurred. For female subjects, peaks in correlations were mostly in the middle and older age windows, whereas in males, peaks were mostly in the younger and middle age windows. We found for both genders that for each of the possible pairwise correlations, the influence of age was significantly dependent on ApoE genotype (all Pr<0.0001). We also found for female and male subjects that the epsilon32- and epsilon43- specific age changes in the correlations were each significantly different from those for the epsilon33 genotype (Pr<0.0001), with two exceptions for males (marginally significant differences, P<0.08). We conclude that the influence of ApoE genotypic variation extends far beyond the levels of the gene product, to the dynamics of the relational order among measures of lipid metabolism with age. Moreover, age and common ApoE genotype are not independent predictors of the gender-specific changes in relational order that we observed among these measures of lipid metabolism. These results have implications for the development and application of therapeutic approaches to treat human disease and our enhanced understanding of the role of genetic variation in the dynamic actions of complex adaptive systems with age that occur in response to environmental change. These dynamic actions emerge as the phenotypes that are measures of human health in the population at large.

Genetic structure of five susceptibility gene regions for coronary artery disease: disequilibria within and among regions.

We analyzed two-locus disequilibria for 16 polymorphic loci of seven susceptibility genes for coronary artery disease located in five chromosomal regions distributed across four chromosomes. Included were the genes coding for apolipoprotein B (ApoB, chromosome 2, four marker loci), lipoprotein lipase (LPL, chromosome 8, three marker loci), apolipoproteins AI, CIII AIV (ApoAI-CIII-AIV, chromosome 11, three marker loci), apolipoprotein E (ApoE, chromosome 19, two marker loci), and the low density lipoprotein receptor (LDLR, chromosome 19, four marker loci). Our sample included 540 unrelated individuals from the Rochester, Minn. population. There were no statistically significant deviations of single-locus genotypes from Hardy-Weinberg equilibrium. The strongest associations within genes were for composite diallelic disequilibria; 17/19 were significant (13 at Pr <0.001, 1 at Pr <0.01, 3 at Pr <0.05). These observations suggest marker alleles within genes have a shared evolutionary history reflected by disequilibria that have not been dissipated by recombination. Disequilibrium was not generally concordant with the physical orderings of markers. Only two significant higher-order disequilibria were observed although 12 triallelic disequilibria were at maximum possible values. We observed 19 statistically significant disequilibria (Pr <0.05; 4 composite diallelic, 13 triallelic, and 2 quadriallelic) between 101 pairs of marker loci, where each locus in a pair was from a different unlinked region. These unexpected results are most likely explained by recent historical factors, including worldwide population expansion and amalgamation with continuous admixture, that influence the genetic structure (organization of alleles and non-alleles into genotypes) of a population. We conclude that disequilibria between loci from unlinked regions may be more extensive than is commonly assumed. Our findings also suggest that it is, on average, at least 15 times more likely to not detect significant disequilibrium among unlinked loci when it is really present than to make a false positive inference. Disequilibria between functional loci within or between regions will impact estimates of genetic variance associated with particular functional mutations within a susceptibility gene region.

A biometrical study of the relationship between sodium-lithium countertransport and triglycerides.

We addressed the question: Is there evidence that allelic variation in a single unmeasured gene that has a large effect on maximal activity of erythrocyte sodium-lithium countertransport (Na-Li CNT) also has pleiotropic effects on variation in plasma triglyceride levels? Complex segregation analysis models that included plasma triglyceride levels as a covariate were considered as explanations for interindividual variation in Na-Li CNT. A sample of 711 healthy adults from 254 pedigrees enrolled in the Rochester Family Heart Study was selected for this study. The majority of the pedigrees supported the hypothesis that variations in a single unmeasured non-transmitted environmental factor have large effects on the Na-Li CNT distribution. Only gender-specific first-order covariate parameters were necessary in the complex segregation models suggesting that the form of the relationship between Na-Li CNT and plasma triglyceride level was not influenced by variation in the inferred environmental factor with large effects. Stratification of the sample by this inferred environmental factor resulted in three classes of individuals with significant differences in the distributions of coronary heart disease risk factor traits, as well as interindividual variation in both Na-Li CNT and plasma triglyceride levels. These results, along with other observations from the Rochester Family Heart Study sample, emphasize the complex and multifactorial nature of the causes of interindividual variation in Na-Li CNT. Our study further suggests that new research strategies are needed for studying the relationships between genetic and environmental variation and variation in quantitative traits such as Na-Li CNT that have been identified as risk factors for hypertension.

Genotype-environment interaction: apolipoprotein E (ApoE) gene effects and age as an index of time and spatial context in the human.

We analyzed the age-dependence of the estimates of the parameters of the genetic architecture of plasma ApoE levels associated with ApoE gene variation. Our study sample included 1988 individuals in multigeneration pedigrees from the Rochester, MN, population. We used a 30-yr sliding window across the age range (5-90 yr) to estimate the age dependency of parameters. Additive ApoE allelic variance of transformed plasma ApoE values for both genders, heritabilities for males and phenotypic and residual variance for females peaked in the 20-40-yr age windows and decreased significantly with age (P < 0.05). Phenotypic and residual variance for males and dominance variance for both genders did not vary significantly with age. All parameter estimates were significantly different from zero across all age windows for both genders. Most studies of ApoE have focused on its functions in the pathophysiology of coronary artery disease (CAD) in middle-aged and older individuals. Our findings suggest the greatest role of this gene is in determining phenotype differences among younger and middle-aged individuals. These observed genotypic effects on the plasma ApoE levels may contribute to age-dependent differences in physiological health, growth, and risk of disease.

Apolipoprotein E polymorphism predicts death from coronary heart disease in a longitudinal study of elderly Finnish men.

BACKGROUND: There is ample evidence from cross-sectional studies of an association between allelic variation of the gene coding for apolipoprotein E (apoE) and interindividual variation in plasma lipids, and the presence of coronary heart disease (CHD). There have been no prospective studies, however, to evaluate the usefulness of allelic variation of the apoE gene for predicting CHD. METHODS AND RESULTS: Two samples of elderly Finnish men were followed for 5 years, one in the east (n = 297) and the other in the southwest of Finland (n = 369). At baseline, when the apoE genotypes were assessed, the men were 65 to 84 years old. At the end of the follow-up, the vital status of each man was determined, and cause of death was coded. At baseline, relative frequencies of the three alleles-epsilon 2, epsilon 3, and epsilon 4--were 0.037, 0.827, and 0.136 in the eastern and 0.062, 0.763, and 0.175 in the southwestern samples, respectively (chi 2 = 8.89, df = 2, P < .012 for difference between the samples). During the 5-year follow-up, a total of 28 deaths from CHD were recorded in the eastern and 42 in the southwestern sample. Relative CHD mortality was not heterogeneous between the samples. Among those who died from CHD, there was a doubling of the relative epsilon 4 allele frequency in both samples (chi 2 = 4.70, df = 1, P < .03 for the eastern sample; chi 2 = 7.11, df = 1, P < .01 for the southwestern sample). CONCLUSIONS: Allelic variation in the apoE gene is a statistically significant predictor of CHD death in these samples of elderly Finnish men.

Traversing the biological complexity in the hierarchy between genome and CAD endpoints in the population at large.

An emerging challenge facing those who are concerned about the efficacy of public health programs is to understand how information from the DNA revolution might be used to improve our ability to predict the initiation, progression and severity of a common disease having a complex multifactorial etiology. In the course of research to evaluate the role of information about DNA, combinations of genome types and environmental exposures that predispose to disease will be identified. Such information is expected to be useful in efforts to identify individuals and families at higher risk of disease and to predict their responses to a proposed therapy. This paper begins with a discussion of the features of a realistic biological model for the study of a common multifactorial disease. We present evidence for the complexity in the relationship between genome type variation and variation in risk of coronary artery disease (CAD) and review the preliminary results of our studies to determine whether information about genome type variation can improve our ability to predict the distribution of CAD among individuals in the population at large. Such studies make it apparent that new analytical strategies are necessary to deal with the plethora of genome type information available for the evaluation of risk of a common disease like CAD. This shift in the research paradigm will build upon new strategies to understand the organization of natural systems that are coming from outside the mainstream of genetic research.

Biological complexity and strategies for finding DNA variations responsible for inter-individual variation in risk of a common chronic disease, coronary artery disease.

Most common chronic diseases of humans aggregate, but do not segregate, in families. The segregation-linkage research paradigm has not provided great insights into their genetic etiology. In this paper, using coronary artery disease as an example, we discuss hierarchical organization, coherence, emergent properties and dynamism as features that characterize the complexity of genotype-phenotype relationships. We summarize a research strategy for evaluating the contribution of genetic and environmental factors to the prediction of inter-individual variation in risk of disease. We then review a statistical strategy that employs cladistic theory to identify individuals carrying mutant DNA sequences responsible for an observed association between marker variation in a gene and inter-individual variation in biological traits that determine risk of a common multifactorial disease. Finding these DNA sequences is a necessary step in our search for an understanding of the nature of the mapping of genetic variation into variation in risk of a disease like coronary artery disease.

Application of cladistics to the analysis of genotype-phenotype relationships.

We seek to understand the relative contribution of allelic variations of a particular gene to the determination of an individual's risk of atherosclerosis or hypertension. Work in progress is focusing on the identification and characterization of mutations in candidate genes that are known to be involved in determining the phenotypic expression of intermediate biochemical and physiological traits that are in the pathway of causation between genetic variation and variation in risk of disease. The statistical strategy described in this paper is designed to aid geneticists and molecular biologists in their search to find the DNA sequences responsible for the genetic component of variation in these traits. With this information we will have a more complete understanding of the nature of the organization of the genetic variation responsible for quantitative variation in risk of disease. It will then be possible to fully evaluate the utility of measured genetic information in predicting the risk of common diseases having a complex multifactorial etiology, such as atherosclerosis and hypertension.

Genetic structure and the search for genotype-phenotype relationships: an example from disequilibrium in the Apo B gene region.

We analyzed allelic associations (disequilibria) for four restriction fragment length polymorphisms (RFLPs) in the region of the 43-kb Apo B gene in a sample of 233 unrelated individuals from Montreal, Canada, sampled for health. This total sample (T) included 160 individuals of known French Canadian (FC) ancestry. We present a rigorous application of current methodology to these samples, including estimation of type II error probabilities and correlations between markers for estimates of disequilibria. We then consider the utility of these estimates of allelic disequilibria for the interpretation of genotype-phenotype relations. Significant deviations from Hardy-Weinberg equilibrium were not predicted by proximity to other markers in disequilibrium. We found significant quadri-allelic disequilibrium for two marker pairs despite absence of significant deviations from Hardy-Weinberg equilibrium for either marker or tri-allelic disequilibrium, respectively. Altogether these results underscore the complexity of the genotypic structure of the data. A combination of nonevolutionary factors, including sampling for health, small sample size and data exclusion due to methodological constraints of not successfully typing all members of the sample for every RFLP, is a likely explanation for this complexity. These types of factors are common to many RFLP studies. Patterns of composite di-allelic disequilibrium indicated that some RFLP allele pairs may have a longer shared evolutionary history than others and that disequilibrium is not predicted by distance between RFLPs. Type II error probabilities were generally much higher than those for type I errors. Correlations between marker pairs for disequilibria were generally not high. We show from a review of 14 published studies of association between the XbaI RFLP and variation in a total of 15 lipid traits that deviations from Hardy-Weinberg equilibrium can cause substantial differences in the estimation of variability associated with phenotypic differences among marker genotypes relative to Hardy-Weinberg conditions.

Heterogeneity of the blood pressure distribution among Solomon Islands societies with increasing acculturation.

This study illustrates the very complex nature of gene by environmental interactions influencing the blood pressure (BP) distribution in a series of genetically distinctive populations undergoing rapid acculturation. We report the results of two BP and anthropometric surveys on Solomon Islands societies separated by an interval of 14 to 19 years. While differences in acculturation existed at the time of the initial survey, the interval between surveys was marked by rapid acculturation in almost all societies. Seven of the eight societies originally covered were included in the resurvey, and a large but variable proportion of the original sample subjects was recovered in the follow-up. Because the genetic relationships of the societies have been described, we were able to establish the following points concerning the role of genetic differences in determining the distribution of BP among these populations and, more important, the interaction of these genetic differences with changes associated with increasing acculturation: 1) In the initial survey, mean adjusted systolic and diastolic BPs were significantly heterogeneous among societies within and among genetically related clusters of societies (genetic clusters) and sexes. At the same time, rank differences in these means were not associated with rank differences in acculturation status among societies ignoring cluster membership. 2) Importantly, in the follow-up survey increasing acculturation resulted in the disappearance of significant differences in mean systolic and diastolic BP among genetic clusters in males, despite continued significant heterogeneity among societies within genetic clusters. In females, differences among genetic clusters persisted, but the degree of significance was substantially less with increasing acculturation. We interpret these changes as evidence for genotype by environment interaction. 3) There were significant differences in interindividual variances of both systolic and diastolic BPs among genetic clusters in the first survey. Ranks of these variances were not significantly associated with acculturation rank. In the follow-up survey, however, most societies showed striking increases in the variance of both systolic and diastolic BPs with increased acculturation. These increases in variance of both systolic and diastolic BPs may be related to a) shifts in demography and/or anthropometry of some societies; b) increased range and intensity of environmental factors affecting BP and associated with increased acculturation; and/or c) genotype by environmental interactions. 4) The correlation between systolic and diastolic BP decreased over the interval for all societies within and among genetic clusters. This trend was partly the result of larger changes in variances for systolic than diastolic BP in the resurveys.(ABSTRACT TRUNCATED AT 400 WORDS)

Lipid and apolipoprotein levels in six Solomon Island societies differ from those in a U.S. white population.

Levels of plasma cholesterol, triglycerides, and apolipoproteins (apo) AI, AII, and E in 560 males and 744 females from six Solomon Island societies were compared with levels in age- and sex-matched participants in the Rochester Family Heart Study (RFHS). The overall average cholesterol, triglyceride, apo AI, and apo AII levels for all the Solomon Island societies were significantly lower than levels for the RFHS (P less than 0.001). The mean level of apo E for these societies was significantly higher than levels in RFHS in spite of the fact that the levels of triglycerides were significantly lower. Normally, apo E is a major constituent of triglyceride-rich very-low-density lipoprotein (VLDL). For both sexes, none of the Solomon Island societies showed a significant correlation of plasma cholesterol levels with apo E. In the RFHS, this correlation was 0.50 in males and 0.43 in females. Mean apo E levels are estimated to be 4.15-6.0% of the high-density lipoprotein (HDL) protein in the different Solomon Island societies. This study establishes a distinctive Solomon Island lipid profile characterized by the high apo E levels, which appear to be associated primarily with the HDL particle, whereas, in normal Western populations, it is associated primarily with VLDL, and only small quantities are associated with HDL.