The ADH3(1-1) genotype appears to substantially increase the risk of ethanol-related oral cancer, thus providing further evidence for the carcinogenicity of acetaldehyde.

To examine occupation, smoking, and other risk factors for nasal cancer, a case-control study was conducted among 160 patients, who were admitted to four hospitals in North Carolina and Virginia between 1970 and 1980, and 290 controls. Employment in the furniture industry was not associated with squamous cell tumors, but such employment increased the risk of nasal adenocarcinoma by fivefold. In addition, approximately threefold excess risks of adenocarcinoma were observed for those employed in other industries involving possible exposure to wood dust. Elevated risks among males were also associated with occupational exposures to chromates (relative risk (RR) = 5.1) and chemicals (RR = 2.7). Among females, an excess risk was associated with employment in the textile industry, particularly for jobs involving dust exposure (RR = 2.3). Although there was no evidence that alcohol consumption affected risk, heavy cigarette smokers were at a two- to threefold excess risk (predominantly for squamous cell tumors); in addition, there was evidence that there was an elevated risk associated with the use of snuff. Elevated risks were also associated with histories of nasal polyps (RR = 8.3), recurrent nose bleeds (RR = 2.0), and sinus trouble (RR = 2.7). These findings provide leads for further studies, and underscore the importance of distinguishing between histologic types.

Circulating antibodies to certain nuclear antigens are of diagnostic use and contribute to our ability to determine prognosis in terms of predicting particular organ system involvement in the rheumatic diseases. Examples include antibodies to DNA and to the Sm antigen associated with systemic lupus erythematosus (SLE) (l), antibodies to nuclear ribonucleoprotein n-RNP associated with mixed connective tissue disease (2), and antibodies to another nucleic acidic protein antigen, PM-1, which have a high specificity for polymyositis (3). A recent report of 16 patients who had circulating Sm antibodies suggested that this serologic pattern might be indicative of a more benign form of SLE with nonprogressive nephritis (4). The purpose of the present study was to compare the disease patterns in patients with circulating Sm and DNA antibodies and to determine the relative importance of the presence of Sm antibody or the absence of DNA antibody in the disease which occurred. MATERIALS AND METHODSCalf thymus nuclear extract, extractable nuclear antigen (ENA), and DNA were prepared by previously published techniques (5). The Sm antigen was prepared from the washed DNA-histone fraction of the nuclear extract. The resulting pellet was treated with DNase (Worthington Biochemical Corporation) for one hour at 37OC in the presence of 5 M/ml of MgCI,. The resulting supernatant, which had Sm antigen activity but no RNP activity, was further concentrated by 30% saturated ammonium sulfate, lyophilized, and stored at -10°C. This partially purified Sm antigen was then used in hemagglutination to determine the Sm antibody titer.All sera were analyzed for the presence of Sm and n-RNP by passive hemagglutination (2), immunodiffusion (6), and counterimmunoelectrophoresis (7) according to pre-

In June 1982, the sexual practices of 245 homosexual male outpatients of private physicians were evaluated in relationship to decreased numbers of helper T lymphocytes, an abnormality that is characteristic of the acquired immunodeficiency syndrome (AIDS). Three risk groups were defined a priori--85 high-risk men from central Manhattan ("New York"), 96 intermediate-risk men from Washington, DC, with AIDS-area homosexual contacts ("Washington-exposed"), and 64 low-risk Washington, DC, men without such contacts ("Washington-unexposed"). An increasing number of homosexual partners was correlated with decreasing helper T-cell counts (R = -0.29, p = 0.009) and decreasing helper:suppressor T-cell ratios (R = -0.32, p = 0.005) in the entire study group combined and in New York subjects separately. Suppressor T-cell counts were unrelated to the number of partners in all three groups. Increasingly frequent receptive anal intercourse correlated with decreasing helper T-cell counts most clearly in the New York City group (R = -0.23, p = 0.04), somewhat less so in the Washington-exposed group (R = -0.18, p = 0.07), and not at all in the Washington-unexposed group (R = -0.09, p = 0.48). This association persisted in the New York and Washington-exposed groups after adjusting for seven other sexual practices, the number of homosexual partners, and five other potentially confounding variables. A transmissible agent associated with receptive anal intercourse best explains these data. The cause of these low helper T-cell counts may also be the cause of AIDS.