Autonomy (vs. sociotropy) and depressive symptoms in quitting smoking: Evidence for trait-congruence and the role of gender

Addictive Behaviors 31 (2006) 1744 – 1760

Autonomy (vs. sociotropy) and depressive symptoms in quitting smoking: Evidence for trait-congruence and the role of gender J. Lee Westmaas a,*, Roberta Ferrence b, T. Cameron Wild c a

Department of Psychology, State University of New York at Stony Brook, Stony Brook, NY 11794-2500, USA Ontario Tobacco Research Unit, University of Toronto and Centre for Addiction and Mental Health, Canada c Department of Public Health Sciences and Centre for Health Promotion Studies, University of Alberta, Canada b

Abstract According to Beck’s cognitive theory of depression, autonomy (high achievement concerns) and sociotropy (high interpersonal concerns) are vulnerability factors for depression when achievement or interpersonal stressors, respectively, are experienced. This hypothesis was tested among men and women attempting to quit smoking, an achievement stressor that can provoke depressive symptoms. Smokers recruited from the community (N = 210) provided information about their quit attempt through mailed questionnaires. For the 48-h period following the quit, relationships among autonomy, sociotropy, coping, depressive symptoms and lapsing were assessed. Structural equation models supported the trait-congruence hypothesis because greater autonomy, but not sociotropy, was associated with elevated depressive symptoms among both men and women smokers. However, results were stronger for men (b = .47, p = .0001) than for women (b = .20, p = .05). After accounting for autonomy’s relationship with depressive symptoms, greater autonomy was inversely associated with lapsing among men (b = .35, p = .01), but not women. Results point to the potential usefulness of a theoretical approach to understanding relationships between depressive symptoms and smoking cessation, and indicate that autonomous personality may be an important factor in smoking cessation in men. D 2005 Elsevier Ltd. All rights reserved. Keywords: Autonomy; Coping; Depressive symptoms; Smoking; Sociotropy

* Corresponding author. Tel.: +1 631 632 7842; fax: +1 631 632 7876. E-mail address: [email protected] (J.L. Westmaas). 0306-4603/$ - see front matter D 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.addbeh.2005.12.013

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

1745

Encouraging smokers to quit is an important public health goal but is difficult for the majority of current smokers. Tobacco use claims at least 419,000 lives every year (Centers for Disease Control and Prevention, 2002) and is responsible for at least 29% of all cancer deaths in the U.S. (American Cancer Society, 2002). Most smokers regard quitting as a very difficult process, a perception consistent with the low abstinence rates of 3% to 5% among self-quitters (Hughes, Keely, & Naud, 2004). During early abstinence, smokers experience a variety of withdrawal symptoms that can include intense feelings of depression, irritability, insomnia, and other somatic and emotional states (Hughes, 1992a). Several investigations have focused on the occurrence of depressive symptoms during quitting. These studies have suggested that depression-prone smokers are more likely to experience depressive symptoms during abstinence (Ginsberg, Hall, Reus, & Munoz, 1995; Niaura et al., 1999), which makes quitting more difficult (Brandon, 1994; Kassel, Stroud, & Paronis, 2003). Studies also indicate that in North America depression or depressive symptoms are more common among women than men (Culbertson, 1997) but depressed affect does not appear to account for women’s greater difficulty in quitting (Wetter et al., 1999). If there is indeed a relationship between depression-proneness and difficulty quitting it would be valuable to have a theoretical framework that could account for these findings. The present study uses trait constructs from Beck’s cognitive theory of depression vulnerability (Beck, 1983, 1987) to examine the occurrence of depressive symptoms during quitting, and does so for women and men separately. The present study is also relevant to research on personality in smoking behavior. Previous research has implicated personality variables such as hostility, sensation-seeking, impulsivity, and neuroticism in smoking cessation outcomes (Gilbert, Crauthers, Mooney, McClernon, & Jensen, 1999; Lipkus, Barefoot, Williams, & Siegler, 1994; Reuter & Netter, 2001), but findings have been inconsistent and not grounded in psychological theory. The present study uses a theoretical framework to investigate the relationship between personality vulnerability for depressive symptoms and ability to abstain soon after a quit attempt.

1. Cognitive vulnerability and the trait-congruence hypothesis According to Beck’s (1983, 1987) cognitive model of depression, two personality styles predispose individuals to depression, depending on the nature of stressors experienced. Referred to as the btraitcongruenceQ hypothesis, it posits that (i) individuals high in sociotropy have high needs for affiliation, acceptance, and support, and are vulnerable to interpersonal stressors, and (ii) autonomous individuals experience excessive concern and worry about personal achievement, maintaining independence and freedom, and are vulnerable to achievement-related stressors. Studies support the validity of a sociotropic personality style by demonstrating that sociotropic individuals experience dysphoric or physiological stress reactions in response to primarily interpersonal stressors (Allen, Horne, & Trinder, 1996; Bieling & Alden, 1998; Bruch, Rivet, Heimberg, Hunt, & McIntosh, 1999; Ewart, Jorgensen, & Kolodner, 1995). The evidence for an autonomous style, however, is mixed (Coyne & Whiffen, 1995; Segal, Shaw, & Vella, 1989). However, a recent investigation that used both a clinical sample of adult depressed patients, as well as matched controls from the community, found support for the congruence hypothesis for both sociotropy and autonomy (Mazure, Gruce, Maciejewski, & Jacobs, 2000). These findings suggested that previous use of undergraduate samples may have contributed to conflicting

1746

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

findings. Newer measures of sociotropy and autonomy, the Personal Style Inventory, have also been developed to reduce overlap between constructs and to exclude items assessing affective states (Robins et al., 1994). Confirmatory factor analyses have demonstrated construct validity for these newer measures of sociotropy and autonomy (Bagby, Parker, Joffe, Schuller, & Gilchrist, 1998). However, research on the relationships of these measures with depressive symptoms in non-clinical, non-student samples is lacking. Because quitting smoking is a context that can precipitate depressive symptoms, and because quitting represents a significant achievement stressor, the present study hypothesized that quitting should be associated with depressive symptoms among autonomous but not sociotropic smokers. Cross-sectional relationships between sociotropy/autonomy and depressive symptoms were thus examined, and separately for men and women.

2. Gender, depressive symptoms and quitting smoking Vulnerability to depression, as indicated by a history of prior episodes, is a risk factor for failure to achieve abstinence from smoking (Anda et al., 1990; Blazer, Kessler, McGonagle, & Swartz, 1994; Glassman, 1993; Wilhelm, Arnold, Niven, & Richmond, 2004, see Hitsman, Borelli, McChargue, Spring, & Niaura, 2003 for exceptions). Women experience elevated rates of depression (Culbertson, 1997), and in several studies, women appear to have greater difficulty quitting smoking (cf. Ward, Klesges, Zbikowski, Bliss, & Garvey, 1997; Wetter et al., 1999). Women’s greater propensity to experience depressive symptomatology has been advanced as a reason for why women in some studies have lower cessation success (Borelli, Bock, King, Pinto, & Marcus, 1996). However, Wetter et al. (1999) did not find evidence that negative affect or stress variables were responsible for women’s greater difficulty in quitting. More research is clearly needed to elucidate the role of depressive symptoms in the ability to quit smoking specifically for men and women. Examining evidence for the trait-congruence hypothesis separately for men and women in the context of quitting smoking may lead to greater understanding of relationships among gender, depressive symptoms, and ability to quit.

3. Trait-congruence, depressive symptoms and quitting smoking The present study examined evidence for the trait-congruence hypothesis by hypothesizing that because of the nature of quitting smoking, autonomous personality should be more strongly linked with depressive symptoms compared to sociotropic personality. In attempting to quit smoking, not acquiescing to urges to smoke requires a high level of self-control and motivation, as well as planning out how one will respond to the usual triggers of smoking. As such, abstaining from smoking in the initial days of a quit attempt should represent a significant achievement stressor for smokers, and lapsing a significant achievement failure for smokers motivated to quit. According to the trait-congruence hypothesis, highly autonomous smokers who attempt to quit on their own, but who lapse early in the quit attempt, should be particularly prone to depressive symptoms because lapsing represents a failure to achieve their goal. Even if highly autonomous smokers do not lapse, however, the difficulty of controlling urges or withdrawal symptoms, which should challenge highly autonomous individuals’ need for control and independence, should also trigger depressive symptoms, much more than among sociotropic individuals. This would undermine quitting success because experiencing depressive

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

1747

symptoms, or negative affect in general, has been consistently linked with lapsing during a quit attempt (Kassel et al., 2003). The hypothesized association between greater autonomy and depressive symptoms during a quit attempt should also be stronger for men than for women because of gender-related socialization experiences. Men are encouraged to develop autonomous qualities such as independence and an achievement orientation, whereas women are socialized to emphasize interpersonal relationships and intimacy (Deaux & Lewis, 1984; Eagly & Steffen, 1984). These gender stereotyped qualities occupy opposite poles of the agency-communion dimension (Bakan, 1966; Eagly & Wood, 1991; Lueptow, Garovich, & Lueptow, 1995), but both men and women may possess qualities from each end of the continuum (Kirsh & Kuiper, 2002). Thus, associations between autonomy and depressive symptoms should occur among both men and women, but because of gender expectations and socialization experiences emphasizing autonomous qualities for males, the relationship between autonomy and depressive symptoms should be stronger for men. While attempting to quit is hypothesized to increase highly autonomous smokers’ depressive symptoms, and therefore the likelihood of lapsing, the nature of a highly autonomous personality, with its emphasis on achievement, will-power and self-control, is an orientation that could increase the likelihood of abstinence in the early stages of quitting. Thus, after controlling for the effects of autonomy on depressive symptoms, greater autonomy may be associated with an increased likelihood of abstinence in the early stages of a self-initiated quit attempt for both men and women.

4. Personality vulnerability and coping with quitting One factor that may determine the extent to which autonomy is associated with depressive symptoms during quitting is how autonomous individuals cope. Little is known about how highly autonomous (or sociotropic individuals) cope with significant stressors. However, coping is an important factor in the ability to quit smoking (Shiffman, Hickcox et al., 1996; Shiffman, Paty, Gnys, Kassel, & Hickcox, 1996). Therefore, how autonomous or sociotropic individuals cope with quitting could influence the extent to which depressive symptoms are experienced, or how successful they are in abstaining. Recently, there has been a revival of interest in links between personality and coping (Suls & David, 1996), with a majority of these investigations focusing on the bBig 5Q personality dimensions (e.g., O’Brien & DeLongis, 1996). However, coping tendencies associated with personality vulnerability for depression have not been investigated. In the context of quitting smoking, Curry and Marlatt (1985) identified two categories of cognitive coping reported by smokers that appear to conceptually map on to the constructs of autonomy and sociotropy, and which were associated with resisting smoking during a quit attempt. They labeled one category of coping cognitions as achievement thoughts. This category consisted of thoughts reflecting a desire to prove to oneself that one could quit smoking. These achievement coping thoughts should be especially likely to be experienced by highly autonomous individuals attempting to quit, as autonomy is characterized by strivings for self-control, independence, and achieving goals. Another category, which Curry and Marlatt (1985) labeled bsocial concern copingQ, referred to bthoughts about letting other people downQ, bnot wanting others to be upset with you for smokingQ, and bnot wanting to harm others by smoking (e.g., dI thought of all the people I would be letting down if I were to start smoking again.T)Q (Curry and Marlatt, 1985, p. 249). These thoughts reflect a strong desire

1748

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

for positive interpersonal relations, which has been identified as one of several important motivational drives governing behavior (Novacek & Lazarus, 1990). These thoughts are sociotropic in nature because sociotropic individuals, by definition, are highly concerned about interpersonal interactions, and desire approval and acceptance. It was thus hypothesized that sociotropic smokers would be especially likely to experience these coping cognitions in order to help them cope with cravings or the urge to smoke during a quit attempt. Thus, although sociotropy should be unrelated to depressive symptoms during quitting (according to the trait-congruence hypothesis), sociotropy may be indirectly associated with short-term quitting if sociotropic coping cognitions help sociotropic smokers resist smoking. The present study used structural equation modeling to examine cross-sectional relationships among autonomy, sociotropy, coping, depressive symptoms, and ability to abstain in the 2-day period following a quit attempt. This period was chosen because as much as 50% of smokers relapse within 2 days of their quit attempt (e.g., Gulliver, Hughes, Solomon, & Dey, 1995). Analyses were conducted separately for men and women because there is evidence of gender differences in depression vulnerability and cessation success (Borelli et al., 1996) and in the prevalence of sociotropy and autonomy (Kwon, Campbell, & Williams, 2001; Sato & McCann, 1998; Scheibe, Preuschhof, Cristi, & Bagby, 2003). In addition, because of male socialization experiences that emphasize independence and self-control, associations between autonomy and depressive symptoms were expected to be stronger among men than among women. Based on the above literature, it was hypothesized that during a quit attempt, 1. For both men and women greater autonomy, but not sociotropy, will be associated with elevated depressive symptoms. 2. Sociotropy will be positively associated with social concern coping, which in turn will be associated with resisting smoking. 3. Greater autonomy will be associated with resisting smoking. In addition, autonomy will be positively associated with achievement coping, which in turn will be associated with resisting smoking.

5. Methods 5.1. Overview of study Smokers who responded to newspaper advertisements were sent a battery of questionnaires that included assessments of psychosocial variables, and measures that were to be completed at the end of the first and second day of their quit attempt. There was no-face-to-face contact with participants so the likelihood of socially desirable responding was expected to be low. Although smokers were reminded by telephone to complete measures at the instructed times, because we could not be absolutely sure if these instructions were followed, associations observed among variables should be considered cross-sectional relationships. However, it was possible to address whether the hypothesized association between autonomy and depressive symptoms was due to autonomous smokers feeling depressed because they lapsed. This was accomplished by examining whether autonomy was associated with depressive symptoms among participants who did not lapse. If among non-lapsers the relationship between autonomy and depressive symptoms remained the same, this would suggest that lapsing was not a cause of autonomous smokers’ depressive symptoms.

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

1749

5.2. Participants Two hundred and ten smokers were recruited through advertisements placed in newspapers serving the greater Toronto metropolitan area (approx. 5 million residents). Twelve advertisements were placed in weekend and weekday issues, over a period of 1 month in January 1998. Three hundred calls received were from eligible smokers and these individuals were mailed questionnaires. We received completed questionnaires from 210 smokers and compensated them $15.00 (Canadian) for their participation (approx. US$10). Of the 210 participants, 93 (44%) were men, and 117 (56%) were women. Eighty-six percent of the sample was Caucasian, 2.9% Asian/Pacific Islander, 2.4% Mixed, 1.4% African, and 1.0% Native American Indian. The remainder (6.7%) did not provide this information. Participants ranged in age from 17 to 71, with a mean age of 41 (S.D. = 12) years. Fifty-five percent of participants had at least some post-secondary education. For 39% of participants, household income was less than $20,000 per annum, for 19% it was between $20,000 and $30,000, and for 15% it was between $30,000 and $40,000. The remainder (27%) were in households earning greater than $40,000 annually (all amounts in Canadian dollars). About a third of participants (34%) were either married or currently living with a partner, 25% were divorced or separated, and 37% reported never having been married. Participants smoked an average of 18 cigarettes per day (S.D. = 9). The average age when participants began smoking was 17 years. The majority (76%) had never sought treatment for smoking. The average number of previous life-time quit attempts was five (S.D. = 7). 5.3. Procedure Newspaper advertisements recruited smokers who were serious about quitting on their own. The display ads indicated that participants would set a quit date and report on their quit attempt. Individuals who called were informed of the following eligibility requirements: (a) planning to quit in the next 20 days, (b) currently smoke at least 10 cigarettes per day, (c) not currently receiving treatment for smoking cessation, and (d) never previously diagnosed with emphysema, lung cancer, myocardial infarction, or stroke (because we wanted to target smokers whose primary motivation for quitting was not influenced by having experienced these illnesses, and to reduce attrition due to health complications). Eligible participants were asked to provide their name, phone number, and address so that a questionnaire packet could be mailed to them. Smokers were informed that the study did not provide treatment for smoking cessation. On receiving the questionnaire packet, participants were asked to complete a baseline survey that included measures assessing socio-demographic information, personality, smoking dependence, and other variables. In the letter accompanying the packet, participants were asked to complete the background questionnaire and to select a quit date that was within 20 days of receiving the packet. They were also asked to complete specific questionnaires included in the packet at the end of the first and second days of their quit attempt and to subsequently return all questionnaires in the pre-stamped envelope included. A week after mailing of the questionnaire packets, participants were contacted and asked for their quit dates. They were also reminded to complete the post-quit-date measures during a reminder phone call placed within 5 days of the quit date. The first 2 days following the quit attempt were assessed because the majority of nicotine withdrawal symptoms are at their peak during this period (Hughes, 1992; Hughes et al., 1992) and many smokers

1750

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

lapse (cf. Gulliver et al., 1995). In the post-quit-date questionnaires, participants were asked to describe the situation in which they were most tempted to smoke (bhigh-risk situationQ), various details about the high-risk situation, negative affect at the time of the high-risk situation, how much they had smoked that day, and how they coped with quitting. Four months later, participants were sent a letter that asked them how many cigarettes per day they currently smoked. 5.4. Sociodemographic and background variables Single items were used to assess sociodemographic variables (age, sex, marital status, income, education and number of children). Smoking dependence was assessed using the 6-item Fagerstrom Test for Nicotine Dependence (Heatherton, Kozlowski, Frecker, & Fagerstrom, 1991). Motivation to quit was assessed with two items that asked participants how motivated they were to quit smoking and how strongly they desired to quit smoking. Responses were on Likert scales that ranged from 1 [bnot at all (motivated)Q] to 5 [bvery much (motivated)Q] and were averaged to create an index of motivation to quit. On the survey that participants were instructed to complete on the second day of their quit attempt, a question asked whether they used a nicotine patch or chewed nicotine gum during the 48-h post-quit period. At the time of the study, other nicotine products were not yet available, nor were any NRTs available over the counter. 5.4.1. Sociotropy/autonomy The revised Personal Style Inventory (PSI; Robins et al., 1994) was used to assess sociotropy and autonomy. Various authors from different theoretical perspectives have described two similar personality styles that confer vulnerability to depression but have used different terms to describe them. These range from bdependencyQ and bself-criticismQ (Blatt, D’Afflitti, & Quinlan, 1976) to banxious attachmentQ and bcompulsive self-relianceQ (Bowlby, 1977). Although several scales have been developed to assess these dimensions, including the Sociotropy and Autonomy scales of Beck and colleagues (SAS: Beck, Epstein, Harrison, & Emery, 1983), these scales have suffered from various psychometric problems (Robins et al., 1994). The Sociotropy and Autonomy scales of the PSI (24 items each) were developed to address these shortcomings and have demonstrated good psychometric properties. The measure of sociotropy included items assessing concern about what others think (e.g., bI am very sensitive to criticism by othersQ), dependency (e.g., bI find it difficult if I have to be alone all dayQ), and pleasing others (e.g., bIt is hard for me to say dnoT to other people’s requestsQ). The measure of autonomy included items assessing perfectionism/self-criticism (e.g., bI feel badly about myself when I am not actively accomplishing thingsQ), need for control (e.g., bI feel controlled when others have a say in my plansQ), and defensive separation (e.g., bI don’t like relying on others for helpQ). Respondents indicated their agreement with each item on a scale of 1 (strongly disagree) to 6 (strongly agree). Good internal consistency has been reported for the scales (all a N .80; Robins et al., 1994). Both sociotropy and autonomy have good test–retest stability but the sociotropy construct has demonstrated greater evidence of convergent and discriminant validity. For example, sociotropy is highly correlated with the Interpersonal Dependency subscale of the Depressive Experiences Questionnaire (DEQ, Welkowitz, Lish, & Bond, 1985) but only moderately with the Beck Depression Inventory (BDI, Robins et al., 1994). Additionally, Allen and colleagues demonstrated that higher scoring participants on a slightly longer (earlier) version of the PSI sociotropy subscale showed greater physiological arousal to personal rejection scripts than to achievement failure or neutral scripts (Allen et al., 1996). In a daily diary study

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

1751

with dysphoric individuals, more sociotropic individuals were more likely to make social comparisons in interpersonal domains (Giordano, Wood, & Michela, 2000). 5.5. Post-quit date measures 5.5.1. Social concern and achievement coping Curry and Marlatt’s (1985) descriptions and examples of social concern and achievement coping were used to develop additional items for measures of social concern and achievement coping cognitions. Items were generated by study staff and face validity was ascertained by several senior investigators. This resulted in the elimination of several items and the rewording of others, in order to achieve a final set of items for social concern and achievement coping. At the end of each of the first 2 days post-quit, participants were asked about the extent to which they experienced thoughts bto help you resist your urge to smoke.Q Participants indicated whether or not they experienced each thought by placing a check mark in a bYesQ or bNoQ column. Eight bsocial concern copingQ items assessed the occurrence of thoughts about interpersonal approval if the smoker were to resist smoking, or about interpersonal disapproval if the smoker were to lapse during the high-risk situation. Examples of items were bPeople would think badly of you if you gave in to your urgesQ, bPeople you like or love would be happy if you could resist smokingQ, and bOther people would be disappointed if you gave in to your urges.Q After a reliability analysis, one item was dropped; internal consistency of the final 7-item measure was adequate for each day post-quit (as N .77). The number of byesQ responses was counted to provide an index of social concern coping for each day post-quit. The correlation was high across days (N .60) so these indices were averaged to create a measure of social concern coping for the 48-h post-quit date period. Achievement coping was assessed by 5 items. Examples of items were bYou can achieve anything you put your mind toQ, and bIf other people can quit then you can too.Q Items assessing achievement cognitions showed good internal reliability for both days (as N .71). The same procedure as above was followed to create a measure of achievement coping for the 48-h post-quit date period. 5.5.2. Depressive symptoms Participants completed the Center for Epidemiological Studies Depression Scale (CES-D; Radloff, 1977) at the end of their second quit day and with reference to the 2-day post-quit-date period. The 20item CES-D scale includes items that assess depressed mood, feelings of guilt and worthlessness, feelings of helplessness and hopelessness, psychomotor retardation, loss of appetite, and sleep disturbance. Participants indicated how often they felt that way during the 48-h post-quit period, using a scale ranging from 1 [rarely or none of the time] to 4 [most or all of the time]. The CES-D scale has shown good internal reliability and convergent and discriminant validity (Radloff, 1977) and is an appropriate measure of depressed mood in the general population (Shaver & Brennan, 1991). Internal consistency was .90 in the present sample. 5.5.3. Smoking lapse Self-reported smoking lapse within the first 48 h of the quit attempt was examined because the majority of nicotine withdrawal symptoms are at their peak during this period (Hughes, 1992). Many smokers attempting to quit lapse during this period (cf. Gulliver et al., 1995). At the end of the first and second days of their quit attempt, participants responded to specific items that assessed whether or not

1752

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

Table 1 Intercorrelations, means, and standard deviations of study variables (N = 198)

Autonomy Sociotropy Social concern coping Achievement coping Depressive symptoms Lapsinga M S.D.

Autonomy

Sociotropy

– .42** .16* .06 .36** .06 3.77 .67

– .32** .04** .24** .03 3.69 .82

Social concern coping

– .36** .30** .12 6.31 3.69

Achievement coping

– .03 .06 8.30 2.32

Depressive symptoms

– .21** 19.32 11.19

*p b .05, **p b .01. a Point–biserial correlations.

they had smoked (lapsed) during their high-risk situation, and how much they had smoked in the last 24 h. If participants smoked a puff or more, they were coded as having smoked on that day. Prior research has demonstrated that in non-intervention studies, self-report of smoking is a valid indicator of smoking (see Velicer, Prochaska, Rossi, & Snow, 1992 for a review). A final variable indicating whether or not participants had smoked during the 48-h post-quit date period was created. Abstinence from smoking was coded as 0, and lapsing was coded as 1.

6. Results 6.1. Sample characteristics The mean Fagerstrom score for the sample was 4.7 (S.D. = 2.56). In the 2-day period following the quit, 53.7% of women and 60.2% of men had lapsed. This difference between men and women was not statistically significant according to a Chi-square test [v 2 (1, N = 210) = .64, p = .48]. More than three-quarters of participants (76%) provided data at the 4-month follow-up. Nonresponders were slightly younger (M = 37.6, S.D. = 11.46) compared to responders (M = 42, S.D. = 12.6; t = 2.24, p = .03). Non-responders did not differ from responders on nicotine dependence (Ms = 4.7, p = .99) or motivation to quit (Ms = 4.1 vs. 4.2, p = .35). With non-responders coded as having smoked, 11.8% of men (11) from the entire sample, and 16.2% of women (19) were no longer smoking at the 4-month follow-up. This difference between men and women was not statistically significant [v 2 (1, N = 158) = .40, p = .68]. Responses to the measure assessing motivation to quit indicated that participants were highly motivated, with a mean score of 4.2 out of 5 (S.D. = .77). Motivation to quit was somewhat lower among 2-day lapsers (M = 4.08, S.D. = .79) than among non-lapsers (M = 4.33, S.D. = .72; t = 2.36, p = .01). Twelve smokers (5.7%) had used either the nicotine patch or gum and were excluded from analyses reported below. Table 1 reports correlations among variables used in structural equation models, and their means and standard deviations. Means for sociotropy and autonomy were comparable to those obtained in other samples (Crosby Ouimette, Klein, Anderson, Riso, & Lizardi, 1994; Robins et al., 1994).

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

1753

6.2. Data analyses Hypotheses were examined using the AMOS 5.0 structural equation modeling (SEM) procedure (Arbuckle & Wothke, 1999). With AMOS, hypothesized relationships among variables are represented by a path diagram. The goodness-of-fit of an entire model is evaluated against the observed data. Path coefficients are interpreted as standardized regression coefficients that indicate the strength of the relationship between two variables, controlling for all other associations depicted in the model. The comparative fit index (CFI; Bentler, 1990) and the root mean square error of approximation (RMSEA; Browne & Cudeck, 1993; Steiger, 1990) are commonly used to determine goodness-of-fit. RMSEA values of .06 indicate a good fit (Hu & Bentler, 1999), with decreasing values indicating better fit. Hypotheses dictated the paths for the model tested. Paths were added from each coping variable to depressive symptoms because of the possibility that coping could mitigate the occurrence of depressive symptoms during quitting. In addition, a path from depressive symptoms to lapsing was included because of prior research indicating a reliable relationship between negative affect and lapsing during quit attempts (Kassel et al., 2003). The approximately even split between lapsing and abstinence for women and men indicated that this variable was appropriate for use in structural equation models for the 2-day post-quit period (Tabachnick & Fidell, 1996). Shared variance between autonomy and sociotropy was modeled (double-headed arrow) so that any relationships observed between autonomy and depressive symptoms, lapsing or coping controlled for sociotropy (and similarly, any relationships observed between sociotropy and depressive symptoms, lapsing, or coping controlled for autonomy). This model proved to be a good fit to the data, based on fit indices [RMSEA= 0, CFI = 1.0; v 2 (4, N = 198) = 1.55, p = .82]. Figs. 1 and 2 illustrate path coefficients for men and women, respectively. .45***

SOCIOTROPY

AUTONOMY

-.09

ACHIEVEMENT COPING

-.35** -.04

.03

.24*

.05

.47***

DEPRESSIVE SYMPTOMS

-.02

.50***

.14

SOCIAL CONCERN COPING

-.20*

SMOKING LAPSE

Fig. 1. Structural model indicating associations among autonomy, sociotropy, depressive symptoms, coping, and lapsing among men.

1754

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760 .42***

SOCIOTROPY

AUTONOMY

.16

ACHIEVEMENT COPING

-.02

.20*

-.12

DEPRESSIVE SYMPTOMS

.35***

.11

.12

.39***

SOCIAL CONCERN COPING

.20* -.22*

-.07

SMOKING LAPSE

Fig. 2. Structural model indicating associations among autonomy, sociotropy, depressive symptoms, coping, and lapsing among women.

6.3. Short-term cessation (2 days post-quit date) 6.3.1. Men Among men, greater autonomy was associated with elevated depressive symptoms (b = .47, p = .0001). As hypothesized, sociotropy was not associated with depressive symptoms (b = .05, p = .60). These coefficients differed significantly from each other (t = 2.65, p = .01). Greater depressive symptomatology was associated with an increased likelihood of lapsing among men (b = .50, p = .0001). Among men, greater autonomy was also associated with a reduced likelihood of lapsing (b = .35, p = .003), controlling for all other associations depicted in the model. Achievement coping was not associated with autonomy, nor with depressive symptoms or the likelihood of lapsing. Among men, greater sociotropy was associated with greater social concern coping (b = .24, p = .02). Social concern coping was not associated with depressive symptoms (b = .14, p = .15), but greater social concern coping was associated with a reduced likelihood of lapsing (b = .20, p = .05). 6.3.2. Women Results of the model for women indicated that similar to men, greater autonomy was associated with greater depressive symptoms (b = .20, p = .04). In addition, sociotropy was not associated with depressive symptoms (b = .02, p = .81). This difference in coefficients, however, was not statistically significant (t = 1.43, p = .15). As was the case for men, autonomy was not associated with achievement coping, and achievement coping was not associated with depressive symptoms. Unlike the case for men, autonomy was not associated with lapsing (b = .16, p = .12). Similar to results for men, sociotropy was associated with social concern coping (b = .35, p = .0001), but among women, greater social concern coping was associated with elevated depressive symptoms

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

1755

(b = .39, p = .0001). Simultaneously, however, greater social concern coping was associated with a reduced likelihood of lapsing (b = .22, p = .05). The relationship between greater autonomy and heightened depressive symptomology did not differ significantly between men and women (t = 1.76, p = .08). However, the relationship between greater autonomy and a reduced likelihood of lapsing was significantly stronger for men that it was for women (t = 3.28, p = .001). There was no difference between men and women in the relationship between sociotropy and social concern coping (t = .84, p = .40), but the relationship between social concern coping and depressive symptoms was marginally greater for women compared to men (t = 1.87, p = .06). The relationship between greater social concern coping and a reduced likelihood of lapsing did not differ between women and men (t = .17, p = .87). 6.4. Autonomy and depressive symptoms among non-lapsers To evaluate whether the association between greater autonomy and elevated depressive symptoms could be attributed to autonomous men feeling depressed because they lapsed, a similar model was evaluated using only men smokers who did not lapse during the 48-h post-quit date period [RMSEA= 0, CFI = 1.0; v 2 (4, N = 34) = 2.24, p = .69]. This model indicated that greater autonomy was still significantly associated with depressive symptoms among men who had remained abstinent (b = .64, p = .001). This relationship was also significantly greater than the null relationship between sociotropy and depressive symptoms among abstaining men (b = .02, p = .90; t = 2.04, p = .05). As in the first model, achievement-oriented cognitive coping was not associated with autonomy or depressive symptoms. To evaluate whether the association between greater autonomy and greater depressive symptoms could be a result of autonomous women feeling depressed because they lapsed, a similar model was evaluated using only women smokers who did not lapse during the 48-h post-quit date period [RMSEA= 0, CFI = 1.0; v 2 (4, N = 49) = 2.24, p = .69]. This good-fitting model indicated that neither autonomy (b = .21, p = .16) nor sociotropy (b = .14, p = .39) was associated with depressive symptoms among women who had remained abstinent. As in the first model, achievement-oriented cognitive coping among abstaining women was not associated with autonomy or depressive symptoms. However, among abstaining women, greater sociotropy was again associated with greater social concern coping (b = .41, p = .001), and greater social concern coping was associated with greater depressive symptoms (b = .37, p = .02). 6.5. Four-month cessation Logistic regression models examined the relationships of sociotropy and autonomy with smoking status at the 4-month follow-up among participants who provided follow-up data. Sociotropy and autonomy were independent variables and smoking status the dependent variable. For men, neither sociotropy nor autonomy predicted 4-month smoking status [v 2 (2, N = 61) = .56, p = .78]. Among women, however, the logistic regression model was significant [v 2 (2, N = 88) = 7.83, p = .02]. Univariate results indicated that the odds of being a smoker at the 4-month follow-up was higher among more sociotropic women (OR = 2.14, p = .05).

1756

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

7. Discussion Although depressive symptoms have received a significant amount of attention in the last 15 years as a potential impediment to smoker’s ability to quit, this literature has not benefited from a firm psychological–conceptual or theoretical model that could be used to guide research. The present study uses a theoretically based approach to examine depressive symptoms in smokers by using trait constructs from Beck’s cognitive theory (Beck, 1983, 1987). The importance of gender in this approach was also recognized. Rather than using a clinical or undergraduate sample, which limits the generalizeability of findings, the present study recruited community residents from a large metropolitan area involved in an ongoing attempt to quit smoking on their own. Results obtained add to existing research on the traitcongruence hypothesis, which thus far has produced mixed findings for autonomy (Robins, 1995). Results of the study supported the trait-congruence hypothesis that an achievement stressor such as attempting to quit smoking should produce depressive symptoms among highly autonomous rather than highly sociotropic smokers. Specifically, results indicated that for both men and women greater autonomy, but not sociotropy, was associated with elevated depressive symptomatology in the first 2 days of an unaided quit attempt. However, the hypothesis was supported more strongly among men because for men, the relationship between greater autonomy and heightened depressive symptoms was stronger than the relationship between sociotropy and depressive symptoms. In contrast, among women, the autonomy–symptoms relationship did not differ significantly from the sociotropy– symptoms relationship. Among men, the positive relationship between autonomy and depressive symptoms could not be attributed simply to autonomous men feeling depressed because they lapsed. Among men who had not lapsed 2 days post-quit date, autonomy was still significantly associated with depressive symptomatology. For highly autonomous men, it may be the challenges to their strong desires for self-control and independence from cigarettes, in the face of strong urges to smoke, that contributed to their feelings of depression. The positive association between depressive symptoms and lapsing indicate that autonomy could be a significant risk factor for failing to quit among men smokers who attempt to quit on their own. The present study also investigated achievement-oriented coping thoughts as a possible factor that would mitigate the effects of autonomous personality on greater depressive symptoms. However, this coping dimension was unrelated to both autonomy and depressive symptoms among both men and women. Other types of cognitive coping pertinent to autonomous personality, but not measured in the current study, may be important mediators of the autonomy–depressive symptoms association. Future research should examine cognitive mediators of this relationship, particularly among smokers attempting to quit. Results also indicated that, accounting for the significant relationships between autonomy and depressive symptoms, and between depressive symptoms and lapsing, greater autonomy was associated with a decreased likelihood of lapsing among men. If longitudinal analyses can demonstrate that autonomous personality in men is predictive of successfully resisting smoking during a quit attempt, it would suggest that if autonomous men can be taught how to cope with depressive symptoms during quitting, interventions that take advantage of autonomous men’s achievement orientation could potentially increase the likelihood of cessation. Although sociotropy was hypothesized to be unrelated to depressive symptoms (consistent with the trait-congruence hypothesis), results demonstrated that this cognitive vulnerability dimension may nevertheless have implications for smoking cessation, particularly among women. Women scoring

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

1757

more highly on sociotropy were more likely to use social concern coping thoughts during their attempt to quit smoking, and this was true even among non-lapsers (indicating that these coping thoughts were not a result of having lapsed). While social concern coping thoughts were positively associated with depressive symptoms among both lapsers and non-lapsers, after accounting for this relationship, social concern coping cognitions were associated with resisting smoking in the shortterm. Social concern coping items included thoughts about how displeased others would be if one were to fail to quit. These cognitions might have caused distress in women, even if they had not lapsed, because in general women are socialized to be nurturant and concerned about the impact of their behavior on others (Deaux & Lewis, 1984; Eagly & Steffen, 1984). Overall, these results suggest that while greater social concern coping may be a factor in depressive symptoms among women smokers attempting to quit, and that highly sociotropic women are particularly likely to engage in this form of coping, social concern coping thoughts may help women (and men) resist smoking in the short-term. Results also indicated that although sociotropy itself was not associated with lapsing in the short-term among men or women, highly sociotropic women were less likely to be abstinent 4 months following the quit attempt. Future research should examine the mechanisms behind this association. 7.1. Limitations of research The relationships examined in the present study were cross-sectional and therefore the causal direction of associations between variables could not be ascertained. However, additional analyses conducted determined whether relationships between cognitive personality vulnerability and depressive symptoms were due to lapsing. These analyses indicated that lapsing was not a factor in relationships observed between autonomous personality and depressive symptoms among men or women. Nevertheless, future research on the contributing roles of sociotropy and autonomy to depressive symptoms during smokers’ quit attempts should employ longitudinal designs and include repeated assessments of mood and coping. The extent to which depressive symptoms were the result of physiological withdrawal that autonomous men experienced, or a result of appraisal processes or coping specific to autonomous men, or both, is not known. In addition, the association between autonomous personality and depressive symptoms found among men might also be obtained among men smokers who are not attempting to quit. However, the fact that there is an association between autonomy and depressive symptoms when smokers are attempting to quit is important because depressive symptoms increase the risk of lapsing (Brandon, 1994; Kassel et al., 2003). The demonstration of an association between greater autonomy and a decreased likelihood of lapsing also points to the potential importance of this construct in smoking cessation among men. Biochemical validation of smoking was not conducted. Nevertheless, because the study required minimal contact with investigators, and because there was no intervention, participants were less likely to have misreported their smoking (Velicer et al., 1992). To further address the generalizability of the results, it would be important to replicate them in populations of smokers who differ in sociodemographic variables such as age and socioeconomic status. Finally, we could not be absolutely certain that measures were completed at the instructed times. However, using the dates that participants were required to indicate on measures, the mean number of days between completion of the baseline measures and the quit date was 4.2 days. This suggested that

1758

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

participants followed the instructions outlined by completing the baseline measures, setting a quit date shortly thereafter, and completing daily measures during the quit period. 7.2. Conclusion The present study used Beck’s cognitive theory of depression (Beck, 1983, 1987) to attempt to understand links between depressive symptoms and smoking cessation early in a quit attempt when smokers are vulnerable to lapsing. The data supported the conceptualization of quitting smoking as an achievement stressor that elicits depressive symptoms among autonomous rather than sociotropic smokers. However, evidence was stronger for men than for women. Results also indicated that among men, autonomous personality was associated with a reduced likelihood of lapsing. Future research addressing relationships among autonomy, depressive symptoms, and lapsing has the potential to help men and women smokers high in autonomy resist smoking during a quit attempt.

Acknowledgements The authors wish to thank Daniel Klein, Marvin Goldfriend, and anonymous reviewers for their feedback on an earlier draft of this manuscript. This research was supported by a grant from the Centre for Addiction and Mental Health, Toronto, Ontario, Canada. The authors wish to thank the research assistants who helped conduct this study: Linda Couper, Meytal Elhav, Thusi Iswaran, Ailin Oshi, and Jessica Sinclair.

References Allen, N. B., Horne, D. J. de L., & Trinder, J. (1996). Sociotropy, autonomy, and dysphoric emotional responses to specific classes of stress: A psychophysiological evaluation. Journal of Abnormal Psychology, 105, 25 – 33. American Cancer Society. (2002). Cancer facts and figures. Atlanta, GA7 American Cancer Society. Anda, R. F., Williamson, D. F., Escobedo, L. G., Mast, E. E., Giovino, G. A., & Remington, P. L. (1990). Depression and the dynamics of smoking: A national perspective. Journal of the American Medical Association, 264, 1541 – 1545. Arbuckle, J. L., & Wothke, W. (1999). AMOS 4.0 user’s guide. Chicago7 Smallwaters. Bagby, R. M., Parker, J. D. A., Joffe, R. T., Schuller, D., & Gilchrist, E. (1998). Confirmatory factor analysis of the Revised Personal Style Inventory (PSI). Assessment, 5, 31 – 43. Bakan, D. (1966). The duality of human existence: An essay on psychology and religion. Chigago7 Rand McNally. Beck, A. T. (1983). Cognitive therapy of depression: New perspectives. In P. J. Clayton, & J. E. Barrett (Eds.), Treatment of depression: Old controversies and new approaches (pp. 265 – 290). New York7 Raven Press. Beck, A. T. (1987). Cognitive model of depression. Journal of Cognitive Psychotherapy, 1, 2 – 27. Beck, A. T., Epstein, N., Harrison, R. P., & Emery, G. (1983). Development of the Sociotropy – Autonomy Scale: A measure of personality factors in psychopathology, Unpublished manuscript. Philadelphia: University of Pennsylvania. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238 – 246. Bieling, P. J., & Alden, L. E. (1998). Cognitive–interpersonal patterns in dysphoria: The impact of sociotropy and autonomy. Cognitive Therapy and Research, 22, 161 – 178. Blatt, S. J., D’Affliti, J. P., & Quinlan, D. M. (1976). Experiences of depression in normal young adults. Journal of Abnormal Psychology, 85, 383 – 389. Blazer, D., Kessler, R., McGonagle, K., & Swartz, S. (1994). The prevalence and distribution of major depression in a national community sample: The national comorbidity survey. American Journal of Psychiatry, 151, 979 – 986.

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

1759

Borelli, B., Bock, B., King, T., Pinto, B., & Marcus, B. H. (1996). The impact of depression on smoking cessation in women. American Journal of Preventive Medicine, 12, 378 – 387. Bowlby, J. (1977). The making and breaking of affectional bonds: I. Aetiology and psychopathology in the light of attachment theory. British Journal of Psychiatry, 130, 201 – 210. Brandon, T. H. (1994). Negative affect as motivation to smoke. Current Directions in Psychological Science, 3, 33 – 37. Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen, & J. S. Long (Eds.), Testing structural equation models (pp. 445 – 455). Newbury Park, CA7 Sage. Bruch, M. A., Rivet, K. M., Heimberg, R. G., Hunt, A., & McIntosh, B. (1999). Shyness and sociotropy: Additive and interactive relations in predicting interpersonal concerns. Journal of Personality, 67, 373 – 406. Centers for Disease Control and Prevention. (2002). Annual smoking-attributable mortality, years of potential life lost, and economic costs — United States, 1995–1999. Morbidity and Mortality Weekly Report, 51, 300 – 303. Coyne, J. C., & Whiffen, V. E. (1995). Issues in personality as diathesis for depression: The case of sociotropy-dependency and autonomy-self-criticism. Psychological Bulletin, 118, 358 – 378. Crosby Ouimette, P., Klein, D. N., Anderson, R., Riso, L. P., & Lizardi, H. (1994). Relationship of sociotropy/autonomy and dependency/self-criticism to DSM-III-R personality disorders. Journal of Abnormal Psychology, 103, 743 – 749. Culbertson, F. (1997). Depression and gender. An international review. American Psychologist, 52, 25 – 31. Curry, S. J., & Marlatt, G. A. (1985). Unaided quitters’ strategies for coping with temptations to smoke. In S. Shiffman, & T. A. Wills (Eds.), Coping and substance use (pp. 243 – 265). Orlando7 Academic Press. Deaux, K., & Lewis, L. L. (1984). The structure of gender stereotypes: Interrelationships among components and gender label. Journal of Personality and Social Psychology, 46, 991 – 1004. Eagly, A. H., & Steffen, V. (1984). Gender stereotypes stem from the distribution of women and men into social roles. Journal of Personality and Social Psychology, 46, 735 – 754. Eagly, A. H., & Wood, W. (1991). Explaining sex differences in social behavior: A meta-analytic perspective. Personality and Social Psychology Bulletin, 17, 306 – 315. Ewart, C. K., Jorgensen, R. S., & Kolodner, K. B. (1998). Sociotropic cognition moderates blood pressure response to interpersonal stress in high-risk adolescent girls. International Journal of Psychophysiology, 28, 131 – 142. Gilbert, D. G., Crauthers, D. M., Mooney, D. K., McClernon, F. J., & Jensen, R. A. (1999). Effects of monetary contingencies on smoking relapse: Influences of trait depression, personality, and habitual nicotine intake. Experimental and Clinical Psychopharmacology, 7, 174 – 181. Ginsberg, D., Hall, S. M., Reus, V. I., & Munoz, R. F. (1995). Mood and depression diagnosis in smoking cessation. Experimental and Clinical Psychopharmacology, 3, 389 – 395. Giordano, C., Wood, J. V., & Michela, J. L. (2000). Depressive personality styles, dysphoria, and social comparisons in everyday life. Journal of Personality and Social Psychology, 79, 438 – 451. Glassman, A. H. (1993). Cigarette smoking: Implications for psychiatric illness. American Journal of Psychiatry, 150, 546 – 553. Gulliver, S. B., Hughes, J. R., Solomon, L. J., & Dey, A. N. (1995). An investigation of self-efficacy, partner support and daily stresses as predictors of relapse to smoking in self-quitters. Addiction, 90, 767 – 772. Heatherton, T. F., Kozlowski, L. T., Frecker, R. C., & Fagerstrom, K. (1991). The Fagerstrom test for nicotine dependence: A revision of the Fagerstrom Tolerance Questionnaire. British Journal of Addiction, 86, 1119 – 1128. Hitsman, B., Borelli, B., McChargue, D. E., Spring, B., & Niaura, R. (2003). History of depression and smoking cessation outcome: A meta-analysis. Journal of Consulting and Clinical Psychology, 71, 657 – 663. Hu, L.-T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1 – 55. Hughes, J. R. (1992). Tobacco withdrawal in self-quitters. Journal of Consulting and Clinical Psychology, 60, 689 – 697. Hughes, J. R., Gulliver, S. B., Fenwick, J. W., Valliere, W. A., Cruser, K., Pepper, S., et al. (1992). Smoking cessation among self-quitters. Health Psychology, 11, 331 – 334. Hughes, J. R., Keely, J., & Naud, S. (2004). Shape of the relapse curve and long-term abstinence among untreated smokers. Addiction, 99, 29 – 38. Kassel, J. D., Stroud, L. R., & Paronis, C. A. (2003). Smoking, stress, and negative affect: Correlation, causation, and context across stages of smoking. Psychological Bulletin, 129, 270 – 304. Kirsh, G. A., & Kuiper, N. A. (2002). Individualism and relatedness themes in the context of depression, gender, and a self-schema model of emotion. Canadian Psychology, 43, 76 – 90.

1760

J.L. Westmaas et al. / Addictive Behaviors 31 (2006) 1744–1760

Kwon, P., Campbell, D. G., & Williams, M. G. (2001). Sociotropy and autonomy: Preliminary evidence for construct validity using TAT narratives. Journal of Personality Assessment, 77, 128 – 138. Lipkus, I. M., Barefoot, J. C., Williams, R. B., & Siegler, I. C. (1994). Personality measures as predictors of smoking initiation and cessation in the UNC Alumni Heart Study. Health Psychology, 13, 149 – 155. Lueptow, L. B., Garovich, L., & Lueptow, M. B. (1995). The persistence of gender stereotypes in the face of changing sex roles: Evidence contrary to the sociocultural model. Ethology and Sociobiology, 16, 509 – 530. Mazure, C. M., Bruce, M. L., Maciejewski, P. K., & Jacobs, S. C. (2000). Adverse life events and cognitive – personality characteristics in the prediction of major depression and antidepressant response. American Journal of Psychiatry, 157, 896 – 903. Niaura, R., Britt, D. M., Borreli, B., Shadel, W. G., Abrams, D. B., & Goldstein, M. G. (1999). History and symptoms of depression among smokers during a self-initiated quit attempt. Nicotine and Tobacco Research, 1, 251 – 257. Novacek, J., & Lazarus, R. S. (1990). The structure of personal commitments. Journal of Personality, 58, 693 – 715. O’Brien, T. B., & DeLongis, A. (1996). The interactional context of problem-, emotion-, and relationship-focused coping: The role of the big five personality factors. Journal of Personality, 64, 775 – 813. Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385 – 401. Reuter, M., & Netter, P. (2001). The influence of personality on nicotine craving: A hierarchical multivariate statistical prediction model. Neuropsychobiology, 44, 47 – 53. Robins, C. J. (1995). Personality–event interaction models of depression (1995). European Journal of Personality, 9, 367 – 378. Robins, C. J., Ladd, J., Welkowitz, J., Blaney, P. H., Diza, R., & Kutcher, G. (1994). The personal style inventory: Preliminary validation studies of new measures of sociotropy and autonomy. Journal of Psychopathology and Behavioral Assessment, 16, 277 – 300. Sato, T., & McCann, D. (1998). Individual differences in relatedness and individuality: An exploration of two constructs. Personality and Individual Differences, 24, 847 – 859. Scheibe, S., Preuschhof, C., Cristi, C., & Bagby, R. M. (2003). Are there gender differences in major depression and its response to antidepressants. Journal of Affective Disorders, 75, 223 – 235. Segal, Z. V., Shaw, B. F., & Vella, D. D. (1989). Life stress and depression: A test of the congruency hypothesis for life event content and depressive subtype. Canadian Journal of Behavioural Science, 21, 389 – 399. Shaver, P. R., & Brennan, K. A. (1991). Measures of depression and loneliness. In J. Robinson, P. Shaver, & L. Wrightsman (Eds.), Measures of Personality and Social Psychological Attitudes (pp. 195 – 289). Academic Press. Shiffman, S., Hickcox, M., Paty, J., Gnys, M., Kassel, J., & Richards, T. (1996). Progression from a smoking lapse to relapse: Prediction from abstinence violation effects, nicotine dependence, and lapse characteristics. Journal of Consulting and Clinical Psychology, 64, 993 – 1002. Shiffman, S., Paty, J. A., Gnys, M., Kassel, J. A., & Hickcox, M. (1996). First lapses to smoking: Within-subjects analysis of real-time reports. Journal of Consulting and Clinical Psychology, 64, 366 – 379. Steiger, J. H. (1990). Structural model evaluation and modification: An interval estimation approach. Multivariate Behavioral Research, 24, 173 – 180. Suls, J., & David, J. P. (1996). Coping and personality: Third time’s the charm? Journal of Personality, 64, 993 – 1005. Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics. New York, NY7 HarperCollins College Publishers. Velicer, W. F., Prochaska, J. O., Rossi, J. S., & Snow, M. G. (1992). Assessing outcome in smoking cessation studies. Psychological Bulletin, 111, 23 – 41. Ward, K. D., Klesges, R. C., Zbikowski, S. M., Bliss, R. E., & Garvey, A. J. (1997). Gender differences in the outcome of an unaided smoking cessation attempt. Addictive Behaviors, 22, 521 – 533. Welkowitz, J., Lish, J. D., & Bond, R. N. (1985). The depressive experiences questionnaire: Revision and validation. Journal of Personality Assessment, 49, 89 – 94. Wetter, D. W., Kenford, S. L., Smith, S. S., Fiore, M. C., Jorenby, D. E., & Baker, T. B. (1999). Gender differences in smoking cessation. Journal of Consulting and Clinical Psychology, 67, 555 – 562. Wilhelm, K., Arnold, K., Niven, H., & Richmond, R. (2004). Grey lungs and blue moods: Smoking cessation in the context of lifetime depression history. Australian and New Zealand Journal of Psychiatry, 38, 896 – 905.