Is alcohol a cofactor of HIV and AIDS? Evidence from immunological and behavioral studies

The authors aim to critically examine empirical research on the effects of alcohol on HIV and AIDS from the immunological and behavioral fields. In vitro immunological studies demonstrate that social drinking increases the susceptibility of human cells to HIV infection. Animal studies show that acute and chronic alcohol ingestion increases rare of progression from retrovirus to clinical illness. In humans with HIV, no experimental evidence shows that alcohol is a cofactor of AIDS. Findings from behavioral studies show that a link between social drinking and risk of HIV is weak. No experimental evidence demonstrates that chronic drinking influences rate and course of disease progression to AIDS in humans who are HIV+. It is premature to promote the role of alcohol as a cofactor in HIV and AIDS.

Testing market efficiency: Evidence from the NFL sports betting market

This article examines the efficiency of the National Football League (NFL) betting market. The standard ordinary least squares (OLS) regression methodology is replaced by a probit model. This circumvents potential econometric problems, and allows us to implement more sophisticated betting strategies where bets are placed only when there is a relatively high probability of success. In-sample tests indicate that probit-based betting strategies generate statistically significant profits. Whereas the profitability of a number of these betting strategies is confirmed by out-of-sample testing, there is some inconsistency among the remaining out-of-sample predictions. Our results also suggest that widely documented inefficiencies in this market tend to dissipate over time.

Genetic heterogeneity in familial acute myelogenous leukemia: Evidence for a second locus at chromosome 16q21-23.2

The identification of genes responsible for the rare cases of familial leukemia may afford insight into the mechanism underlying the more common sporadic occurrences. Here we test a single family with 11 relevant meioses transmitting autosomal dominant acute myelogenous leukemia (AML) and myelodysplasia for linkage to three potential candidate loci. In a different family with inherited AML, linkage to chromosome 21q22.1-22.2 was recently reported; we exclude linkage to 21q22.1-22.2, demonstrating that familial AML is a heterogeneous disease. After reviewing familial leukemia and observing anticipation in the form of a declining age of onset with each generation, we had proposed 9p21-22 and 16q22 as additional candidate loci. Whereas linkage to 9p21-22 can be excluded, the finding of a maximum two-point LOD score of 2.82 with the microsatellite marker D16S522 at a recombination fraction theta = 0 provides evidence supporting linkage to 16q22. Haplotype analysis reveals a 23.5-cM (17.9-Mb) commonly inherited region among all affected family members extending from D16S451 to D1GS289, In order to extract maximum linkage information with missing individuals, incomplete informativeness with individual markers in this interval, and possible deviance from strict autosomal dominant inheritance, we performed nonparametric linkage analysis (NPL) and found a maximum NPL statistic corresponding to a P-value of .00098, close to the maximum conditional probability of linkage expected for a pedigree with this structure. Mutational analysis in this region specifically excludes expansion of the AT-rich minisatellite repeat FRA16B fragile site and the CAG trinucleotide repeat in the E2F-4 transcription factor. The ''repeat expansion detection'' method, capable of detecting dynamic mutation associated with anticipation, more generally excludes large CAG repeat expansion as a cause of leukemia in this family.