Stress and suicide in the Nurses' Health Study.

STUDY OBJECTIVES: Although stress is thought to be a risk factor for suicide, most research has been retrospective or has focused on attempted suicides or suicide ideation. This study examined prospectively the associations between self perceived stress, diazepam use, and death from suicide among adult women. DESIGN: A cohort study was conducted with 14 years of follow up. Stress at home and at work were assessed by questionnaire and scored on a four point scale: minimal, light, moderate, or severe. SETTING: Eleven states within the United States. PARTICIPANTS: Female nurses (n=94 110) who were 36 to 61 years of age when they answered questions on stress and diazepam use in 1982. RESULTS: During 1 272 000 person years of observation 73 suicides were identified. After adjustment for age, smoking, coffee consumption, alcohol intake, and marital status, the relation between self reported stress and suicide remained U shaped. Compared with the light home and work stress categories, which had the lowest incidences of suicide, risks were increased among women reporting either severe (relative risk (RR) = 3.7, 95% confidence intervals (CI) 1.7 to 8.3) or minimal (RR=2.1, 95% CI 1.0 to 4.5) home stress and either severe (RR=1.9, 95% CI 0.8 to 4.7) or minimal (RR=2.4, 95% CI 0.9 to 6.1) work stress. When responses to home and work stress were combined, there was an almost fivefold increase in risk of suicide among women in the high stress category. Risk of suicide was over eightfold among women reporting high stress or diazepam use compared with those reporting low stress and no diazepam use. CONCLUSIONS: The relation between self reported stress and suicide seems to be U shaped among adult women. The excess risk for those reporting minimal stress may reflect denial or undiagnosed depression or an association with some other unmeasured risk factor for suicide.

Consumer perception of risk associated with filters contaminated with glass fibers.

The filters in Eclipse, a new cigarette-like smoking article marketed by R. J. Reynolds Tobacco Company, are contaminated with glass fibers, fragments, and particles. Reported herein are the results of a study in which consumers were questioned about their opinions as to whether exposure to glass fibers in such a filter poses an added health risk beyond that from smoking and whether the manufacturer has an obligation to inform consumers about the glass contamination problem. The study queried 137 adults who were interviewed while waiting at a Division of Motor Vehicles office in Erie County, New York in 1997. All but one person expressed the view that the presence of glass fibers on the filters poses an added health risk beyond that associated with exposure to tobacco smoke alone. Nearly all expressed the position that the cigarette manufacturer has a duty to inform the public about the potential for glass exposure.

Mismeasurement and the resonance of strong confounders: correlated errors.

Confounding in epidemiology, and the limits of standard methods of control for an imperfectly measured confounder, have been understood for some time. However, most treatments of this problem are based on the assumption that errors of measurement in confounding and confounded variables are independent. This paper considers the situation in which a strong risk factor (confounder) and an inconsequential but suspected risk factor (confounded) are each measured with errors that are correlated; the situation appears especially likely to occur in the field of nutritional epidemiology. Error correlation appears to add little to measurement error as a source of bias in estimating the impact of a strong risk factor: it can add to, diminish, or reverse the bias induced by measurement error in estimating the impact of the inconsequential risk factor. Correlation of measurement errors can add to the difficulty involved in evaluating structures in which confounding and measurement error are present. In its presence, observed correlations among risk factors can be greater than, less than, or even opposite to the true correlations. Interpretation of multivariate epidemiologic structures in which confounding is likely requires evaluation of measurement error structures, including correlations among measurement errors.

Mismeasurement and the resonance of strong confounders: uncorrelated errors.

Greenland first documented (Am J Epidemiol 1980; 112:564-9) that error in the measurement of a confounder could resonate--that it could bias estimates of other study variables, and that the bias could persist even with statistical adjustment for the confounder as measured. An important question is raised by this finding: can such bias be more than trivial within the bounds of realistic data configurations? The authors examine several situations involving dichotomous and continuous data in which a confounder and a null variable are measured with error, and they assess the extent of resultant bias in estimates of the effect of the null variable. They show that, with continuous variables, measurement error amounting to 40% of observed variance in the confounder could cause the observed impact of the null study variable to appear to alter risk by as much as 30%. Similarly, they show, with dichotomous independent variables, that 15% measurement error in the form of misclassification could lead the null study variable to appear to alter risk by as much as 50%. Such bias would result only from strong confounding. Measurement error would obscure the evidence that strong confounding is a likely problem. These results support the need for every epidemiologic inquiry to include evaluations of measurement error in each variable considered.

Adolescents' knowledge of medical terminology and family health history.

Compared 309 youths ages 11 to 15 years and their parents with respect to their comprehension of terms for seven common medical disorders: heart attack, stroke, atherosclerosis, ulcer, hypertension, diabetes, and cancer. For two thirds of the adolescent sample, accuracy of reporting of these disorders among the parents and grandparents was assessed. Results indicated considerable variation among disorders with respect to both comprehension of terms and accuracy of family health history. Adolescents' age was a major predictor of knowledge of medical terms (r = .41). Age was not related to accuracy of family health information. Consonant with this finding, adolescents' level of accuracy regarding family health history was generally similar to that of previous adult samples, suggesting that family health information is acquired and retained at an early age. Adolescents were more accurate concerning parents' compared with grandparents' history of hypertension.

Win some, lose some: parental hypertension and heart rate change in an incentive versus response cost paradigm.

Fowles (1983), citing evidence from separate studies, suggests that both incentive and response cost paradigms increase heart rate and should be subsumed under Gray's (1975) 'appetitive motivational system'. Shock avoidance and loss of reward (response cost) contingencies, while aversive, appear to evoke this motivational system; consequently both should elicit heart rate increases independent of anxiety. The present investigation compared magnitude of heart rate changes observed under conditions of winning and losing money. Results showed: no differences between incentive and response cost conditions; no effect of state anxiety on heart rate in these conditions, despite an elevation of state anxiety on the task day relative to a subsequent relaxation day assessment; and some evidence for the presence under both such appetitive conditions of cardiovascular hyperresponsivity among offspring of hypertensive parents. The results suggest a need for systematic parametric studies of experimental conditions.

Cardiovascular responsivity to stress: family patterns and the effects of instructions.

Cardiovascular responses to two stressful tasks, as well as to instructions regarding the first task, were assessed in two generations of the same biological family (preadolescent children and their middle-aged parents). Information on the presence of hypertension and other cardiovascular disorders was also obtained for the grandparent generation. Results for the middle-aged normotensive adults confirmed previous observations among young adults of cardiovascular hyperresponsivity among offspring of hypertensive parents; hyperresponsivity was also observed during an instructions period prior to active participation in the task. Personality variables (Type A; hostility) were not related to the observed hyperresponsivity. Assessments of similarity of cardiovascular responses between preadolescent children and their parents were more equivocal; previous studies showing positive results have, for the most part, evaluated children who were somewhat older than those in the present study.

Accuracy of knowledge of family history of cardiovascular disorders.

There has been increased interest in the use of familial trends in physical and psychological disorders for identifying individuals at risk; research on individuals who have relatives with certain health problems may contribute to knowledge of etiology. In addition, accurate family health information may allow targeting of prevention and early detection programs to minimize cost and maximize utility. This study compared 292 undergraduates' reports of their parents' and grandparents' histories of six medical disorders with the parents' reports. Results showed moderate agreement concerning the parents' health, but substantial disagreement concerning the grandparents' health. Demographic and personality variables did not predict accuracy of students' reports of parental hypertension. Among the disorders, differences in agreement of reporting arose; the most salient condition, heart attack, had the highest agreement for both parents and grandparents. These results suggested that caution is necessary in evaluating family health information from subjects in research and from target individuals in prevention programs. This problem is especially serious for disorders such as hypertension, which is less salient and which also shows age-related penetrance, with the disorder often not evident until the 5th or 6th decade of life.

Inaccuracy of family health information: implications for prevention.

There is a conflict between the rest of the family's right to medical confidentiality and the individual's need to know about health risks for which she or he may show increased susceptibility. Research has shown increased risk in the development of a variety of disorders for those with a positive family history. In many cases the familial predisposition appears to interact with alterable environmental factors (e.g., diet). This suggests a need for targeting of preventive efforts at those at highest risk. As success of early detection and prevention programs increase, there will appear an increased need for accurate family health history information as an aid in early identification. Research on accuracy of family health information (Hastrup, Hotchkiss & Johnson, this issue) shows extensive inaccuracy. Implications of inaccurate information for self-directed and public health prevention are described. Research and progress in prevention will likely accelerate, continuing to alter the traditional view from one of the individual with a personal health history to one in which individuals share a familial (genetic + shared environment + similar behavior patterns) risk with others, who may also benefit from preventive programs.

Sex differences in cardiovascular stress responses: modulation as a function of menstrual cycle phases.

An experiment was conducted to study the effects of two laboratory stresses (unsignalled shock-avoidance reaction time and cold pressor tasks) on cardiovascular responses of young adult women during the follicular or the luteal phases of the menstrual cycle. A comparison group of male subjects, matched for family history of hypertension and for time between stress and rest sessions, was also tested. Results showed reduced cardiovascular responsiveness to the reaction time task (heart rate and blood pressure) for the women in the follicular phase of their cycle as compared with those in the luteal phase or with the males. Cardiovascular hyperresponsivity to this task has been previously related to familial history of hypertension. No differential changes were observed during the cold pressor test. These results are discussed in terms of possible hormonal effects on blood pressure regulation, and with respect to methodological considerations for assessing stress reactivity among women.

The role of imagination in the disease process: post-Cartesian history.

A previous paper traced the pre-Cartesian history of the concept of imagination as a causal variable in physiopathology. The present paper continues that history, showing the prohibitive impact of mind--body dualism and sampling the views of some eighteenth- and nineteenth-century supporters of the theory. Contemporary research has produced abundant evidence supporting the historic belief that imagination has an around function and a direct link to physiopathology. This literature is surveyed and possible therapeutic applications are discussed.