Personality and Individual Differences 34 (2003) 971–982 www.elsevier.com/locate/paid Stress and illness in American and Russian college students
Dmitri Poltavski, F. R. Ferraro*
Department of Psychology, University of North Dakota, Box 8380, Grand Forks, ND 58202-8380, USA
Received 15 June 2001; received in revised form 24 January 2002; accepted 24 March 2002

Abstract
The differences in perceptions of potentially stressful events among 139 college students from Russia and
USA were investigated in relation to somatic illness. Individual interpretations were assessed using the Life
Events Survey. The instrument was administered in conjunction with the State-Trait Anxiety Index.
Results indicated more perceived stress by the Russian student sample but more frequent incidence of reported illness in the American student sample. American female college students showed greater anxiety levels and more perceived stress in comparison to male students. Females also reported a greater number of illness occurrences. # 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Stress; Illness; Cross-cultural; Russia; College students

1. Introduction
For centuries, the maintenance of the internal physiological equilibrium has been considered vital for the well-being of any living organism. Having subsumed both emotional and physiological states of equilibrium under the term ‘‘homeostasis’’ Cannon (1932) thus brought psychology into the realm of physiology. Selye (1976) expanded the notion of homeostatic disregulation brought upon by stressors of both physical and mental nature and proposed a mechanism of adaptation of an organism to the demands of its environment. Specifically, Selye believed that brief, mild and controllable shifts in homeostasis triggered efficient adaptive physiological, cognitive and behavioral responses that could actually be beneficial to the organism in its development. Only situations characterized by more severe, uncontrollable and persistent stressors were considered by Selye as deleterious to both psychological and physical conditions of the organism.
Selye grouped various disease states resultant from overexposure to different stressors under the notion of the ‘‘General Adaptation Syndrome’’.
* Corresponding author. Tel.: +1-701-777-2414; fax: +1-701-777-3454.
E-mail address: f_ferraro@und.nodak.edu (F. R. Ferraro).
0191-8869/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0191-8869(02)00082-X

972

D. Poltavski, F. R. Ferraro / Personality and Individual Differences 34 (2003) 971–982

Neurophysiological studies provided evidence in support of the mechanism of the stress response and stress-induced illness postulated by Selye (Chrousous & Gold, 1992). Specifically, it has been shown that the general adaptational response following the onset of an objectively or subjectively stressful stimulus is characterized by two principal components: the release of the corticotropin-releasing hormone (CRH) particularly by the neurons of the paraventricuar nucleus in the hypothalamus (Swanson, Sawchenko, Rivier, & Vale, 1983) and the activation of the locus ceruleus- noradrenergic system in the brain stem (Chrousous & Gold, 1992). While the former component leads to the activation of the pituitary and adrenal glands with subsequent excitation of the sympathetic nervous system, the latter initiates a series of noradrenergic firings throughout a dense network of neurons in the brain resulting in enhanced arousal, vigilance, as well as increased anxiety (Dunn & Berridge, 1990). It has been demonstrated that catabolic reactions of the organism that generally promote enhanced availability of energy resources to cope with a stressful stimulus are triggered by the release of the CRH, and become sufficiently maladaptive in the long run if the stressor or a series of stressors persist (e.g. Gold, Goodwin, & Chrousos, 1988).
Human research has shown susceptibility to upper respiratory infection following the onset of undesirable events in the lives of the participants (Stone, Reed, & Neale, 1987). The researchers suggested that decreased production of a particular immuoglobulin (IgA) could be linked to a weakened immune function associated with an increasing likelihood of an upper respiratory illness. In another study (Cohen, Tyrrell, & Smith, 1991) nasal drops containing a cold virus were administered to the participants and the incidence of illness was recorded. The researchers found that the likelihood of developing a cold was significantly greater if the subjects reported experiencing stressful events in the past year during which they did not feel in control, were overwhelmed and/or threatened. In medical students the highest incidence of acute infections was found to coincide with the time of their final examinations, which has been linked to the suppression of the immune system (Glaser et al., 1987).
Animal studies have shown a positive correlation between stressor intensity and the severity of the preexisting heart condition in hamsters (McCabe, Sheridan, Weiss, Kaplan, Natelson, & Pare,
2000). Psychological factors such as stress predictability and control have also been found critical in gastric ulcer development in rats, which exhibited fewer ulcers when electric shock was predictable and/or could be controlled (Seligman, 1968; Weiss, 1971). Social status has also been shown to play an important role in the development of arteriosclerosis in male cynomolgus monkeys when exposed to an unstable social environment (Manuck, Marsland, Kaplan, & Williams, 1995).
The relevance of psychological and social factors to somatic illness has been emphasized in
Lazarus and Folkman’ s (1987) contextual model of stress. According to this model perception of a stimulus as a stressor occurs only when its initial appraisal renders harmful relevance to the person’s well-being. The model also suggests initiation of behavioral coping following secondary appraisal of the adequacy of the resources available to meet the demands imposed by the stressor.
1.1. The present study
Since psychological and social factors (e.g. stressor predictability and control, social status, stability of the environment) have been shown to at least aggravate preexisting somatic conditions in both animals and humans, and while personal interpretations of stressors are important

D. Poltavski, F. R. Ferraro / Personality and Individual Differences 34 (2003) 971–982

973

in behavioral responses, then similar environments could lead to different behavioral outcomes including pathology, and conversely, similar behaviors could be observed under different environmental/social conditions. In the present study the researchers have chosen to focus on the latter possibility, i.e. to investigate the role of stress in somatic illness across two culturally different environments: American and Russian.
The USA and Russia have been shown to differ on the individualism-collectivism scale (Triandis, 1995). The characteristics of individualism and collectivism are seen as expressions of individual personality traits such as ideocentrism and allocentrism on the societal level (Searle-White,
1996). While the former refers to the priority of personal goals over those of a group, self-reliance and competitiveness, the latter describes someone who shares his/her own sphere of responsibility with others, and is a functional member of a group endorsing interdependence (Searle-White,
1996). In this conceptualization, individuals within a culture would show variability along these personal traits, but overall it is expected that in a collectivist culture there would be more people who score higher on allocentrism while the same would hold true for idiocentrism in an individualist culture (Searle-White, 1996). Russians have been found to think in terms of interdependence, and crossing between spheres of responsibility, whereas Americans tend to think in terms of personal independence, and autonomy in the sphere of control (Triandis, Bontempo,
Betancourt, & Bond, 1986). In a more recent study by Brief and Comrey (1993) an attempt has been made to draw a personality profile for a Russian sample. By using the Comrey Personality
Scales (CPS) the researchers found that the Russian sample as a whole tended to be significantly more lethargic, neurotic and defensive than the American one but not more introverted or lacking in compulsion. Despite earlier reports of interdependence and crossing between spheres of responsibility among Russians (Triandis, 1988) an unexpected finding of the study was that the
Russian sample was significantly more egocentric than the American respondents. One possible explanation of this finding could be a changing social infrastructure that after the collapse of the
Soviet Union in1992 has been developing with orientation towards the West. Balatsky and
Diener (1993) also reported lower psychological well-being among Russians, while Kassinove and
Sukhodolsky’s findings (1995) indicated that Russians are more pessimistic.
Despite the above reports of negative affect (pessimism, lethargy, neuroticism) Russian college students do not seem to exaggerate either their positive or negative feelings (Kassinove &
Eckhardt, 1994). The researchers found that American college students reported more negative and positive affect, while Russians scored significantly higher (more rational) on the subscales of the Survey of Personal Beliefs. On the basis of the findings, the researchers suggested that either
Americans tend to socialize through a greater affective range (e.g. tendency to catastrophize), or
Russians are less likely to magnify negative life events possibly due to their upbringing in the society that did not encourage expression of strong feelings which could have affected their responses on the paper and pencil tests, which they might have treated with a degree of suspicion.
Since objectively Russians might be experiencing more stressful life events, which could be of higher intensity on both social and political levels (Jose et al., 1998), their susceptibility to stressrelated somatic illness might be greater than that of the Americans if their coping strategies are not significantly different from those of Americans. The findings of a reduced emotional response among Russians (Kassinove & Eckhardt, 1994) suggest differences in the copying style. Russian adolescents have also been found to report using more problem solving and social support strategies than their American counterparts (Jose et al., 1998).

974

D. Poltavski, F. R. Ferraro / Personality and Individual Differences 34 (2003) 971–982

Behavioral coping can also involve substance abuse. A greater proportion of Russian adolescents in comparison with the American youth, for example, have been reported to start using alcohol before age 12 (Parker, Levin, & Harford, 1996). More prevalent use of alcohol and tobacco products has also been observed in Russian adults (Bobak, McKee, Rose, & Marmot,
1999).
In view of this evidence it was expected that a stronger relationship between self-reported stress and illness in the American group, and lower levels of reported stress in the Russian sample would be observed. Nevertheless, since cigarette smoking and alcohol consumption have been linked to the induction of stress-like physiological processes and since Russians have been reported to use these substances significantly more than Americans, it was expected that Russians would also report a greater number of illness occurrences. Additionally, it was hypothesized that the anxiety scores of the Russian participants would be lower than those of the American students. No gender differences were predicted.

2. Method
2.1. Participants
A total of 68 Russian participants (34 males and 34 females), ages 20–25, was obtained from the
Moscow State Teacher Training University. The volunteers were recruited in the department of
English and Linguistics and were required to have English proficiency of at least 525 as measured by TOEFL (Test of English as a Foreign Language) taken by the students independently of the research project. This score corresponds to the proficiency level required for undergraduate admission to the University of North Dakota (UND Academic catalog, 1999–2001).
The American sample comprised 71 American college students (25 males and 46 females), ages
21–25, recruited from the student population of the University of North Dakota. Efforts were made to recruit non-dormitory residents since none of the participants in the Russian sample resided in dormitories. The participants were also selected on the basis of their field of study
(humanities) to match the homogeneous composition of the Russian group. Participation was not contingent on any monetary or academic rewards.
2.2. Materials
A modified version of the Life Experience Survey (LES) originally developed by Sarason,
Johnson, and Siegel (1978) was administered as a measure of life changes. The questionnaire was chosen on the basis of its provisions for individualized ratings of fairly ambiguous events such as
‘‘change of residence’’ on the scale from À3 to +3 thus indicating not only the intensity of the stressor but also its desirability. The rating of ‘‘À3’’ indicated that an event was extremely unfavorable to the respondent while a ‘‘+3’’ rating corresponded to an extremely favorable impact. A rating of ‘‘0’’ showed no reported impact on the participant. The original LES contained 57 items divided between two sections: Section 1 designed for all respondents, comprised a list of 47 specific events plus three blank spaces in which participants could indicate other events that they may have experienced. Section 2 contained the remaining 10 items specifically designed to capture

D. Poltavski, F. R. Ferraro / Personality and Individual Differences 34 (2003) 971–982

975

life events in student populations. The LES items were chosen to represent life changes frequently experienced by individuals in the general population, and 34 of the listed events were similar to those found in the Schedule of Recent Experiences (SRE; Holmes & Rahe, 1967), previously widely used in life stress research. Sarason et al. (1978) added several gender-specific items dealing with such issues as abortion, serious injury or illness of a close friend, engagement, breaking up with a boyfriend/girlfriend, and so forth.
The instrument was modified in the present study to further adapt it to the student population and to include a potentially stress-mediating variable of adverse behavioral habits that was intended to assess nicotine and alcohol use in both samples. Several items from Section 1 of the original LES were removed (e.g. retirement from work, ending of formal schooling, son or daughter leaving home, etc.) as they were not directly applicable to the student population under investigation. At the same time the student section was expanded to include such items as problems with instructor, moving away from home for the first time, roommate conflict, holding an office in a student club, outstanding personal achievement (receiving good grades, making a sporting team, etc.).
The survey considered events in the retrospect of up to 6 months prior to the completion of the questionnaire. Three types of scores were calculated: a positive change score (the sum of the impact ratings of those events designated as positive by the participant), a negative change score
(sum of negative ratings), a total change score (derived by adding the two values for the positive and negative change)
Finally, a section on cigarette smoking and alcohol consumption was added. Cigarette subsection evaluated a number of cigarettes a person smoked generally and when feeling anxious, nervous, distressed, or upset. The alcohol subsection was similarly constructed and included a number of drinks (of established preference, e.g., beer, wine and/or hard liquor etc.) a person consumed regularly and in the state of anxiety. In the end of the questionnaire a 12-item list of common physical conditions was provided that included such items as common cold, influenza, headaches, diarrhea, tonsillitis etc., followed by three blank spaces to indicate episodes of sickness not specified. Each item on the sickness scale was rated in terms of its frequency of occurrence in the past 6 months and degree of severity (attended class anyway, too ill to attend class, went and saw a doctor, required a hospital stay).
The LES was administered in conjunction with the State-Trait Anxiety Index (STAI) (Spielberger, Gorsuch, & Lushene, 1970) as a standardized measure of anxiety levels. Anxiety levels were found to have a positive correlation with the levels of experienced stress (Lazarus, 1991), and the STAI was thus thought to be predictive of a stressful condition. The STAI clearly differentiates between the temporary condition of ‘‘state anxiety’’ and the more general and longstanding quality of ‘‘trait anxiety’’. The essential qualities evaluated by the STAI state anxiety scale are feelings of apprehension, tension, nervousness, and worry. This scale consists of 20 statements that evaluate how respondents feel ‘‘right now, at this moment’’. Scores on the state anxiety scale increase in response to physical danger and psychological stress, and decrease as a result of relaxation training (Spielberger et al., 1970). The STAI consists of 20 statements that assess how respondents feel ‘‘generally’’. On this scale, consistent with the trait anxiety construct, psychoneurotic and depressed patients normally exhibit high scores. Individuals respond to each item of both scales on a four-point Likert scale, indicating the magnitude of applicability of a particular statement to him/her (with 1 being the least applicable and 4 being very applicable). 976

D. Poltavski, F. R. Ferraro / Personality and Individual Differences 34 (2003) 971–982

2.3. Procedure
In previous cross-cultural studies with Russian samples (e.g. Kassinove & Eckhardt, 1994; Jose et al., 1998) the researchers used translated materials and employed a technique of back translation to ensure comprehensibility and comparability. It has been noted, however, that the translated measures were originally designed for the American population and therefore, significant aspects of Russians’ daily functioning might not have been captured (Jose et al., 1998).
Kassinove and Eckhardt (1994) also points out that the Russian language does not accommodate as many adjectives describing emotionality as does English which could indicate relative predilection towards emotionality in the American culture. It has been, therefore, decided to administer the LES in its original language (English) to both samples.
Participants in the American sample were approached individually outside the classroom on the campus territory and asked to fill out the LES and STAI questionnaires. The order in which the surveys were filled out was not deemed relevant and was, therefore, dependent on individual preference. Unlike the American group, the Russian sample was given questionnaires through their instructors both collectively (during a regular class meeting) and individually. In both conditions professors of English were available to clarify possible difficulties in comprehension.

3. Results
The data were analyzed with series of 2 (gender)Â2 (country) analyses of variance (ANOVA).
Between-group differences were assessed on the following dependent measures: years of education, scores on the subscales of the LES and scores on the subscales of the STAI (Table 1).
Additionally, Pearson product–moment correlation coefficients were also obtained for each pair of the dependent measures.
3.1. Analyses of variance
3.1.1. Education
Significant main effects were found for both gender (F=8.35, P