Description of cold.dat Variable Name Variable Description Format ------------------------------------------------------------- id subject id number f6.0 clincold S got clinical cold f1.0 0=no 1=yes infctd S got infected f1.0 0=no 1=yes infwsym S got infected & had symptoms f1.0 0=no 1=yes lmucwtto log-10 of post-challenge mucus weight f4.2 lpremuc log-10 of pre-challenge mucus weight f4.2 virus challenge virus 3=RV2 4=RV9 5=RV14 6=RSV 7=CV f2.0 serostat serologic status f1.0 0=negative 1=positive sex gender f1.0 0=male 1=female agecat age (dichotomized) f1.0 0=18-33 1=34-54 educatn education f1.0 range is 0 to 8 (no schooling to doctoral degree) allergy S allergic to food, drugs, etc. f1.0 0=no 1=yes daylite number of hours of daylight f5.2 nmbrmate number of roommates f1.0 range is 0 to 2 mateinf S had an infected roommate f1.0 0=no 1=yes ponderal weight/(height*height*height) f5.2 smokstat smoking status f1.0 0=nonsmoker 1=smoker lavecot log-10 of average cotinine level f4.2 drnkrate ave. # of alcoholic drinks per day f5.2 drnkcat categorized daily drinking rate f1.0 0=0 1=(0,1] 2=(1,2] 3=more than 2 weekdrk weekday drinking rate f2.0 wkenddrk weekend drinking rate f2.0 diet diet index f5.2 (higher score is better) aerobic exercise index f4.2 (higher score is better) sleepqul quality of sleep index f2.0 (higher score is better) lgmonocy log-10 # of monocytes f3.2 lgneutro log-10 # of neutrophils f3.2 lglympho log-10 # of lymphocytes f3.2 lgsigato log-10 total serum IgA level f4.2 lgnigato log-10 total nasal IgA level f4.2 lgsigeto log-10 total serum IgE level f4.2 lgnigeto log-10 total nasal IgE level f4.2 lgnprotn log-10 total nasal protein level f4.2 sfesteem self-esteem score f5.2 (higher score = more self-esteem) control personal control f6.2 (higher score = more control) extravsn extraversion f5.2 (higher score = more extraverted) psi psychological stress index f2.0 (higher score = more stress) lgtotsle log-10 total # stressful life events f4.2 slecat categorized # stressful life events f1.0 0=0 to 2 1=more than 2 pss perceived stress scale score f5.2 (higher score = more stress) negmood negative affect score f5.2 (higher score = more negative affect) ---------------------------------------------------------- Source: data kindly supplied to jh by Sheldon Cohen on condition that they be used exclusively for teaching purposes. See Cohen S et al. "Psychological Stress and Susceptibility to the Common Cold" New Eng J of Med 1991; 325:606-612. Cohen S et al. "Negative Life Events, Perceived Stress, Negative Affect and Susceptibility to the Common Cold" Journal of Personality and Social Psychology 1993; 64(1) 131- 140. Cohen S et al. "Smoking, Alcohol Consumption and Susceptibility to the Common Cold" Am J Public Health 1993; 83(9) 1277-1283. jh 1997.01.05 Summary, Introduction and Methods sections of "Negative Life Events, Perceived Stress, Negative Affect, and Susceptibility to the Common Cold" Sheldon Cohen, David A. J. Tyrrell, and Andrew P. Smith Journal of Personality and Social Psychology 1993; 64(1) 131- 140. After completing questionnaires assessing stressful life events, perceived stress, and negative affect, 394 healthy Ss were intentionally exposed to a common cold virus, quarantined. and monitored for the development of biologically verified clinical illness. Consistent with the hypothesis that psychological stress increases susceptibility to infectious agents, higher scores on each of the 3 stress scales were associated with greater risk of developing a cold. However, the relation between stressful life events and illness was mediated by a different biologic process than were relations between perceived stress and illness and negative affect and illness. That these scales have independent relations with illness and that these relations are mediated by different processes challenges the assumption that perceptions of stress and negative affect are necessary for stressful life events to influence disease risk. It is commonly believed that life stressors increase susceptibility to infectious disease. When demands imposed by events exceed ability to cope, a psychological stress response is elicited (Lazarus & Folkman, 1984). This response is composed of negative cognitive and emotional states. In turn, these states are thought to alter immune function through autonomic nerves that connect the central nervous system to immune tissue (D. L. Felten, Felten, Carlson, Olschowka, & Livnat, 1985; S. Y. Felten & Olschowka, 1987), through the action of hormones whose release is associated with negative affectivity (Shavit, Lewis Terman, Gale, & Liebeskind, 1984), or through stress-elicited changes in health practices such as smoking and alcohol consumption (Cohen & Williamson, 1991; Kiecolt-Glaser & Glaser, 1988). Direct connections between stress and varisus functions of the immune system have been found in both field (e.g., Kiecolt Glaser & Glaser, 1991) and laboratory settings (e.g., Manuck, Cohen, Rabin, Muldoon, & Bachen, 1991; Naliboffet al., 1991). However, it is unclear whether the immune changes related to stress in these studies are of the type or magnitude that would influence susceptibility to infection (Jemmott & Locke,1984; Laudenslager, 1987). There is also research directly assessing the relation between stress and upper respiratory infections in community samples (see review in Cohen & Williamson, l 991). Several of these studies provide evidence that social stressors increase risk for verified upper respiratory disease (Graham, Douglas, & Ryan,1986; Meyer & Haggerty,1962). This work, however, did not control for the possible effects of stressful events on exposure to infectious agents (as opposed to their effects on host resistance) or provide evidence about other behavioral and biologic mechanisms through which stress might influence susceptibility to infection. Moreover, the literature on this topic is not entirely consistent with several studies failing to find a relation between stress and respiratory disease (Alexander & Summerskill, 1956; Cluff, Cantor, & Imboden, 1966). Viral-challenge studies, in which volunteers who complete stress scales are intentionally exposed to a cold or influenza virus, have provided only weak support for a relation between stress and susceptibility to upper respiratory infections (Broadbent, Broadbent, Phillpotts, & Wallace, 1984; Totman, Kiff, Reed, & Craig,1980; Greene, Betts, Ochitill, Iker, & Douglas, 1978; Locke & Heisel,1977). This work, however, suffers from a wide range of methodological flaws (Cohen & Williamson, 1991) Individual studies suffer from insufficient sample sizes, concurrent administration of drugs, lack of information on overall rates of infection in response to the dose of virus administered, and lack of controls for important predictors of susceptibility such as preexisting antibodies to the infectious agent, gender, and age (see Jackson et al., l960). They also fail to control for the possible role of stress- elicited changes in health practices such as smoking and alcohol consumption. Part of the problem in establishing a relation between stress and new cases of disease is that there is little agreement within or across disciplines on how stress should be defined or measured. As addressed earlier, this article is concerned with psychological stress, that is, negative cognitive and emotional states elicited when persons perceive that their demands exceed their ability to cope (Lazarus & Folknman, 1984). However, even within this constrained area, there is considerable controversy as to how such a state should be assessed (e.g., Cohen, 1986; Dohrenwend & Shrout, 1985; Lazarus, DeLongis, Folkman, & Gruen, 1985). In an article published in a medical journal (Cohen, Tyrrell, & Smith, 1991), we reported that psychological stress, operationally defined as an index including negative life events, perceived stress, and negative affect, predicted susceptibility to colds among 394 initially healthy persons we intentionally exposed to upper respiratory viruses. We justified the use of the stress index on the basis of a factor analysis indicating that the three measures formed a single principal component, providing evidence that the scales measure a common underlying concept. By using a single index, we were able to substantially reduce the number of analyses and hence Type I error. The psychological stress index provided a robust and valid measure in that it showed a linear relation with risk for developing colds, and in that the relation between the stress index and susceptibility was unaffected by controls for a series of environmental, psychological, behavioral, and immunological factors. However, combining the three scales masked information that examination of the individual scales might tell us about the stress process. Although the scales are correlated with one another, each taps a somewhat different component of the psychological stress experience. In this article, we present further data from our prospective study of persons intentionally exposed to upper respiratory vi ruses. Healthy persons were administered the three stress scales, three personality scales, and measures of health practices, and then they were experimentally exposed to one of five cold viruses or placebo. The association between stress and the development of biologically verified clinical disease was examined with use of a control for baseline (prechallenge) antibodies to the challenge virus, the identity of the challenge virus, allergic status, weight, the season, the number of subjects housed together, the infectious status of any subjects sharing housing, and various demographic factors. We examine the relation between each of the three separate stress scales and risk for clinical colds, evaluate potential pathways through which each might influence susceptibility, and discuss the differences in terms of the components of psychological stress that each of the scales assess. Method The subjects were 154 men and 266 women who volunteered to participate in trials at the Medical Research Council's Common Cold Unit (CCU) in Salisbury, England. All reported no chronic or acute illness or regular medication regimen on their applications and were judged in good health following clinical and laboratory examination on arrival at the unit. Pregnant women were excluded. Volunteers' ages ranged from l8 to 54 years,with a mean of 33.6 and standard deviation of 10.6. Twenty-two percent did not complete their secondary education, 51% completed secondary education but did not attend a university, and 27% attended a university for at least 1 year. Volunteers were reimbursed for their traveling expenses and received free meals and accommodations. The trial was approved by the Harrow District Ethical Committee and informed consent was obtained from each volunteer after the nature and possible consequences of the study were fully explained. During their first 2 days at the CCU, volunteers were given a thorough medical examination, completed a series of self- reported behavioral protocols including psychological stress, personality, and health practice questionnaires, and had blood drawn for immune and cotinine assessments. Subsequently, volunteers were exposed using nasal drops, to a low infectious dose of one of five respiratory viruses: rhino virus types 2 (RV2; n = 86), 9 (RV9; n = 122), and 14 (RV14; n = 92) respiratory syncytial virus (RSV; n = 40), and coronavirus type 229E (CV; n = 54). An additional 26 volunteers received saline. One or two viruses were used in each individual trial, and volunteers were randomly assigned to virus and saline conditions. Viral doses were in tended to simulate those that occur in person-to-person transmission, and they resulted in illness rates of between 20% and 60%. For 2 days before and 7 days after viral challenge, volunteers were quarantined in large apartments (alone or with 1 or 2 others). Starting 2 days before viral challenge and continuing through 6 days after the challenge, each volunteer was examined daily by a clinician using a standard respiratory sign-symptom protocol (Beare & Reed, 1977). Examples of items on the protocol include sneezing, watering of eyes, nasal stuffiness, nasal obstruction, postnasal discharge, sinus pain, sore throat, hoarseness, cough, and sputum. The protocol also included an objective count of the number of tissues used daily by a volunteer and body temperature (oral) assessed twice each day. Approximately 28 days after the viral challenge a second serum sample was collected by volunteers' own physicians and shipped back to the CCU. All investigators were blind to volunteers' psychological status and to whether they received virus or saline. Psychological Stress Three kinds of measures of psychological stress were used: (a) number of major stressful life events judged by the respondent as having a negative impact, (b) perception that current demands exceed capabilities to cope, and (c) current negative affect. The major stressful life events scale consisted of events that might happen in the life of the respondent (41 items) or close others (26 items). The events were a subset of those appearing in the List of Recent Experiences (Henderson, Byrne, & Duncan-Jones, 1981 ) and were chosen because of their potential for negative impact and the relatively high frequency of occurrence in population studies. Respondents were asked which of the items had occurred during the last 12 months. They were asked to rate each event they reported as having either a positive or negative impact on their lives. A few items such as death of a spouse or child were assumed to be consensually negative, and the respondent was not asked for an impact rating. The scale score was the number of negative events (either consensual or respondent rated) reported by the subject. Because the scores were highly skewed (57% reported two or fewer events, range was 0-14), with some irregularities in the smoothness of the distribution, we used two different approaches to life-event analyses. In the first, we retained the continuous scaling by using a log1O transformation of the raw data. In the second, we transformed life events into a dichotomized variable: two or fewer events versus more than two events. The lO-item Perceived Stress Scale (PSS-lO;Cohen & Williamson, 1988) was used to assess the degree to which situations in life are perceived as stressful (reliability in this sample, alpha = .85). Items in the PSS- 10 were designed to tap how unpredictable, uncontrollable, and overloading respondents find their lives. finally, the negative affect scale included 15 items from Zevon and Tellegen's (1982) list of negative emotions. The items included distressed, nervous, sad, angry, dissatisfied with self, calm (reverse scored), guilty, scared, angry at yourself, upset, irritated, depressed, hostile, shaky and content (reverse scored). A 5- point (0-4) Likert-type response format was used to report affect intensity during the last week (alpha = .84). Both the PSS-1O and negative affect scores were approximately normally distributed, and their raw (continuous) scores were used as predictors in the regression analyses. Stressful life events were correlated .35, p < .001, and .35 p < .001, with PSS- 10 and negative affect, respectively. Negative affect was correlated .65, p < .001, with the PSS-10. Point biserial correlations between dichotomized stressful life events and PSS-10 and life events and negative affect were .32, p < .001, and .32, p < .001, respectively. For comparison, we also present data based on analyses using the psychological stress index used in our earlier article (Cohen, Tyrrell, & Smith, 1991). All three stress scales formed a single principal component, with loadings of .66, .86, and .86, respectively, providing evidence that the scales measure a common underlying concept. Hence we formed an index combining the three measures as an indicator of psychological stress. Because life events were not normally distributed, an index based on normalized scores was not appropriate. Instead, the index was created by quartiling each scale and summing quartile ranks for each subject (I for lowest quartile and 4 for highest), resulting in a stress index ranging from 3 to 12. The quartiles were 0, 1-2, 3-4, and 5-14 for the life- events scale; 0-10, 11-14, 15-18, and 19-33 for the PSS- 10; and 0-7, 8-13, 14-20, and 21-49 for the negative affect scale. Index scores were approximately normally distributed. Infections and Clinical Colds Infectious diseases result from the growth and action of microorganisms or parasites in the body (see Cohen & Williamson, 1991). Infection is the multiplication of an invading microorganism. It is possible for a person to be infected (for the microorganism to replicate) without developing clinical symptoms. Clinical disease occurs when infection is followed by the development of symptomatology characteristic of the disease. Biological verification of infection can be accomplished by establishing that an infectious agent is present or replicating in tissue, fluid, or both. We use two common procedures for detecting a replication of a specific virus. In the viral isolation procedure, nasal secretions are cultured (put in a medium that stimulates virus replication). If the virus is present in nasal secretions, it will grow in the medium and can be detected. Alternatively, we can indirectly assess the presence of a replicating virus by looking at changes in serum antibody levels to that virus. Antibodies are protein molecules that attach themselves to invading microorganisms and mark them for destruction or prevent them from infecting cells. An invading microorganism (i.e., infection) triggers the immune system to produce antibody. (Antibody is also called immunoglobulin [Ig]). Because each antibody recognizes only a single type of microorganism, the production of antibody to a specific infectious agent is evidence for the presence and activity of that agent. Assays for viral isolation and viral-speciffc antibody levels. Nasal wash samples for viral isolation were collected before inoculation and on Days 2-6 after viral inoculation. They were mixed with broth and stored in aliquots at -70C. Rhinoviruses were detected in O-Hela cells, respiratory syncytial virus in Hep2 cells, and coronavirus in the C- 16 strain of continuous human fibroblast cells. When a characteristic cytopathic effect was observed, the tissue culture fluids were passaged into further cultures and identity tests on the virus were performed. Rhinoviruses and coronaviruses were confirmed by neutralization tests with specific rabbit immune serum, and respiratory syncytial virus by immunofluorescent staining of culture cells. Levels of neutralizing antibodies and of specific antiviral immunoglobulin A (IgA) and immunoglobulin G (IgG) were determined before and 28 days after the viral challenge. Neutralizing antibodies (for rhino viruses only) were determined by neutralization tests with homologous virus (Al Nakib & Tyrrell, 1988). Results were recorded as the highest dilution showing neutralization, and a fourfold rise was regarded as significant. Suitable neutralizing tests were not available for IgA and IgG levels for rhinoviruses (Barclay & Al Nakib, 1987), CV (Callow, 1985), and RSV (Callow, 1985) were determined by enzyme-linked immunosorbent assays. This test detects antibody which correlates with neutralization titers, is associated with resistance to infection, and increases in response to infection (Al Nakib & Tyrrell, 1988). Operational definitions of infection and clinical colds. A volunteer was deemed infected if a virus was isolated after the challenge or if there was a significant rise in viral- specific serum antibody after the challenge, that is, a fourfold increase in neutralizing antibody (rhino viruses only) or an IgG or IgA increase of two standard deviations greater than the mean of nonchallenged volunteers (all viruses). Eighty-two percent (325) of the volunteers receiving virus were infected. Nineteen percent (5) of the volunteers receiving saline were infected. We attributed infections among the saline group to volunteer transmission of virus to others housed in the same apartment. A control for person-to-person transmission is included in the data analysis. At the end of the trial, the clinician judged the severity of each volunteer's cold on a scale ranging from nil (0) to severe (4). Ratings of mild cold (2) or greater were considered positive clinical diagnoses. Volunteers also judged the severity of their colds on the same scale. Clinician diagnosis was in agreement with self-diagnosis for 94% of the volunteers. Volunteers were defined as having developed clinical colds if they were both infected and diagnosed by the clinician as having a clinical cold. Of the 394 volunteers participating in the trials, 38% (148) developed clinical colds. None of the 26 saline controls developed colds. Seven persons with positive clinical diagnoses but no indication of infection were excluded from the sample because we assumed the illness was caused by pretrial exposure to another virus. Analyses including them (by definition no clinical cold and no infection) resulted in identical conclusions. Body Temperature and Mucus Weights Because clinical diagnoses can be influenced by how subjects present their symptoms, we independently evaluated the associations between each stress measure and two clinical signs not subject to self-presentation bias: body temperature and mucus weights. Body temperature (degrees centigrade) was taken each morning and afternoon with an oral thermometer. An average daily temperature ([morning + after noon ] /2) was calculated for the day before and each day following the viral challenge. Mucus weights were determined by collecting tissues used by subjects in sealed plastic bags. The bags were weighed, and the weight of the tissues and the bags was subtracted. Daily mucus weights (in grams) were calculated for both before and after the viral challenge. The prechallenge measure was based on the mucus weight on the day before challenge. The postchallenge measures were based on the weights from Days I to 5 after the challenge. To obtain an approximately normally distributed variable, we used the log10 of both pre and postchallenge mucus weights in the analyses. Standard Control Variables We used a series of control variables that might provide alternative explanations for a relation between stress and illness. These include prechallenge serostatus for the experimental virus, age, gender, education, allergic status, weight, season, number of others the volunteer was housed with, whether an apartment mate was infected, and challenge virus. Prechallenge serostatus refers to whether a subject had antibody to the virus before experimental exposure, that is, was previously exposed to the virus. Serostatus was defined as positive when a volunteer had a neutralizing prechallenge antibody titer greater than 2 for rhinoviruses and a prechallenge antibody level greater than the sample median for CV and RSV. Forty-three percent of volunteers were seropositive before the challenge, including 55% for RV2, 48% for RV9, 20% for RV14, 50% for RSV, and 50% for CV. Age and gender were based on self-report. Because age was not normally distributed it was scored categorically on the basis of a median split: 18-33 or 34-54. Scores on education were based on a 9-point self-report scale ranging from no schooling (O) to doctoral degree (8). Allergic status was based on physician interview questions regarding allergies to food, drugs, or other allergens. Persons reporting any allergy were defined as allergic. A ponderal index (weight/height cubed) was used to control for volunteers' weight. We used the number of hours of daylight on the first day of the trial as a continuous measure of the season. Number of daylight hours was correlated .80 (<0.001) with the average temperature on the same day. A control for the possibility that person-to-person transmission rather than the virus inoculation might be responsible for infection or clinical colds was also included. Because person-to-person transmission would only be possible if an apartment mate was infected by the viral challenge, the control variable indicated whether any housing mate was infected. Finally, challenge virus is a categorical variable representing the experimental virus to which a volunteer was exposed. Health Practice Measures Health practices including smoking, drinking alcohol, exercise, quality of sleep, and dietary practices were assessed as possible pathways linking stress and susceptibility. Cotinine as assessed in serum by gas chromatography was used as a biochemical indicator of smoking rate because it provides an objective measure of nicotine intake that is not subject to self-report bias (Feyerabend & Russell, 1990; Jarvis, Tunstall-Pedoe, Feyerabend, Vesey, & Saloojee, 1987). We used the log10 of the average of the two (before and 28 days after challenge) cotinine measures as an indicator of smoking rate.(The correlation between the two measures was .95, p < .001, N= 348). The correlation between the log10 average cotinine and the log10 self-reported number of cigarettes smoked per day was .96 (p < .001. N= 372). The remaining health practices were assessed by questionnaire before the viral challenge (see Cohen et al., 1991). Average number of alcoholic drinks per day was calculated using separate estimates of weekday and weekend drinking. A half pint, bottle. or can of beer, glass of wine, and shot of whisky contain approximately equal amounts of alcohol and were each treated as a single drink. We use the log10 of average number of drinks per day as an indicator of drinking. The exercise index included items on the frequency of walking, running, jogging, swimming, aerobics, and work around the house. The quality of sleep index included items on feeling rested, difficulty falling asleep, and awakening early; and the dietary habit index was made up of items designed to assess concern with a healthy diet and included frequencies of eating breakfast, fruits, and vegetables. Personality Measures Because psychological stress could reflect stable personality styles rather than responses to environmental stressors, self-esteem and personal control (two personality characteristics closely associated with stress) were assessed before the viral challenge. Self-esteem was measured by the self-regard and social confidence subscales of the Feelings of Inadequacy Scale (Fleming & Watts, 1980; alpha = .89), and personal control by the personal efficacy and interpersonal control subscales of the Spheres of Control Scale (Paulhus, 1983; alpha = .76). A third personality characteristic, introversion-extraversion was also assessed because of an existing literature suggesting that introverts were at higher risk for infection (Broadbent et al., 1984; Totman et al, 1980). It was assessed by the Eysenck Personality Inventory (Eysenck & Eysenck, 1964; alpha = .80). As expected, none of the saline control subjects developed colds and hence the analyses include only persons receiving a virus. The primary analysis tests whether each of the psychological stress measures is associated with greater rates of clinical colds. Secondary analyses assess the importance of the two components of the definition of a clinical cold, infection and illness (clinical symptomatology), in accounting for an association of stress and clinical colds. Specifically, we determined whether the relation between stress and colds was attributable to in creases in infection or to increases in diagnosed colds among infected persons. Logistic regression was used to predict these dichotomous outcomes (Hosmer & Lemeshow, 1989). We used a regression procedure that provides estimated regression coefficients for each independent variable adjusted for all other variables in the equation. Probability values are based on the change in log-likelihood that would result if each variable were entered as the last variable in a stepwise regression. This is analogous to testing whether a variable accounts for a significant increment in explained variance in a linear regression model. We report a sequential series of analyses. In the first analysis, only the psychological stress measure was entered as a predictor. In the second, we entered the standard control variables along with the stress measure and tested whether there was a significant change in log-likelihood when the stress index was added to the equation. Education, weight, season, and number of apartment mates were entered as continuous variables and the remainder of the standard controls as dummy (categorical) variables (Hosmer & Lemeshow, 1989). Additional analyses tested possible roles of health practices and personality variables in the relation between stress and clinical colds. In all cases, we added these variables to a regression equation that included standard controls, and we report the adjusted coefficient and the probability value based on the change in log-likelihood when the stress measure is added to the equation. These analyses are extremely conservative, testing the significance of the stress measures after adding as many as 16 control variables. Repeated measures analysis of covariance was used in supplementary analyses of continuous outcomes: postchallenge mucus weights and body temperature. In each case, prechallenge scores and standard control variables were used as covariates. ----------------------------------------- Introduction section of "Smoking, Alcohol Consumption, and Susceptibility to the Common Cold" Sheldon Cohen, David A. J. Tyrrell, Michael A H Russell, Martin J Jarvis and Andrew P. Smith Am J Publ Health 1993; 83:1277-1283. Both smoking and alcohol consumption are believed to suppress host resistance and thereby increase the risk of upper respiratory infections. In the case of smoking, epidemiological studies indicate an increased risk of serologically confirmed influenza among otherwise healthy smokers(1-3) although a study of rhinovirus colds failed to find a relation between smoking and risk of illness.(4) Because exposure to infectious agents was not controlled for in the epidemiological work, it is possible that the increased number of influenza cases among smokers was attributable to smokers having more close contacts with infected persons rather than to decreased host resistance. Moreover, the recording of illnesses in all these studies depended on persons seeking medical care. Smokers may be more likely to seek medical care when they are bothered by mild symptoms because they are aware of being at risk for serious respiratory disease. Alternatively, they may be less likely to seek care because they regard mild respiratory symptoms as normal. Although the use of alcohol, especially in excess, is generally viewed as being immunosuppressive, it is not clear that documented alcohol-elicited changes in immune function are of clinical significance (5-6) There are data supporting an increased incidence of bacterial infections among alcoholics.(7-5) However, these relations are often attributed to complications of alcoholism, including nutritional deficiencies, alcoholic cirrhosis, hygienic factors, and life-style.(5-9) Up until now, the relation between alcohol consumption and susceptibility to common upper respiratory infections in healthy, nonalcoholic humans has not been studied. We present data from a prospective study of the independent effects of smoking and alcohol consumption on susceptibility to the common cold. Healthy persons were questioned about their smoking and alcohol consumption, and had blood drawn for immune and cotinine assessments. Subsequently, they were exposed to one of five respiratory viruses and monitored for the development of clinical illness. Demographic, environmental, immunological, and psychological variables that might provide alternative explanations for associations between drinking and smoking and the incidence of biologically verified colds were also assessed and controlled for in data analyses. By experimentally exposing subjects to viral agents, we eliminated the possibility that smoking or drinking resulted in exposure to the infectious agent. Moreover, by closely monitoring for the development of infection and illness, we eliminated biases that can occur when persons must seek medical care to receive a diagnosis. ------------------------------------------ Abstract and Introduction sections of "Psychological Stress and Susceptibility to the Common Cold" Sheldon Cohen, David A. J. Tyrrell, and Andrew P. Smith Abstract ======== BACKGROUND. It is not known whether psychological stress suppresses host resistance to infection. To investigate this issue, we prospectively studied the relation between psychological stress and the frequency of documented clinical colds among subjects intentionally exposed to respiratory viruses. METHODS. After completing questionnaires assessing degrees of psychological stress, 394 healthy subjects were given nasal drops containing one of five respiratory viruses (rhinovirus type 2, 9, or 14, respiratory syncytial virus, or coronavirus type 229E), and an additional 26 were given saline nasal drops. The subjects were then quarantined and monitored for the development of evidence of infection and symptoms. Clinical colds were defined as clinical symptoms in the presence of an infection verified by the isolation of virus or by an increase in the virus-specific antibody titer. RESULTS. ... Introduction ============ STRESSFUL life events are commonly believed to suppress host resistance to infection. When demands imposed by events exceed a person's ability to cope, a psychological stress response composed of negative cognitive and emotional states is elicited. Psychological stress, in turn, is thought to influence immune function through autonomic nerves innervating lymphoid tissue or hormone-mediated alteration of immune cells. Stress may also alter immune responses through the adoption of coping behaviours such as increased smoking and alcohol consumption. There is substantial evidence that stressful life events and perceived stress are associated with changes in immune function. Although psychological stress is often described as suppressing immune response, the implications of stress- induced immune changes for susceptibility to disease have not been elucidated. There is some direct evidence from previous studies that psychological stress increases the risk of verified acute infectious respiratory illness. These studies, however, did not control for the possible effects of stressful events on exposure to infectious agents (as opposed to their effects on resistance) or provide evidence about other behavioral and biologic mechanisms through which stress might influence a person's susceptibility to infection. Moreover, the literature on this topic is not entirely consistent; several studies have failed to find a relation between stress and respiratory disease. We present data from a prospective study of the association between psychological stress and susceptibility to the common cold. Healthy persons were assessed for degree of stress and then experimentally exposed to one of five cold viruses (394 subjects) or placebo (26 subjects). The association between stress and the development of biologically verified clinical disease was examined with use of control for base-line (prechallenge) serologic status, the identity of the challenge virus, allergic status, weight, the season, the number of subjects housed together, the infectious status of any subjects sharing housing, and various demographic factors. In further analyses we tested the possibility that a relation between stress and susceptibility to illness could be attributed to differences in health practices or differences in base-line white-cell counts or total antibody levels. A final analysis investigated the possibility that differences in personality rather than environmental factors causing stress might account for the association between stress and clinical colds. ...