Jose R. Fernandez

A mixture model approach for the analysis of microarray gene expression data

Microarrays have emerged as powerful tools allowing investigators to assess the expression of thousands of genes in different tissues and organisms. Statistical treatment of the resulting data remains a substantial challenge. Investigators using microarray expression studies may wish to answer questions about the statistical significance of differences in expression of any of the genes under study, avoiding false positive and false negative results. We have developed a sequence of procedures involving finite mixture modeling and bootstrap inference to address these issues in studies involving many thousands of genes. We illustrate the use of these techniques with a dataset involving calorically restricted mice.

A panel of ancestry informative markers for estimating individual biogeographical ancestry and admixture from four continents: utility and applications

Autosomal ancestry informative markers (AIMs) are useful for inferring individual biogeographical ancestry (I‐BGA) and admixture. Ancestry estimates obtained from Y and mtDNA are useful for reconstructing population expansions and migrations in our recent past but individual genomic admixture estimates are useful to test for association of admixture with phenotypes, as covariate in association studies to control for stratification and, in forensics, to estimate certain overt phenotypes from ancestry. We have developed a panel of 176 autosomal AIMs that can effectively distinguish I‐BGA and admixture proportions from four continental ancestral populations: Europeans, West Africans, Indigenous Americans, and East Asians. We present allele frequencies for these AIMs in all four ancestral populations and use them to assess the global apportionment of I‐BGA and admixture diversity among some extant populations. We observed patterns of apportionment similar to those described previously using sex and autosomal markers, such as European admixture for African Americans (14.3%) and Mexicans (43.2%), European (65.5%) and East Asian affiliation (27%) for South Asians, and low levels of African admixture (2.8–10.8%) mirroring the distribution of Y E3b haplogroups among various Eurasian populations. Using simulation studies and pedigree analysis we show that I‐BGA estimates obtained using this panel and a four‐population model has a high degree of precision (average root mean square error [RMSE]=0.026). Using ancestry–phenotype associations we demonstrate that a large and informative AIM panel such as this can help reduce false‐positive and false‐negative associations between phenotypes and admixture proportions, which may result when using a smaller panel of less informative AIMs. Hum Mutat 29(5), 648–658, 2008.

Bedtime Dosing of Antihypertensive Medications Reduces Cardiovascular Risk in CKD

Time of ingestion of hypertension medications can affect circadian patterns of BP, but whether this translates into an effect on clinical outcomes is unknown. Here, in an open-label trial, we randomly assigned 661 patients with CKD either to take all prescribed hypertension medications upon awakening or to take at least one of them at bedtime. We measured 48-hour ambulatory BP at baseline and 3 months after any adjustment in treatment or, at the least, annually. After a median follow-up of 5.4 years, patients who took at least one BP-lowering medication at bedtime had an adjusted risk for total cardiovascular events (a composite of death, myocardial infarction, angina pectoris, revascularization, heart failure, arterial occlusion of lower extremities, occlusion of the retinal artery, and stroke) that was approximately one-third that of patients who took all medications upon awakening (adjusted HR 0.31; 95% CI 0.21 to 0.46; P < 0.001). Bedtime dosing demonstrated a similar significant reduction in risk for a composite outcome of cardiovascular death, myocardial infarction, and stroke (adjusted HR 0.28; 95% CI 0.13 to 0.61; P < 0.001). Furthermore, patients on bedtime treatment had a significantly lower mean sleep-time BP and a greater proportion demonstrated control of their ambulatory BP (56% versus 45%, P = 0.003). Each 5-mmHg decrease in mean sleep-time systolic BP was associated with a 14% reduction in the risk for cardiovascular events during follow-up (P < 0.001). In conclusion, among patients with CKD and hypertension, taking at least one antihypertensive medication at bedtime improves control of BP and reduces the risk for cardiovascular events.

Influence of Time of Day of Blood Pressure–Lowering Treatment on Cardiovascular Risk in Hypertensive Patients With Type 2 Diabetes

OBJECTIVE We prospectively investigated in hypertensive patients with type 2 diabetes if bedtime treatment with ≥1 hypertension medications exerts better blood pressure control and cardiovascular risk reduction than conventional therapy, in which all medications are ingested in the morning. RESEARCH DESIGN AND METHODS We conducted a prospective, randomized, open-label, blinded end point trial on 448 hypertensive patients with type 2 diabetes, 255 men/193 women, mean ± SD age 62.5 ± 10.8 years, randomized to ingest all their prescribed hypertension medications upon awakening or ≥1 of them at bedtime. Ambulatory blood pressure was measured for 48 h at baseline and again annually or even more frequently (quarterly) after adjustments in treatment. RESULTS After a median follow-up of 5.4 years, patients ingesting ≥1 hypertension medications at bedtime showed a significantly lower cardiovascular risk (adjusted by age and sex) than subjects ingesting all medications upon awakening (hazard ratio 0.33 [95% CI 0.21–0.54]; P < 0.001). The difference between groups in the adjusted risk of major events (cardiovascular death, myocardial infarction, and stroke) was also statistically significant (0.25 [0.10–0.61]; P = 0.003). Patients treated at bedtime showed significantly lower sleep time blood pressure mean and higher prevalence of controlled ambulatory blood pressure (62.5 vs. 50.9%; P = 0.013). There was a significant 12% cardiovascular risk reduction per each 5 mmHg decrease in asleep systolic blood pressure during follow-up (P < 0.001). CONCLUSIONS Among patients with diabetes, treatment with ≥1 hypertension medications at bedtime, compared with all medications upon waking, resulted in improved ambulatory blood pressure control and significantly reduced cardiovascular morbidity and mortality.

Ancestral proportions and their association with skin pigmentation and bone mineral density in Puerto Rican women from New York city.

Hispanic and African American populations exhibit an increased risk of obesity compared with populations of European origin, a feature that may be related to inherited risk alleles from Native American and West African parental populations. However, a relationship between West African ancestry and obesity-related traits, such as body mass index (BMI), fat mass (FM), and fat-free mass (FFM), and with bone mineral density (BMD) in African American women has only recently been reported. In order to evaluate further the influence of ancestry on body composition phenotypes, we studied a Hispanic population with substantial European, West African, and Native American admixture. We ascertained a sample of Puerto Rican women living in New York (n=64), for whom we measured BMI and body composition variables, such as FM, FFM, percent body fat, and BMD. Additionally, skin pigmentation was measured as the melanin index by reflectance spectroscopy. We genotyped 35 autosomal ancestry informative markers and estimated population and individual ancestral proportions in terms of European, West African, and Native American contributions to this population. The ancestry proportions corresponding to the three parental populations are: 53.3±2.8% European, 29.1±2.3% West African, and 17.6±2.4% Native American. We detected significant genetic structure in this population with a number of different tests. A highly significant correlation was found between skin pigmentation and individual ancestry (R2=0.597, P<0.001) that was not attributable to differences in socioeconomic status. A significant association was also found between BMD and European admixture (R2=0.065, P=0.042), but no such correlation was evident with BMI or the remaining body composition measurements. We discuss the implications of our findings for the potential use of this Hispanic population for admixture mapping.

Large-scale SNP analysis reveals clustered and continuous patterns of human genetic variation

Understanding the distribution of human genetic variation is an important foundation for research into the genetics of common diseases. Some of the alleles that modify common disease risk are themselves likely to be common and, thus, amenable to identification using gene-association methods. A problem with this approach is that the large sample sizes required for sufficient statistical power to detect alleles with moderate effect make gene-association studies susceptible to false-positive findings as the result of population stratification [1, 2]. Such type I errors can be eliminated by using either family-based association tests or methods that sufficiently adjust for population stratification [3–5]. These methods require the availability of genetic markers that can detect and, thus, control for sources of genetic stratification among populations. In an effort to investigate population stratification and identify appropriate marker panels, we have analysed 11,555 single nucleotide polymorphisms in 203 individuals from 12 diverse human populations. Individuals in each population cluster to the exclusion of individuals from other populations using two clustering methods. Higher-order branching and clustering of the populations are consistent with the geographic origins of populations and with previously published genetic analyses. These data provide a valuable resource for the definition of marker panels to detect and control for population stratification in population-based gene identification studies. Using three US resident populations (European-American, African-American and Puerto Rican), we demonstrate how such studies can proceed, quantifying proportional ancestry levels and detecting significant admixture structure in each of these populations.

Effects of Time of Day of Treatment on Ambulatory Blood Pressure Pattern of Patients With Resistant Hypertension

Patients with resistant hypertension present high prevalence of a non-dipper blood pressure pattern. Recent results indicate that non-dipping is related partly to the absence of 24-hour therapeutic coverage in patients treated with single morning doses. Accordingly, we investigated the impact of treatment time on the blood pressure pattern in 700 patients with resistant hypertension on the basis of clinic measurements who were studied by 48-hour ambulatory monitoring. Among them, 299 patients received all their medication on awakening, and 401 were taking ≥1 antihypertensive drug at bedtime. The percentage of patients with controlled ambulatory blood pressure was double in patients taking 1 drug at bedtime (P=0.008). Among the 578 patients with true resistant hypertension, subjects receiving 1 drug at bedtime showed a significant reduction in the 24-hour mean of systolic and diastolic blood pressure (3.1 and 1.6 mm Hg, respectively; P<0.011). This reduction was much more prominent during nighttime (5.1 and 3.0 mm Hg; P<0.001). Accordingly, the diurnal/nocturnal blood pressure ratio was significantly increased by 2.7 and the prevalence on non-dipping reduced (56.9 versus 81.9%; P<0.001) in patients taking 1 drug at bedtime. Compared with patients receiving all drugs on awakening, subjects with 1 drug at bedtime also showed significant reductions in the average values of glucose, cholesterol, fibrinogen, and urinary albumin excretion (P<0.011). In patients with resistant hypertension, pharmacological therapy should take into account when to treat with respect to the rest–activity cycle of each patient to improve control and to avoid the non-dipper pattern associated to higher cardiovascular risk.

Treatment of non-dipper hypertension with bedtime administration of valsartan.

Background Previous results have indicated that valsartan administration at bedtime, as opposed to upon wakening, may improve the diurnal: nocturnal ratio of blood pressure without loss in 24-h coverage and efficacy. Objectives To investigate the administration time-dependent antihypertensive efficacy of valsartan in non-dipper patients. Methods We studied 148 non-dipper patients with grade 1–2 essential hypertension, aged 53.0 ± 12.6 years, who were randomly assigned to receive valsartan (160 mg/day) as a monotherapy either on awakening or at bedtime. Blood pressure was measured every 20 min during the day and every 30 min at night for 48 consecutive hours before and after 3 months of treatment. Physical activity was simultaneously monitored every minute by wrist actigraphy to accurately calculate the diurnal and nocturnal means of blood pressure on a per subject basis. Results The significant blood pressure reduction after 3 months of valsartan (P < 0.001) was similar for both treatment times (13.1 and 8.5 mmHg reduction in the 24-h mean of systolic and diastolic blood pressure with morning administration; 14.7 and 10.3 mmHg with bedtime administration; P > 0.126 for treatment-time effect). The diurnal: nocturnal ratio of blood pressure was significantly increased only when valsartan was administered before bedtime, which resulted in 75% of the patients in this group reverting to dippers, a significant increase in the percentage of patients with controlled blood pressure over 24 h, and a reduction in urinary albumin excretion. Conclusions In non-dipper hypertensive patients, dosing time with valsartan should be chosen at bedtime, for improved efficacy during the nocturnal resting hours, as well as the potential associated reduction in cardiovascular risk.

Polymorphism in the Transcription Factor 7-Like 2 (TCF7L2) Gene Is Associated With Reduced Insulin Secretion in Nondiabetic Women

Recently, the transcription factor 7-like 2 (TCF7L2) gene on chromosome 10q25.2 has been linked with type 2 diabetes among Caucasians, with disease associations noted for single nucleotide polymorphisms (SNPs) rs12255372 and rs7903146. To investigate mechanisms by which TCF7L2 could contribute to type 2 diabetes, we examined the effects of these SNPs on clinical and metabolic traits affecting glucose homeostasis in 256 nondiabetic female subjects (138 European Americans and 118 African Americans) aged 7–57 years. Outcomes included BMI, percent body fat, insulin sensitivity (Si), acute insulin response to glucose (AIRg), and the disposition index (DI). Homozygosity for the minor allele (TT) of SNP rs12255372 occurred in 9% of individuals and was associated with a 31% reduction in DI values in a recessive model. The at-risk allele TT was also associated with lower AIRg adjusted for Si in both ethnic groups, whereas rs12255372 genotype was not associated with measures of adiposity or with Si. The T allele of rs12255372 was also associated with increased prevalence of impaired fasting glucose. Genotypes at rs7903146 were not associated with any metabolic trait. Lower Si and higher AIRg observed in the African-American compared with the European-American subgroup could not be explained by the TCF7L2 genotype. Our data suggest that the TCF7L2 gene is an important factor regulating insulin secretion, which could explain its association with type 2 diabetes.

Resistance Training Conserves Fat-free Mass and Resting Energy Expenditure Following Weight Loss

Objective: To determine what effect diet‐induced ∼12 kg weight loss in combination with exercise training has on body composition and resting energy expenditure (REE) in premenopausal African‐American (AA) and European‐American (EA) women.