هزینه های تناسب اندام و مزایای صفات اختلال شخصیت

Abstract Extreme personality traits in humans often have detrimental life consequences, so they have long been supposed to be diseases. However, many other species display personality variants that are maintained due to their fitness advantages; in this case, they are construed as strategies. To examine the fitness costs and benefits of pathological personality traits in humans, we measured features of the A (socially odd, distrustful), B (incentive-seeking, selfish) and C (fearful, inhibited) clusters with the Personality Diagnostic Questionnaire-4 + (PDQ-4 +) in a sample of 738 outpatients. Fitness relevant parameters like mating success, reproductive output, self preservation, and access to status were assessed with the Life Outcome Questionnaire. No fitness advantages were found for high-A subjects. In contrast, high-B subjects tripled low-B subjects with regard to mating success and had 39% more offspring. Further, high-C subjects outperformed low-C subjects in attaining status and avoiding risks. These findings help explain the commonness of some extreme personality traits in humans, and suggest that they should be seen as evolutionary strategies rather than as diseases.

Introduction A diagnosis of personality disorder (PD) is made when a personality trait – whether anxiousness, impulsivity, mistrust or dominance – is so pronounced and pervasive that leads to dysfunctional outcomes, that is, to enduring distress or role impairment. PDs can cause notable clinical maladaptation, and their psychosocial consequences can be as serious as those of other severe mental disorders (Skodol, Johnson, Cohen, Sneed, & Crawford, 2007, and references therein). In fact, PDs have been found to predict poor quality of life ahead of other mental disorders, sociodemographic variables and somatic health ( Cramer, Torgersen, & Kringlen, 2006). As a result, they have long been construed as diseases ( Kendell, 2002). There is an increasing awareness, however, that the commonness of these and other mental disorders is an unsolved evolutionary paradox (Keller & Miller, 2006). Indeed, despite the fact that PDs are harmful and show heritabilities of up to 45% (Jang, Livesley, & Vernon, 1996), they have not been eroded by natural selection but remain in the population at prevalences as high as 9% (Lenzenweger, Lane, Loranger, & Kessler, 2007). It has been suggested that mutation rate may be quicker than purifying selection in this case, or that some mental disorders are simply undesirable but irrelevant to fitness (Keller & Miller, 2006). Actually, the paradox lies in the meaning we give to ‘harmful’. Whereas clinical maladaptation refers to the failure to attain socially valued goals and well-being, evolutionary misfit is the relative inability of the individual to pass his or her genes onto the next generation. The two concepts do not necessarily coincide, and so some clinical conditions may enhance fitness after all (Nesse, 2001). Phobias, for example, despite their associated distress and malfunctioning, have long been understood as ancestral adaptations against evolutionarily recurrent dangers like predators, heights, deep waters or conspecifics (Marks & Nesse, 1994). Phobias may be performing their evolved function still today: While 94,000 people are killed annually by snake bites and 449,000 by drowning (Kasturiratne et al., 2008 and Peden and McGee, 2003), no one dies due to either snake or water phobias. This may also be the case of psychopathy. Exploiting or harming others is often not detrimental for the subject and can constitute an effective – though risky – way of enhancing his or her own fitness (Buss & Duntley, 2008). In essence, natural selection does not shape organisms for happiness or goodness but for fitness (Buss, 2000), so selective forces may successfully maintain a number of clinical conditions in the population (Keller and Miller, 2006, Nesse, 2001 and Troisi, 2005). If this is the case, these traits should be seen as strategies rather than as diseases or defects. Current evidence in other species, although admittedly limited, partially supports this possibility. Personality variation in nonhumans has shown to bear upon central components of fitness such as survival, mating and reproduction (Réale and Dingemanse, 2011 and Smith and Blumstein, 2008). Furthermore, this variation seems to be maintained by tradeoffs, so that different temperaments either favor distinct fitness components or provide advantage in distinct periods or environments (Carter et al., 2010, Dingemanse et al., 2004, Réale and Festa-Bianchet, 2003 and Réale et al., 2010). This would even hold for traits that appear to be maladaptive in certain contexts (Sih, Bell, & Johnson, 2004). In the case of humans, personality has been also shown to exert a nontrivial influence on fitness (Eaves et al., 1990 and Jokela et al., 2009a) and this relationship may involve the same kind of tradeoffs as in nonhumans. For example, whereas extraversion is associated with indicators of premature death such as hospitalization due to accident, it also leads to more mates and extra-pair copulations, an effective strategy of gene dissemination (Nettle, 2005). For its part, conscientiousness enhances survival (Roberts, Kuncel, Shiner, Caspi, & Goldberg, 2007), but may produce losses of immediate opportunities as in the case of mating (Nettle, 2006). Contrary to the general feeling, even PDs may be not unequivocally detrimental. Some studies have found PDs to be roughly inconsequential unless they are accompanied by concomitant mood, anxiety, impulse-control or substance abuse disorders (Lenzenweger et al., 2007) and certain PDs have proved advantageous in the social or achievement domains (Furnham et al., 2012 and Ullrich et al., 2007). Therefore, it is not implausible that personality variation is maintained in humans by the same mechanisms as in other species. This study aims to examine whether a global measure of PD, as well as three more specific subtypes of pathological personality (A, B and C traits), is associated with a range of fitness-related life outcomes, and, if so, in what way. To this end, we studied 738 subjects who spanned the complete range of personality variation, from “healthy” to “severe PD”. We measured indicators of three fitness components: Survival, mating, and fecundity. In addition, we measured subjects' performance in a number of socially relevant domains expected to affect fitness such as attaining status, acquiring material resources, reaching autonomy, or maintaining non-kin alliances. With this information, we hope to provide new insights into the intriguing question of how personality disorders may be maintained in the population despite being clinical maladaptations.

Results 3.1. Access to status and resources: Education, job and finances Standardized regression coefficients are shown in Table 1. As a whole, PD features were detrimental to status and resources attainment: They curtailed the academic period, restricted educational and job level, prompted job instability, and limited income. However, outcomes varied considerably between clusters. High scorers in cluster A (henceforth high-A subjects) left their studies at age 21, three years before low-A subjects. They tended to attain bachelor level instead of master level, and this attainment took them longer. High-B subjects also showed decreased educational level and performance, though to a lesser extent. As regards employment, both high-A and -B subjects started working at a younger age (before age 18) and changed job more often (4.3 vs 2.7 times). However, unlike high-A subjects who showed both bad changes (dismissal or abandonment without further plans) and good ones (moving to a better job), high-B subjects only reported more bad changes (2.9 vs 1.3). Finally, high-A features limited access to higher job level and salary (600 euros less), whereas high-B subjects maintained jobs for shorter periods, had more difficulty covering basic needs, and needed family or bank loans more often. In contrast, C features were associated with easier access to status and resources. High-C subjects had better academic performance and reached higher educational levels than low-C subjects, although they also showed higher rates of drop out from studies due to psychological problems. High-C subjects started working on average two years later than low-C subjects, due to the longer period spent studying. They were also more reluctant to change jobs, including changes for the better; nonetheless, they attained a higher job level. 3.2. Self-preservation and health Most consequences of PDs in this area were attributable to B traits. They showed the highest association with self-damage in the form of suicidal thoughts and self-lesions. Suicidal acts were four times more frequent in high- than in low-B subjects (.97 vs .23 attempts). B features were also related to illegal drug use, and this was true for virtually every substance (results not shown). Cluster B traits doubled the number of months off work for either medical or psychiatric reasons, as well as the probability of a permanent inability allowance (12% vs 5%). In contrast, C traits presented a slight association with suicidal ideation but like A features were unrelated to the presence and number of attempts, and were protective against drug consumption. 3.3. Mating, reproduction and formation of alliances In the mating domain, the effects of PD total score only obscured outstanding differences among clusters. High-B subjects had 3.5 times as many mates as low-B subjects (7.3 vs 2.1) [ Fig. 1A], including five times more short-term mates and twice as many long-term mates. Unlike the other clusters, they clearly favored briefer relationships. In contrast, A and C features were related to a significant decrease in the number of mates, which was not compensated by more durable relationships. Means for (A) mating success and (B) reproductive success plotted against the ... Fig. 1. Means for (A) mating success and (B) reproductive success plotted against the frequency distribution of A, B and C traits. Note. Personality traits and number of mates/offspring are z-standardized residuals after controlling for age and sex. Error bars represent Wald 95% confidence intervals. Figure options While PD total score was unrelated with reproductive output, cluster B traits increased the probability of having offspring (35% vs 25%) and raised offspring number by 39% (.61 vs .44) [ Fig. 1B]. A and B features showed negative but non-significant trends for both variables. No association was found between personality and age at first reproduction, age at second reproduction or interbirth interval. The latter two results were unsurprising given that only 131 subjects had a second child. To further understand the above results, we examined mating and reproduction separately for each sex (not reported in Table 1). The mating advantages reported for high-B subjects hold in both males (7.6 vs 2.2, b = .440, p < .001) and females (6.5 vs 2.1, b = .378, p < .001) [ Fig. 2A]. By contrast, access to mates was constrained in males with either A traits (3.4 vs 5.8, b = − .211, p = .001) or C traits (3.3 vs 6.0, b = − .242, p < .001), but this relationship was only marginally significant in high-A females (3.4 vs 4.7, b = − .110, p = .050) and non-significant in cluster C females (3.7 vs 4.5, b = − .056, p = .316). Standardized beta coefficients of (A) mating success and (B) reproductive ... Fig. 2. Standardized beta coefficients of (A) mating success and (B) reproductive success regressed onto A, B and C traits separately for each sex. Note. Error bars represent Wald 95% confidence intervals. * p < .05 ** p < .001. Figure options On the other hand, the reproductive advantage for the cluster B was significant for males (.62 vs .40, b = .239, p = .020) but not for females (.58 vs .48, b = .102, p = .387) [ Fig. 2B]. A reproductive disadvantage associated to cluster C emerged in females (.40 vs .64, b = − .191, p = .049), but not in males (.50 vs .51, b = − .012, p = .911). In contrast, high-A males showed a marginally significant trend towards decreasing offspring number (.40 vs .59, b = − .209, p = .059) that was not evident in females (.50 vs .54, b = − .052, p = .592). Incidentally, mating success predicted reproductive output in men (b = .028, p = .036) but not in women (b = − .012, p = .536). Finally, although the PD total score showed some effect on non-kin social alliances, most results exclusively concerned high-A subjects [Table 1]. They had fewer than half as many friends as low-A subjects (2.0 vs 4.6) and met their oldest friend three years later, in post-adolescence. In contrast, B and C features were unrelated to the extension and stability of the social network. In spite of this, it was high-C subjects who considered their network to be insufficient. As for interpersonal conflicts and rule breaking, high-A and -B subjects showed higher frequency of temper tantrums, violent acts, and illegal activities, though the severity of these acts warranted more arrests only for high-B subjects (.33 vs .02). C features thwarted violent acts, deterred illegal incomes, and reduced the number of arrests (.09 vs .23). On average, personality clusters explained 5.3% of life outcome variance (range .2% to 16.3%). R2 coefficients, unstandardized betas, and the full predictions for the lower and upper quartiles of personality variables are provided as supplementary material (Table S1, available on the journal's website at www.ehbonline.org).