Appropriate use of information on family history of disease in recruitment for linkage analysis studies.

When conducting genetic studies for complex traits, large samples are commonly required to detect any of the number of genes with relatively low effect thought to underly such traits. This is because, in contrast to monogenic diseases, complex traits typically result from a number of different genetic pathways (genetic heterogeneity) and any sample is likely to contain a considerable fraction of sporadic cases (phenocopies). Such samples are time-consuming and costly to recruit and analyse. Methods which might be used to decrease sample size include attempting to select families, with the aim of reducing genetic heterogeneity or phenocopy rate within the sample. Selecting cases with positive family history of disease should reduce the phenocopy rate, and this strategy has been employed in linkage studies of complex disease, although evaluations of such a strategy have been equivocal. This paper shows how identity by descent (IBD) distributions may be calculated for affected relative pairs recruited conditional on the affection status of a third relative. These distributions are then used to calculate expected power in affected sib and half-sib linkage studies when recruitment is conditional on family history of disease. We consider the proxy conditions of recruitment conditional on disease in an affected parent or third sibling with single-locus and additive multilocus genetic models. We show that while such selection strategies can reduce power if disease risk alleles are common and environmental heterogeneity low, under models more likely to underly common complex diseases power will generally be increased, and that this effect is greater as more loci are involved. Though the proxy cases studied are more extreme than a general strategy of asking potential recruits whether they have any family history of disease, these results suggest that conditional recruitment is more generally useful than previous studies have suggested.

Age related macular degeneration and sun exposure, iris colour, and skin sensitivity to sunlight.

BACKGROUND/AIM: It has been suggested that sun exposure may be a risk factor for age related macular degeneration (AMD) and that skin sensitivity to sunlight and iris colour could be confounding factors. The aim was to investigate this further in the white population. METHODS: 446 cases with end stage AMD were compared with 283 spouse controls. Data on sun exposure, places of residence, iris colour, subjective assessment of change in iris colour, hair colour at age 20, and skin sensitivity were obtained using a questionnaire. Iris colour was graded clinically by comparison with standard photographs. AMD was graded using stereoscopic colour fundus photographs as well as clinical examination and was defined as the presence of geographic atrophy or choroidal neovascularisation. All variables were included in a multiple logistic regression model including age, sex, and smoking. RESULTS: There was no association between AMD and sun exposure or related factors except for the suggestion of an association between sunburn prone skin type and geographic atrophy which reached borderline significance. CONCLUSIONS: No significant association between AMD and sun exposure, iris colour, change in iris colour, or hair colour was demonstrated.

Smoking and age related macular degeneration: the number of pack years of cigarette smoking is a major determinant of risk for both geographic atrophy and choroidal neovascularisation.

BACKGROUND/AIMS: There is evidence that smoking is a risk factor for age related macular degeneration (AMD). However, not all studies have demonstrated this association and several key questions about the role of smoking in AMD have still to be determined. The aim of this study was to further investigate this relation for both choroidal neovascularisation (CNV) and geographic atrophy (GA). METHODS: To investigate the relation between smoking and the risk of developing age related macular degeneration (AMD) in white people, 435 cases with end stage AMD were compared with 280 controls. All subjects had graded stereoscopic colour fundus photography and AMD was defined as the presence of GA or CNV. Smoking history was assessed using multiple parameters in a detailed questionnaire. RESULTS: Comparison of current and former smokers with non-smokers was consistent with smoking being a risk factor for AMD but did not reach statistical significance. There was a strong association between AMD and pack years of cigarette smoking (p = 0.002), the odds ratio increasing with the amount smoked; for subjects with more than 40 pack years of smoking the odds ratio was 2.75 (95% CI 1.22 to 6.20) compared with non-smokers. Both types of AMD showed a similar relation; smoking more than 40 pack years of cigarettes was associated with an odds ratio of 3.43 (95% CI 1.28 to 9.20) for GA and 2.49 (95% CI 1.06 to 5.82) for CNV. Stopping smoking was associated with reduced odds of AMD and the risk in those who had not smoked for over 20 years was comparable to non-smokers. The risk profile was similar for males and females. Passive smoking exposure was associated with an increased risk of AMD (OR 1.87; 95% CI 1.03 to 3.40) in non-smokers. CONCLUSIONS: The authors have demonstrated a strong association between the risk of both GA and CNV and pack years of cigarette smoking. This provides support for a causal relation between smoking and AMD. They also show an increased risk for AMD in non-smokers exposed to passive smoking. Stopping smoking appears to reduce the risk of developing AMD.

Design and analysis of admixture mapping studies.

Admixture between populations originating on different continents can be exploited to detect disease susceptibility loci at which risk alleles are distributed differentially between these populations. We first examine the statistical power and mapping resolution of this approach in the limiting situation in which gamete admixture and locus ancestry are measured without uncertainty. We show that, for a rare disease, the most efficient design is to study affected individuals only. In a typical African American population (two-way admixture proportions 0.8/0.2, ancestry crossover rate 2 per 100 cM), a study of 800 affected individuals has 90% power to detect at P values <10(-5) a locus that generates a risk ratio of 2 between populations, with an expected mapping resolution (size of 95% confidence region for the position of the locus) of 4 cM. In practice, to infer locus ancestry from marker data requires Bayesian computationally intensive methods, as implemented in the program ADMIXMAP. Affected-only study designs require strong prior information on the frequencies of each allele given locus ancestry. We show how data from unadmixed and admixed populations can be combined to estimate these ancestry-specific allele frequencies within the admixed population under study, allowing for variation between allele frequencies in unadmixed and admixed populations. Using simulated data based on the genetic structure of the African American population, we show that 60% of information can be extracted in a test for linkage using markers with an ancestry information content of 36% at 3-cM spacing. As in classic linkage studies, the most efficient strategy is to use markers at a moderate density for an initial genome search and then to saturate regions of putative linkage with additional markers, to extract nearly all information about locus ancestry.

Cost-effective analysis of candidate genes using htSNPs: a staged approach.

We have previously shown that the selection of haplotype tag single nucleotide polymorphisms (htSNPs) and their statistical analysis in a multi-locus transmission/disequilibrium test (TDT) results in a more cost-effective genotyping strategy in disease association studies of genes by minimising redundancy due to linkage disequilibrium between SNPs. Further savings can be achieved by the use of a two-stage genotyping strategy. This approach is illustrated here in conjunction with the multi-locus TDT in determining whether common alleles of the immune regulatory genes RANK and its ligand TRANCE (RANKL) are associated with type 1 diabetes (T1D). A saving of approximately 75% of potential genotyping reactions could be made with minimal loss of power. There was little evidence from our analysis for association between the TRANCE and RANK genes and T1D in the populations tested.

Testing the possible negative association of type 1 diabetes and atopic disease by analysis of the interleukin 4 receptor gene.

Variations in the interleukin 4 receptor A (IL4RA) gene have been reported to be associated with atopy, asthma, and allergy, which may occur less frequently in subjects with type 1 diabetes (T1D). Since atopy shows a humoral immune reactivity pattern, and T1D results from a cellular (T lymphocyte) response, we hypothesised that alleles predisposing to atopy could be protective for T1D and transmitted less often than the expected 50% from heterozygous parents to offspring with T1D. We genotyped seven exonic single nucleotide polymorphisms (SNPs) and the -3223 C>T SNP in the putative promoter region of IL4RA in up to 3475 T1D families, including 1244 Finnish T1D families. Only the -3223 C>T SNP showed evidence of negative association (P=0.014). There was some evidence for an interaction between -3233 C>T and the T1D locus IDDM2 in the insulin gene region (P=0.001 in the combined and P=0.02 in the Finnish data set). We, therefore, cannot rule out a genetic effect of IL4RA in T1D, but it is not a major one.

Identification of the sources of error in allele frequency estimations from pooled DNA indicates an optimal experimental design.

Genotyping costs still preclude analysis of a comprehensive SNP map in thousands of individual subjects in the search for disease susceptibility loci. Allele frequency estimation in DNA pools from cases and controls offers a partial solution, but variance in these estimates will result in some loss of statistical power. However, there has been no systematic attempt to quantify the several sources of error in previous studies. We report an analysis of the magnitude of variance components of each experimental stage in DNA pooling studies, and find that a design based on the formation of numerous small pools of approximately 50 individuals is superior to the formation of fewer, larger pools and the replication of any of the experimental stages. We conclude that this approach may retain an effective sample size greater than 68% of the true sample size, whilst offering a 60-fold reduction in DNA usage and a greater than 30-fold saving in cost, compared to individual genotyping. The possibility of combining pooling with informed selection of haplotype tag SNPs is also considered. In this way further savings in efficiency may be possible by using pooled allele frequency estimates to infer haplotype frequencies and hence, allele frequencies at untyped markers.

Haplotype tagging for the identification of common disease genes.

Genome-wide linkage disequilibrium (LD) mapping of common disease genes could be more powerful than linkage analysis if the appropriate density of polymorphic markers were known and if the genotyping effort and cost of producing such an LD map could be reduced. Although different metrics that measure the extent of LD have been evaluated, even the most recent studies have not placed significant emphasis on the most informative and cost-effective method of LD mapping-that based on haplotypes. We have scanned 135 kb of DNA from nine genes, genotyped 122 single-nucleotide polymorphisms (SNPs; approximately 184,000 genotypes) and determined the common haplotypes in a minimum of 384 European individuals for each gene. Here we show how knowledge of the common haplotypes and the SNPs that tag them can be used to (i) explain the often complex patterns of LD between adjacent markers, (ii) reduce genotyping significantly (in this case from 122 to 34 SNPs), (iii) scan the common variation of a gene sensitively and comprehensively and (iv) provide key fine-mapping data within regions of strong LD. Our results also indicate that, at least for the genes studied here, the current version of dbSNP would have been of limited utility for LD mapping because many common haplotypes could not be defined. A directed re-sequencing effort of the approximately 10% of the genome in or near genes in the major ethnic groups would aid the systematic evaluation of the common variant model of common disease.

Statistical modeling of interlocus interactions in a complex disease: rejection of the multiplicative model of epistasis in type 1 diabetes.

In general, common diseases do not follow a Mendelian inheritance pattern. To identify disease mechanisms and etiology, their genetic dissection may be assisted by evaluation of linkage in mouse models of human disease. Statistical modeling of multiple-locus linkage data from the nonobese diabetic (NOD) mouse model of type 1 diabetes has previously provided evidence for epistasis between alleles of several Idd (insulin-dependent diabetes) loci. The construction of NOD congenic strains containing selected segments of the diabetes-resistant strain genome allows analysis of the joint effects of alleles of different loci in isolation, without the complication of other segregating Idd loci. In this article, we analyze data from congenic strains carrying two chromosome intervals (a double congenic strain) for two pairs of loci: Idd3 and Idd10 and Idd3 and Idd5. The joint action of both pairs is consistent with models of additivity on either the log odds of the penetrance, or the liability scale, rather than with the previously proposed multiplicative model of epistasis. For Idd3 and Idd5 we would also not reject a model of additivity on the penetrance scale, which might indicate a disease model mediated by more than one pathway leading to beta-cell destruction and development of diabetes. However, there has been confusion between different definitions of interaction or epistasis as used in the biological, statistical, epidemiological, and quantitative and human genetics fields. The degree to which statistical analyses can elucidate underlying biologic mechanisms may be limited and may require prior knowledge of the underlying etiology.

Association between the ancestral haplotype HLA A30B18DR3 and multiple sclerosis in central Sardinia.

Association and linkage studies have established the importance of the major histocompatibility complex (MHC) in the susceptibility for multiple sclerosis (MS). We carried out a case-control study to investigate the ancestral haplotype A30B18DR3 and MS in the Nuoro population of Sardinia, which is isolated and genetically distinct from other populations in the Mediterranean basin and characterized by genetic homogeneity, high level of inbreeding, low migration, high prevalence of MS, high frequency of the relevant haplotype, and high past malaria prevalence. Cases and controls were serologically typed for the currently recognized HLA-A, B, and DR antigens. We used a log-linear approach to fit a wide class of models. We tested our hypothesis comparing different models via a likelihood ratio test. We overcame the complication due to unknown gametic phase using expectation-maximization (EM) algorithm as the estimation method. We estimated confidence intervals for odds ratio by using a profile likelihood approach. We found that: (1) the ancestral haplotype A30B18DR3 was associated to MS after allowing for a possible stratification in cases and controls; (2) DR3 allele was conditional independent on disease status, given A30B18 haplotype; (3) there was a tendency for ORs for the high-risk haplotypes to be higher in the high malaria strata; however, this indication did not achieve statistical significance (P = 0.11).

Bayes estimates of haplotype effects.

We describe a Markov chain Monte Carlo implementation of a Bayesian approach to estimating associations of a trait with a large set of haplotypes recently introduced by Clayton and Jones [Am J Hum Genet 65:1161-9, 2000]. The model uses the length of the longest segment in common between any two haplotypes to define the prior correlation structure for the set of haplotype effects, using an intrinsic autocorrelation model. When applied to the Genetic Analysis Workshop 12 data for trait Q1, we found highly significant variation between haplotypes, using either a structured or unstructured covariance matrix.

Multivariate parametric random effect regression models for fecundability studies.

Delay until conception is generally described by a mixture of geometric distributions. Weinberg and Gladen (1986, Biometrics 42, 547-560) proposed a regression generalization of the beta-geometric mixture model where covariates effects were expressed in terms of contrasts of marginal hazards. Scheike and Jensen (1997, Biometrics 53, 318-329) developed a frailty model for discrete event times data based on discrete-time analogues of Hougaard's results (1984, Biometrika 71, 75-83). This paper is on a generalization to a three-parameter family distribution and an extension to multivariate cases. The model allows the introduction of explanatory variables, including time-dependent variables at the subject-specific level, together with a choice from a flexible family of random effect distributions. This makes it possible, in the context of medically assisted conception, to include data sources with multiple pregnancies (or attempts at pregnancy) per couple.

Evidence for genetic associations between asthma, atopy, and bronchial hyperresponsiveness: a study of 8- to 18-yr-old twins.

We measured asthma in the last 12 mo, diagnosed by a respiratory physician at interview; atopy, defined by a positive skin prick test to any of eight common allergens; and bronchial hyperresponsiveness (BHR) to hypertonic saline, in 381 twin pairs aged 8 to 18 yr selected from the Australian Twin Registry-183 monozygous (MZ) and 198 dizygous (DZ). The associations between twins, as measured by an odds ratio, were greater in MZ pairs compared with DZ pairs for asthma: 25.6 (95% confidence interval 11.3- 57.8) versus 1.9 (1.0-3. 5); atopy: 14.6 (7.1-30.1) versus 2.5 (1.4- 4.5); and BHR: 14.1 (6. 4-31.0) versus 4.2 (2.1-8.6) (all p < 0.002). The associations between each pair of traits within an individual were slightly greater than the association between one trait in a twin and the other trait in the cotwin (cross-trait cross-pair) in MZ pairs. Further, the associations in MZ pairs were greater than in DZ pairs (p < 0.05). Under the assumptions of the classic twin model, these data suggest that the strong cross-sectional associations between these three traits are due to an overlap between the genetic factors involved in each of these three traits.

Unbiased application of the transmission/disequilibrium test to multilocus haplotypes.

When the transmission/disequilibrium test (TDT) is applied to multilocus haplotypes, a bias may be introduced in some families for which both parents have the same heterozygous genotype at some locus. The bias occurs because haplotypes can only be deduced from certain offspring, with the result that the transmissions of the two parental haplotypes are not independent. We obtain an unbiased TDT for individual haplotypes by calculating the correct variance for the transmission count within a family, using information from multiple siblings if they are available. An existing correction for dependence between siblings in the presence of linkage is retained. To obtain an unbiased multihaplotype TDT, we must either count transmissions from one randomly chosen parent or count all transmissions and estimate the significance level empirically. Alternatively, we may use missing-data techniques to estimate uncertain haplotypes, but these methods are not robust to population stratification. An illustration using data from the insulin-gene region in type 1 diabetes shows that the validity and power of the TDT may vary by an order of magnitude, depending on the method of analysis.

Calcaneal ultrasound attenuation in an elderly population: measurement position and relationships with body size and past fractures.

This study demonstrates, the relationship between past fracture, body size and broadband ultrasound attenuation (BUA) and investigates two sites of BUA measurement in a representative elderly population of men and women (n = 2106). We measured BUA at a fixed position and at a consistent anatomic position within the calcaneus. We found fixed BUA was less closely correlated with stature and age than anatomic BUA. Both correlations were substantially weaker in men than in women. Mean BUA was significantly lower in women with a past fracture compared with nonfracturers (fixed BUA 63.3 vs 69.4 dB/MHz, p = 0.0004; anatomic BUA 77.6 vs 81.7 dB/MHz, p = 0.013). However, in women, the fixed BUA was better than the anatomic BUA at discriminating between fracturers and nonfracturers (OR 1.38/SD (95% CI 1.12-1.68) and OR 1.22/SD (0.99-1.52), respectively) when adjusted for body size and age. There was no significant difference in either BUA in men with or without a past fracture. In conclusion, currently the fixed position for BUA measurement is preferable and, whilst we have demonstrated that it is possible to locate an anatomically consistent point in the calcaneus, the position chosen by this study did not provide a measurement with more discriminatory capability than the fixed position. In women, BUA behaves similarly to bone mineral density in relation to stature and in its strength of association with past fracture, while the lack of association in men may reflect differing contributions by bone strength to fracture risk in the sexes.

Fine genetic mapping using haplotype analysis and the missing data problem.

The genetic basis of many human diseases, especially those with substantial genetic determinants, has been identified. Notable amongst others are cystic fibrosis, Huntington's disease and some forms of cancer. However, the detection of genetic factors with more modest effects such as in bipolar disorders and a majority of the cancers, has been more complicated. Standard linkage analysis procedures may not only have little power to detect such genes but they do, at best, only narrow the location of the disease susceptibility gene to a rather large region. Association studies are therefore necessary to further unveil the aetiological relevance of these factors to disease. However, the number of tests required if such procedures were used in extended genome-wide screens, is prohibitive and as such association studies have seen limited application, except in the investigation of candidate genes. In this paper, we discuss a logistic regression approach as a generalization of this procedure so that it can accommodate clusters of linked markers or candidate genes. Furthermore, we introduce an expectation maximization (E-M) algorithm with which to estimate haplotype frequencies for multiple locus systems with incomplete information on phase.

Multi-level modelling of conception in artificial insemination by donor.

Data on insemination with donor's sperm have a crossed hierarchical structure due to the coexistence of female factors (ovulatory cycles within pregnancy attempt within women) and male factors (inseminations within donations within donors). A crossed random multi-level logistic model, taking account of this structure, permits an improved estimation of the fixed effects and provides insights into their influence at each level in the hierarchy. We present an efficient algorithm for fitting such models using alternating EM steps. We further discuss the inclusion of compositional covariates to determine what information the quality of a donation conveys regarding the donor basal fecundability and on the specific sperm donation.

Genotypic relative risks under ordered restriction.

In mapping diseases of complex aetiology, conventional linkage approaches narrow the location of the disease susceptibility locus to quite a large region so that candidate gene association studies are then necessary to further isolate these genes. However, even in the simplest scenario where the candidate locus is bi-allelic, two statistical tests with various correcting factors have been proposed: a chi-square 1 df test (counting chromosomes) which may be slightly conservative and a 2 df chi-square test (counting genotypes) which may lack power because of the extra degree of freedom. This paper introduces a better and more powerful alternative which turns out to be a compromise between the two existing statistical tests. The asymptotic distribution of this test statistic is determined and the efficacy of the 3 tests are compared under different genetic models by simulation.

The Australian Incident Monitoring Study in Intensive Care: AIMS-ICU. The development and evaluation of an incident reporting system in intensive care.

Intensive care units are complex, dynamic patient management environments. Incidents and accidents can be caused by human error, by problems inherent in complex systems, or by a combination of these. Study objectives were to develop and evaluate an incident reporting system. A report form was designed eliciting a description of the incident, contextual information and contributing factors. Staff group sessions using open-ended questions, observations in the workplace and a review of earlier narratives were used to develop the report form. Three intensive care units participated in a two-month evaluation study. Feedback questionnaires were used to assess staff attitudes and understanding, project design and organization. These demonstrated a positive attitude and good understanding by more than 90% participants. Errors in communication, technique, problem recognition and charting were the predisposing factors most commonly chosen in the 128 incidents reported. It was concluded that incident monitoring may be a suitable technique for improving patient safety in intensive care.

Quantification of thrombelastographic changes after blood component transfusion in patients with liver disease in the intensive care unit.

Thrombelastography (TEG) can be used to monitor hemostasis and guide transfusion therapy during orthotopic liver transplantation. However, data are limited regarding the type and quantity of blood components necessary for TEG-guided blood component transfusion in coagulopathic critically ill patients with liver disease. We evaluated changes in four thrombelastogram variables (reaction time, thrombin constant time, alpha angle, and maximum amplitude) in whole blood samples after 74 separate blood component transfusions in 60 critically ill patients with a coagulopathy and liver disease. Only platelets significantly improved TEG variables in patients who received a single type of blood component. Each unit of platelets decreased the reaction and thrombin constant time by 0.43 (P < 0.05) and 0.82 (P < 0.005) min, respectively, increased the alpha angle by 1.5 degrees (P < 0.005), and the maximum amplitude by 1.4 mm (P < 0.005). In patients who received multiple blood components, cryoprecipitate decreased the thrombin constant time by 0.56 min/U (P < 0.05), and each unit of platelets decreased the thrombin constant time by 0.39 min (P < 0.005), and increased the alpha angle and maximum amplitude by 0.63 degrees (P < 0.05) and 0.99 mm (P < 0.005), respectively. We conclude that platelet transfusions, alone or in combination with other blood components, are most effective for improving abnormal TEG variables in coagulopathic critically ill patients with liver disease.