Confounders contributing to the reported associations of coffee or caffeine with disease

The role of caffeine or coffee in causing or promoting the incidence of serious disease is equivocal. Two design factors may account for the discrepancies in reported findings on the effects of coffee drinking: (a) imprecision of measurement and (b) confounding variables. A study of 2,714 white U.S. adults disclosed that, of 32 risk factors analyzed by linear and logistic regression, only sex and cigarette smoking were found to be important potential confounders of caffeine and coffee intake. Partial R2 values of the other 30 risk factors were relatively small and were inconsistent for each sex. It is unlikely that any of these factors could explain any of the reported associations between caffeine or coffee consumption and certain diseases. However, certain weak associations with caffeine or coffee intake should be included in the study design when they are known to be risk factors of a disease under investigation. These factors for men are dietary fat intake, vitamin C intake, and body mass index; and for women are vitamin use, alcohol intake, stress, and perceived health status.     

Radiotherapy of periocular basal cell carcinomas: recurrence rates and treatment with special attention to the medical canthus.

Basal cell carcinomas of the eyelids, especially those in the medial canthal area, may cause extensive local destruction. Recurrent tumours are more aggressive and become progressively more difficult to treat; this is especially true for postirradiated recurrent, medial canthal, basal cell carcinomas. Tumours in this area should thus be treated by a technique which allows tissue sampling in order to gauge the adequacy of the treatment, with the goal being complete extirpation of the tumour. Excision monitored by frozen section control or Mohs' surgery is our recommendation based on a retrospective analyses of 631 eyelid basal cell carcinomas, half of which were primary tumours and half recurrent.     

Inhibition of phosphoribosylpyrophosphate synthetase by 4-methoxy-(MRPP) and 4-amino-8-(D-ribofuranosylamino) pyrimido[5,4-d]pyrimidine (ARPP)

The basis for the antitumor activities of the exocyclic amino nucleosides 4-amino-(ARPP) and 4-methoxy-8-(D-ribofuranosylamino)pyrimido[5,4-d]pyrimidine (MRPP) was investigated. The primary target of these nucleosides appeared to be 5-phospho-alpha-D-ribofuranose-1-pyrophosphate (PRPP) synthetase. MRPP-5'-monophosphate was a competitive inhibitor (Ki = 40 microM) of the activation of this enzyme by the cofactor inorganic phosphate (K alpha = 2.2 mM). Consequently, ARPP and MRPP treatment of WI-L2 cultures rapidly inhibited both de novo pyrimidine and purine synthesis as well as the nucleotide salvage reactions dependent on PRPP, ARPP or MRPP treatment completely prevented [14C]bicarbonate incorporation into acid-soluble pyrimidine and purine nucleotides. The rate of salvage of [8-14C]hypoxanthine to form IMP was decreased by 85%. Treatment of cells with these agents caused a 50% reduction in the steady-state level of PRPP. When the capacity of the treated cells for sustained synthesis of PRPP was examined by adenine incorporation, the rate of adenine uptake was inhibited by greater than 50%. In vivo treatment of BDF1 mice with a single dose of ARPP (173 mg/kg) or MRPP (62 mg/kg) extended the mean life span of the mice, which had been inoculated intraperitoneally 1 day earlier with 1 x 10(6) L1210 murine leukemia cells, by 62 and 82% respectively. These studies indicate that MRPP and ARPP inhibit PRPP synthetase, and that PRPP synthetase may be a viable target in the development of certain antitumor agents.     

Urinary hydroxypyridinium cross-links of collagen in primary hyperparathyroidism.

Urinary concentrations of the collagen cross-links, pyridinoline (PYD) and deoxypyridinoline (DPD), were determined in 87 patients with untreated or surgically treated primary hyperparathyroidism (PHPT). Eighty-four healthy individuals, matched for age and sex, constituted the control group for the excretion of pyridinium cross-links. In addition, a subgroup of 25 patients with PHPT was followed longitudinally for up to 2 yr after successful parathyroidectomy. Mean urinary excretion of PYD (46.8 +/- 2.7 nmol/mmol creatinine) and DPD (17.6 +/- 1.3 nmol/mmol creatinine) was significantly higher in patients with untreated PHPT than in normal subjects (P less than 0.001). In the group undergoing successful parathyroidectomy, mean urinary concentrations of PYD (34 +/- 2.5) and DPD (9.4 +/- 0.8) were similar to those in normal controls and significantly lower than those in the untreated patient population (P less than 0.001). The urinary concentration of both cross-links was significantly correlated with serum levels of both alkaline phosphatase and PTH. Mean urinary concentrations of both cross-link compounds decreased significantly within 6 months in patients followed longitudinally and as early as 2 weeks after surgery in individual patients compared to presurgical baseline values. These changes preceded the reduction in serum alkaline phosphatase and hydroxyproline by approximately 6 months. The results demonstrate that urinary hydroxypyridinium cross-links of collagen are useful indices in the clinical assessment of bone involvement in PHPT.     

Moderate drinking in ex-alcoholics: recent studies

The literature on the occurrence of moderate drinking in ex-alcoholics that has been published since the Rand report in 1976 is reviewed. Although differences in diagnosis, definitions of moderate drinking and length of follow-up make strict comparisons of the studies difficult, the majority of studies seem to indicate that the earlier reports of the frequency of such moderate drinking may be overly optimistic. Depending on the definition of moderate drinking that is used, the longer the interval required for alcoholics to sustain moderate, problem-free drinking, the less likely is such an outcome. Among treated alcoholics, the percentage probably ranges from about 2 to 12%; the percentage may be higher among alcoholics identified in community population samples. Thus far, the only factor common to alcoholics who are able to achieve moderate drinking is their being mild cases (i.e., having fewer lifetime alcohol-related problems than other alcoholics). Factors pertinent in assessing discrepancies between the various studies are discussed.     

Respiratory symptoms associated with the use of azodicarbonamide foaming agent in a plastics injection molding facility

Respiratory health variables were studied cross-sectionally in 227 employees of a plastics molding facility where numerous complaints had been apparently associated with the use of azodicarbonamide foaming agent in injection molding. Pre- and postshift respiratory status measures and azodicarbonamide concentrations were also obtained for 17 employees. Cross-sectional pulmonary function differences by injection molding status were not observed. Modest decrements in pulmonary function measures were observed between start and end of shift but with no dose-effect relationship. A strong association was observed for injection molding workers for eye/nose/throat irritation, cough, and wheezing. Additionally, wheezing, chest tightness, and symptoms of chronic bronchitis were strongly associated with work in injection molding during periods in which azodicarbonamide was in use. These results suggest respiratory symptom causation by some combination of azodicarbonamide itself, reaction products of azodicarbonamide formed during injection molding, or other unidentified agents uniquely associated with the process of injection molding with azodicarbonamide foaming agent.     

Identifiability and exchangeability for direct and indirect effects.

We consider the problem of separating the direct effects of an exposure from effects relayed through an intermediate variable (indirect effects). We show that adjustment for the intermediate variable, which is the most common method of estimating direct effects, can be biased. We also show that even in a randomized crossover trial of exposure, direct and indirect effects cannot be separated without special assumptions; in other words, direct and indirect effects are not separately identifiable when only exposure is randomized. If the exposure and intermediate never interact to cause disease and if intermediate effects can be controlled, that is, blocked by a suitable intervention, then a trial randomizing both exposure and the intervention can separate direct from indirect effects. Nonetheless, the estimation must be carried out using the G-computation algorithm. Conventional adjustment methods remain biased. When exposure and the intermediate interact to cause disease, direct and indirect effects will not be separable even in a trial in which both the exposure and the intervention blocking intermediate effects are randomly assigned. Nonetheless, in such a trial, one can still estimate the fraction of exposure-induced disease that could be prevented by control of the intermediate. Even in the absence of an intervention blocking the intermediate effect, the fraction of exposure-induced disease that could be prevented by control of the intermediate can be estimated with the G-computation algorithm if data are obtained on additional confounding variables.     

Identifiability, exchangeability and confounding revisited

In 1986 the International Journal of Epidemiology published "Identifiability, Exchangeability and Epidemiological Confounding". We review the article from the perspective of a quarter century after it was first drafted and relate it to subsequent developments on confounding, ignorability, and collapsibility.