Genetic and environmental contributions to perfectionism and its common factors

Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at ScienceDirect

Psychiatry Research journal homepage: www.elsevier.com/locate/psychres

Genetic and environmental contributions to perfectionism and its common factors Carmen Iranzo-Tatay a,n, Natalia Gimeno-Clemente b, Maria Barberá-Fons a, M. Ángeles Rodriguez-Campayo g, Luis Rojo-Bofill a, Lorenzo Livianos-Aldana a,c,d, Luis Beato-Fernandez e, Francisco Vaz-Leal f, Luis Rojo-Moreno a,c,d a

Department of Psychiatry, “La Fe” Hospital, Valencia, Spain Psychiatry Investigation Group, Eating Disorders and Children's Psychiatry Department, “La Fe” Hospital, Valencia, Spain c CIBER Epidemiology and Public Health (CIBERESP), Spain d Department of Psychiatry, Medicine School, University of Valencia, Spain e Head of Eating Disorders and Children's Psychiatry Department, Hospital General, Ciudad Real, Spain f Department of Psychiatry, Medicine School, University of Extremadura, Spain g Psychologist, department of Psychiatry, “Hospital Univesitario de Albacete”, Spain b

art ic l e i nf o

a b s t r a c t

Article history: Received 23 September 2014 Received in revised form 9 October 2015 Accepted 15 November 2015

The aims of this study: (1) To evaluate the relative contributions of genetics and environment to perfectionism and it's two constructs: self-oriented, and socially prescribed perfectionism. (2) To clarify genetic and environmental common origins of both personal and social components. Methods: Participants were 258 pairs of adolescent Spanish twins. Socially prescribed and self-oriented perfectionism were assessed using the perfectionism subscale of the Eating Disorder Inventory. Statistics: univariate and bivariate twin models, according to sex. Results; Heritability of self-oriented perfectionism was 23% in boys and 30% in girls, and of socially prescribed perfectionism 39% in boys and 42% in girls. Bivariate analysis suggested a common genetic and environmental pathway model. The genetic correlation between both perfectionisms was 0.981 in boys and 0.704 in girls. The non-shared environmental correlation was 0.254 in boys and 0.259 in girls. Conclusions; genetic influences on perfectionism are moderate during adolescence. Our results point toward a shared genetic component underlying both kind of perfectionism. These findings generate doubts about the hypothesis of a leading role of genetics in the pathogenesis of Self-oriented perfectionism and of environment in socially prescribed. The high genetic correlation seems to indicate that self-oriented and socially prescribed are the same dimension of perfectionism. & 2015 Elsevier Ireland Ltd. All rights reserved.

Keywords: Perfectionism Adolescents Twins Genetics Environment

1. Introduction Perfectionism has been described as a personality trait characterized by overly critical self-evaluation that stands for excessively high performance standards (Burns, 1980; Patch, 1984). Hollender was the first to describe perfectionism from a psychopatholological perspective, as a demand to oneself or others of a higher quality of performance than that required by the situation (Hollender, 1965). Hewitt and Flett described three dimensions of perfectionism that differ in the source and target of perfectionistic motivations and behaviors. First of all, self-oriented perfectionism (SOP) focuses on excessively high standards, including setting n Correspondence to: Hospital la Fe de Valencia, C/Boulevard Sur sin número, Sala de Agudos de Psiquiatría, Torre D, 7ª Planta, CP/46026 Valencia, Spain. E-mail address: [email protected] (C. Iranzo-Tatay).

exacting standards for oneself, evaluating one's own behavior stringently and striving to attain perfection in one's own endeavors as well as striving to avoid failure. Secondly, socially prescribed perfectionism (SPP) assesses the belief that others have unrealistically high standards for the individual, that they stringently evaluate the individual, and that they exert pressure to be perfect. Lastly, other-oriented perfectionism characterized by holding excessively high standards for significant others (Hewitt et al., 1989; Antony et al., 1998). Shafran, Cooper and Fairburn proposed the construct of “clinical perfectionism”, defined as the over-evaluation of the striving for, and achievement of, personally demanding standards, despite adverse consequences (Shafran et al., 2002, 2003, 2006). Perfectionists allow little room for making mistakes and perceive even minor ones as likely to lead to failure. Consequently, they never feel that anything is done completely enough or well enough and their actions are always accompanied by feelings of self-criticism and a sense of

http://dx.doi.org/10.1016/j.psychres.2015.11.020 0165-1781/& 2015 Elsevier Ireland Ltd. All rights reserved.

Please cite this article as: Iranzo-Tatay, C., et al., Genetic and environmental contributions to perfectionism and its common factors. Psychiatry Research (2015), http://dx.doi.org/10.1016/j.psychres.2015.11.020i

C. Iranzo-Tatay et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

2

ineffectiveness (Burns, 1980; Ruggiero et al., 2003). Therefore they constantly have thoughts of failure, guilt, indecisiveness, shame, and low self-esteem (Frost et al., 1990; Shafran and Mansell, 2001; Ruggiero et al., 2003). Investigations of genetic and environmental influences on personality traits have suggested that genes play an important role in the expression of perfectionism. A Japanese study determined that perfectionism, assessed with the perfectionism subscale of the Eating Disorder Inventory (EDI), showed heritability estimates in young adult women of 37% (Kamakura et al., 2003). An Australian survey used a similar sample to evaluate the sources of individual differences in three dimensions of perfectionism: “concern over mistakes” (CM), “personal standards” (PS), and “doubts about actions” (DA). Their results found additive genetic and non-shared environmental influences across all measures of perfectionism, where genetic estimates ranged from 25% to 39% of the variance (Wade and Bulik, 2007). A North-American study found similar heritability estimates in a women twin adult sample, ranging from 29% to 42%. They concluded that, as CM is highly correlated with both PS and DA, it seems to be the driving component of the construct of perfectionism. Thus, CM would behave as a bridge between the other two dimensions (Tozzi et al., 2004). Finally, another North-American study evaluated heritability estimates for CM and DA subscales of the Frost Multidimensional Perfectionism Scale. They observed a pattern of genetic and unique environmental effects that influenced both CM and DA, with heritability estimates of 54% and 47% respectively. Moreover, their results revealed that genetic factors were primarily responsible for associations between anxiety and maladaptive perfectionism (Moser et al., 2012). Nonetheless, the extent to which genetic and environmental factors determine self-oriented perfectionism and socially prescribed perfectionism has not been tested yet, nor whether genetic and environmental effects share common etiological factors. Our goal was to use a population based adolescent twin sample of our region (Valencia, Spain) to clarify these two aspects, by means of the EDI perfectionism subscale. As Shafran pointed out, perfectionism is viewed as being so integral to eating disorders that one of the most widely used measures of eating disorder psychopathology (the ‘Eating Disorder Inventory') includes a subscale on perfectionism (Garner et al., 1983; Shafran and Mansell, 2001). Moreover, even do perfectionism is widely related to eating disorders, it has also been related to other mental disorders, and the EDI-perfectionism scale includes six items measuring perfectionism without relating it to eating habits. Thus it´s not a measure of perfectionism exclusively related to eating pathologies We consider that our work gives a new and original approach to perfectionism by the fact that no other research has been conducted using a bivariate model between both constructs of perfectionism.

2. Method 2.1. Study population Subjects were obtained from a study of risk factors, early detection and prevention of eating disorders, the DITCA-CV program (Programa de Detección e Intervención en Trastornos de la Conducta Alimentaria de la Comunidad Valenciana), that the Public Health and Education administrations of Valencia, Spain, offers from years ago to all compulsory secondary education schools in our region, public and private. The survey was first approved by the Department of Public Health of the regional health administration. The participating schools obtained passive informed consent from parents. The methods for the DITCA-CV have been reviewed in detail elsewhere (see Rojo-Moreno et al. (2013)). The

inclusion criteria were being 12–18 years old and attending class on the day in which the questionnaires were conducted. The global population sample comprised 57,997 schoolchildren (49.1% girls, 50.9% boys) obtained over four consecutive years. Although not a random sample, it included a very high percentage of the school population of our community. The Ethics Committee of the Public Health department approved the use of the scholarships' database that contained information regarding students' initials, date of birth, school center and class group. It is an anonymous database, so that only the students' tutors were able to identify who they were. From this population sample, adolescent twins were identified by matching on family name, date of birth, school and year of the survey. It is relatively easy to identify same sex and opposite sex twin pairs from large cohort studies, as long as name, date of birth and sex identifiers are available (Webbink et al., 2006). Five hundred eighty-four pairs of adolescent twins between 13 and 18 years of age were identified (82 monozygotic [MZ] and 87 dizygotic [DZ] pairs of male twins, 118 MZ and 102 DZ pairs of female twins, and 195 opposite-sex pairs of DZ twins). Given that not all schools participated in facilitating access to possible twin pairs, we cannot affirm that our sample is representative of twins in our region (Comunidad Valenciana). It's a convenience twin sample comprised of twin pairs to whom you were able to access through their school centers. There were no selection bias attributable to parents as children directors or tutors were those who permitted to stablish twin zygosity. For the present study only those twins were both members of the pair responded to the complete EDI-Perfectionism scale were included. Opposite sex pairs were excluded of the univariate and bivariate analysis. Assessment packages for each student were sent to the participating schools. Questionnaires included data of the schoolchildren's age, gender, eating habits and the perfectionism subscale from de Eating Disorders Inventory. 2.2. Perfectionism The EDI (Eating Disorder Inventory) is a self-report instrument (Garner et al., 1983) that has been adapted to the Spanish population (Guimerá and Torrubia, 1987). It is designed to assess a variety of psychological and behavioral characteristics that are common to anorexia nervosa and bulimia nervosa. The EDI consists of eight subscales, one of which is the Perfectionism subscale. Confirmatory factor analysis indicated that, for both genders, the Perfectionism scale is best represented by a multidimensional factor structure with three self-oriented perfectionism items (SOP), such as “My goals are too high”, and three socially prescribed perfectionism items (SPP), such us “My parents set high standards for me” (Sherry et al., 2004). In our sample, the EDIPerfectionism scale showed a Cronbach's Alphas of 0.65. Similar results were described by Shore et al. who reported a Cronbach's Alphas in 354 young women (11–18 years old) of 0.70, and in 169 males (11–18 years old) of 0.62 (Shore and Porter, 1990). Other authors have reported Cronbach's Alphas ranging between 0.73 and 0.82 (Garner et al., 1983; Clausen et al., 2011). 2.3. Zygosity determination School teachers participated in the twin recruitment by confirming that pairs included in the sample were in fact siblings and by determining twins' zygosity. Teachers answered a twin physical similarity questionnaire that is over 98% accurate to determine zygosity (Christiansen et al., 2003). Questions are about childhood similarity. We validated teachers' determination of zygosity in 108 twin pairs, administrating the physical similarity questionnaire to the twins' mother or father and their teacher. Only in 8 cases there were discrepancies between answers from both sources. In those

Please cite this article as: Iranzo-Tatay, C., et al., Genetic and environmental contributions to perfectionism and its common factors. Psychiatry Research (2015), http://dx.doi.org/10.1016/j.psychres.2015.11.020i

C. Iranzo-Tatay et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

cases, the questionnaire was repeated and disagreement was solved in 6 cases, 5 in favor of teacher's first answer and one in favor of parents'. The Cohen kappa statistic was 0.94 (p o0.001). This statistic measures the agreement between two or more observers when they are measuring the same object or target. A kappa of 1 indicates perfect agreement, whereas a kappa of 0 indicates agreement equivalent to chance. Our results indicate a high agreement between teachers and parents and so, that our method to determine zygosity is accurate.

3

the full AE model, which estimates the genetic, and unique environmental correlations, was compared to the following nested submodels, setting one or more of these correlations to zero. Model selection was also based on Akaike's Information Criterion (Akaike, 1987) and the Bayesian Information Criterion. The lowest AIC and BIC values among all models are suggestive of the best fitting model with regard to precision and complexity (Posada and Buckley, 2004). The statistical package used to run both univariate and bivariate twin models, was the Mx program (Neale et al., 2006).

2.4. Statistical analysis 2.4.1. Descriptive statistics We analyzed the data with SPSS v. 17. Descriptive statistics of the psychological traits measured were calculated and the differences between groups (according to sex and zygosity) were examined by means of the Student's t test. 2.4.2. Equal environments assumption (EEA) The equal environments assumption (EEA) holds that trait-relevant environments are equally correlated among members of monozygotic (MZ) and dizygotic (DZ) twin pairs (Mitchell et al., 2007). We tested the EEA by calculating Pearson coefficient correlations (r) between the total value of co-socialization and parents treatment and the intra-pair difference in the Perfectionism scale and sub-scales (self-oriented and socially prescribed). The co-socialization scale is composed by four items regarding differential experiences during childhood, such us “Did your twins attend to the same school?” or “Did your twins practice sports together?” A significant negative correlation (po 0.05) would indicate grater similarities of these psychological traits among those twins with a greater level of common experiences, and so, EEA violation. 2.4.3. Univariate models Classic twin studies assess factors influencing liability to a latent phenotype comparing the resemblance of MZ twins with the resemblance of DZ twins for that phenotype (covariation) (Neale et al., 2006). Based on this knowledge, we used univariate structural equation modeling for the perfectionism scale and its two sub-scales (self-oriented and socially prescribed). This way, the proportion of variance of these traits due to additive genetic effects (A), shared (or common) environmental effects (C) and unique (specific or non-shared) environmental effects (E) was estimated. Initially, a full model (ACE) was fit to the data, followed by an AE model, a CE model, and a model containing only non-shared environment (E). The goal of model fitting is to explain the observed data as an optimal combination of goodness-of-fit and parsimony. Akaike's Information Criterion (Akaike, 1987) states that the more negative the value is, the better fitness of the model has (Neale and Cardon, 1992; Markon and Krueger, 2004; Wade and Bulik, 2007). 2.4.4. Bivariate models Bivariate correlated factor models were used to assess the relationship between both constructs of perfectionism (self-oriented and socially prescribed). This model specifies the correlation between genetic factors (rA), unique environmental factors (rE), and common environmental factors (rC) between each of the traits' causal factors, A, C and E. These correlations represent the extent to which the same genes or environmental factors contribute to the observed phenotypic correlation among the variables. Given that the univariate estimates of C were zero for each of the two perfectionism constructs and that the most parsimonious univariate models for the two constructs were the AE model, only this model was examined in subsequent bivariate analyses. The fit of

3. Results 3.1. Participants characteristics The sample was composed of 258 twin pairs; 149 MZ twins (54.3%; 79 female twin pairs, 61 male twin pairs) and 118 DZ same sex twins (45.7%; 70 female twin pairs, 48 male twin pairs). Mean age at the time of the study was of 14.2 71.1 years. No statistical differences were found due to sex or zygosity. 3.2. Description of perfectionism measures Table 1 presents results for perfectionism scale mean scores obtained according to sex and zygosity. There were no statistical differences between boys (5.10 73.58) and girls (4.61 73.66) for the perfectionism mean scores. The same occurred when mean scores for the two constructs of perfectionism (self-oriented and socially prescribed) were evaluated. Table 1 also presents Pearson correlation coefficient (r) between the two constructs, based on sex and zygosity. Correlations were significant at a 0.01 level in all cases. 3.3. Equal environments assumption (EEA) There were no threshold differences in frequency and quality of contact with parents or measures of co-socialization between MZ and DZ twins. Moreover, there was no statistical negative association between measures of co-socialization and/or in frequency and quality of contact with parents with intra-pair resemblance for any of the constructs of perfectionism. Therefore, no significant violation of the equal environments assumption (EEA) was found. 3.4. Univariate twin analyses Pearson correlation coefficients between members of the twin pair for the different perfectionism subscales are presented in Table 1. To compare the resemblance of MZ twins with DZ twins for a trait or disease is the first step of the evaluation of the degree to which genetic and environmental influences affect phenotypic variation of that trait (Boomsma et al., 2002). When the correlation between MZ twins is higher than between DZ twins, it is possible to deduce that the phenotype will show a heritability estimate. Furthermore, if we compare separately the correlations between boys and girls, we would obtain specific information about quantitative differences in the heritability of the phenotype due to sex. For example, in socially prescribed perfectionism in girls, MZ twin correlations (0.409) were consistently much higher than DZ twin's (0,262) (Table 1). On the contrary, a MZ twin's resemblance similar to DZ twin's indicates a low heritability of the trait. For socially prescribed perfectionism in boys, similar patterns of twin correlations were found, suggesting that environmental factors would probably have an important influence on the trait. Tables 2 and 3 show results of fitting separate univariate ACE models to each perfectionism subscale (total perfectionism, self-

Please cite this article as: Iranzo-Tatay, C., et al., Genetic and environmental contributions to perfectionism and its common factors. Psychiatry Research (2015), http://dx.doi.org/10.1016/j.psychres.2015.11.020i

C. Iranzo-Tatay et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

4

Table 1 Perfectionism subscale of the Eating Disorders Inventory (EDI) mean scores in the twin sample (N ¼ 516). Total perfectionism

N pairs X

S

Intra-pair r

Self oriented perfectionism

Socially prescribed perfectionism

Correlation self oriented-socially prescribed perfectionism

X

S

Intra-pair r

X

S

r

r

Girls

Mz 79 Dz 70 Total 149

4.92 3.78 0.466 4.26 3.51 0.223 4.61 3.66 0.366

1.71 1.44 1.58

2.02 1.90 1.97

0.317 0.143 0.244

3.22 2.83 3.03

2.51 2.19 2.37

0.443 0.247 0.368

0.383* 0.464* 0.420*

Boys

Mz 61 Dz 48 Total 109

5.20 3.57 0.403 4.99 3.60 0.213 5.10 3.58 0.293

1.85 1.78 1.82

1.93 2.16 2.03

0.208 0.113 0.153

3.34 3.21 3.28

2.20 2.21 2.20

0.361 0.297 0.302

0.500* 0.356* 0.432*

Total Mz 140 Dz 118 Total 258

0.435 0.217 0.339

0.269 0.131 0.206

0.409 0.262 0.345

DZ: Dizigotics; MZ: Monozigotics; Intra-pair r: Pearson correlation coefficients between members of the twin pair; r: Pearson correlation coefficient between self-oriented perfectionism and socially-prescribed perfectionism according to sex and zigosity; X: mean. S: Standard deviation. *

Significant correlation at a 0.01 level.

oriented and socially prescribed) in girls and boys respectively. In both sexes and in both constructs of perfectionism, the AE model provided an AIC that showed the most parsimonious explanation. This model indicates the influence of genetic factors (A) and specific environmental factors (E) on its variance, and, at the same time, dismisses a common environmental influence (C). Differences regarding sex were seen for both constructs of perfectionism, meaning that in all cases girls showed a higher genetic component (45%, 30% and 42% in girls versus 40%, 23% and 38% in boys for total, self-oriented and socially prescribed perfectionism respectively). For self-oriented perfectionism among boys, even though the AE model was chosen as the most parsimonious, the 95% confidence interval included 0% which may limit results consistency. This did not happen in the other models, which ensures the validity of the results.

3.5. Bivariate twin analyses Tables 4 and 5 show, in girls and boys respectively, bivariate model-fitting results for both subscales (self-oriented and socially prescribed), as well as their genetic (ra) and environmental correlations (rc, re). In both sexes, the best fitting model (AE-ra,re) partitioned the covariation between self-oriented and socially prescribed perfectionism into shared additive genetic components (A), shared unique environmental components (E). This model shows a high genetic correlation (ra) estimated in girls of 0.704 CI 95%: 0.401–1.000) (Fig. 1) and in boys of 0,981 (CI 95%: 0.425– 1.000) (Fig. 2). A significant environmental correlation (re) of unique environmental factors of 0.259 (CI 95%: 0.077–0.425) in girls, and of 0.254 (CI 95%: 0.065–0.452) in boys can also be observed.

Table 2 Univariate structural equation models in girls of; Total Perfectionism, self-oriented perfectionism (SOP) and socially-prescribed perfectionism (SPP). Goodness of fit statistic and relative contributions of genetic and environmental parameters. Model

Variance Components A: additive genetic

Fit Statistics C: common environment

E: unique environment

χ2

Gl

P

∆χ 2( P*)

AIC

ACE AE CE E

Socially prescribed perfectionism (SPP) n¼158 MZ, 140 DZ 0.247 (0.000–0.555) 0.161 (0.000–0.482) 0.592 (0.444–0.769) 0.421 (0.254–0.561) – 0.579 (0.438–0.746) – 0.363 (0.215–0.495) 0.637 (0.505–0.785) – – 1.000 (1.000–1.000)

3.564 3.935 4.326 25.132

3 4 4 5

0.312 0.415 0.364 0.000

Reference 0.371 (0.543) 0.761 (0.383) 21.567 (0.000)

 2.436  4.065  3.674 15.132

ACE AE CE E

Self oriented perfectionism (SOP) n¼ 158 MZ, 140 DZ 0.302 (0.000–0.467) 0.000 (0.000–0.367) 0.302 (0.115–0.467) – – 0.240 (0.082–0.385) – –

0.698 0.698 0.760 1.000

(0.533–0.897) (0.533–0.885) (0.615–0.918) (1.000–1.000)

3.576 3.576 4.543 13.225

3 4 4 5

0.311 0.466 0.337 0.021

Reference – 0.967 (0.325) 9.649 (0.008)

 2.424  4.424  3.457 3.225

ACE AE CE E

Total Perfectionism n¼158 MZ, 140 DZ 0.411 (0.000–0.587) 0.033 (0.000–0.449) 0.447 (0.276–0.587) – – 0.360 (0.211–0.492) – –

0.556 0.553 0.640 1.000

(0.413–0.738) (0.413–0.722) (0.508–0.789) (1.000–1.000)

2.424 2.441 4.613 25.005

3 4 4 5

0.489 0.655 0.329 0.000

Reference 0.017 (0.895) 2.189 (0.139) 22.581 (0.000)

 3.576  5.559  3.387 15.005

AIC ¼Akaike's Information Criterion; Δχ2 ¼chi-square difference test statistic; GL: degree of freedom; P: likelihood ratio χ2 test; P*: likelihood ratio test associated with the chi-square change for the full ACE saturated model; χ2: Chi-square.

Please cite this article as: Iranzo-Tatay, C., et al., Genetic and environmental contributions to perfectionism and its common factors. Psychiatry Research (2015), http://dx.doi.org/10.1016/j.psychres.2015.11.020i

C. Iranzo-Tatay et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

5

Table 3 Univariate structural equation models in boys of; Total Perfectionism, self-oriented perfectionism (SOP) and socially-prescribed perfectionism (SPP): Goodness of fit statistic and relative contributions of genetic and environmental parameters. Model

Variance Components A: additive genetic

Fit Statistics C: common environment

E: unique environment

χ2

Gl

P

∆χ 2 (P*)

AIC

ACE AE CE E

Socially prescribed perfectionism (SPP) n¼ 122 MZ, 96 DZ 0.130 (0.000–0.546) 0.230 (0.000–0.486) 0.640 (0.451–0.843) 0.388 (0.78–0.560) – 0.612 (0.440–0.822) – 0.331 (0.153–0.490) 0.669 (0.511–0.847) – – 1.000 (1.000–1.000)

1.123 1.280 1.762 13.727

3 4 4 5

0.771 0.865 0.779 0.017

Reference 0.157 (0.692) 0.639 (0.424) 12.604 (0.002)

 4.877  6.720  6.238 0.172

ACE AE CE E

Self oriented perfectionism (SOP) n ¼ 122 MZ, 96 DZ 0.226 (0.000–0.444) 0.000 (0.000–0.326) 0.226 (0.000–0.444) – – 0.159 (0.000–0.337) – –

0.774 (0.556–1.000) 0.774 (0.556–1.000) 0.841 (0.663–1.000) 1.000 (1.000–1.000)

3.414 3.414 3.911 6.660

3 4 4 5

0.332 0.491 0.418 0.247

Reference – 0.497 (0.481) 3.246 (0.197)

 2.586  4.586  4.089  3.340

ACE AE CE E

Total perfectionism n¼ 122 MZ, 96 DZ 0.395 (0.000–0.581) 0.012 (0.000–0.447) 0.409 (0.195–0.581) – – 0.318 (0.138–0.477) – –

0.593 0.591 0.682 1.000

0.268 0.270 1.676 13.081

3 4 4 5

0.966 0.992 0.795 0.023

Reference 0.002 (0.967) 1.408 (0.235) 12.813 (0.002)

 5.732  7.730  6.324 3.081

(0.419–0.824) (0.419–0.805) (0.522–0.862) (1.000–1.000)

AIC ¼Akaike's Information Criterion; Δχ2 ¼ chi-square difference test statistic; GL: degree of freedom; P: likelihood ratio χ2 test; P*: likelihood ratio test associated with the chi-square change for the full ACE saturated model; χ2: Chi-square.

Table 4 Bivariate model fit statistics for self-oriented perfectionism and socially-prescribed perfectionism in girls. Model

Phenotypic correlation

Goodness-of-fit of model

ra: Genetic correlation

AIC

BIC

Estimates AE-ra,re 0.704 AE-re

–

AE-ra

0.924

AE

–

re: Unique environment correlation CI 95%

Estimates

CI 95%

0.4011.000 –

0.259

1346.241

 210.040

1357.194

 206.065

0.687– 1.000 –

–

0.0770.425 0.3430.565 –

1351.919

 208.703

–

–

1410.435

 180.947

0.462

Fig. 1. Girls bivariate correlation factors model path diagram. CI 95%: confidence interval at 95%. AIC: Akaike's Information Criterion. BIC: Bayesian Information Criterion. Table 5 Bivariate model fit statistics for self-oriented perfectionism and socially-prescribed perfectionism in boys. Model

Phenotypic correlation

Goodness-of-fit of model

ra: Genetic correlation

AIC

Estimates AE-ra,re 0.981 AE-re

–

AE-ra

1.000

AE

–

re: Unique environment correlation CI 95%

Estimates

CI 95%

0.425– 1.000 –

0.254

0.804– 1.000 –

–

0.065– 0.452 0.332– 0.573 –

–

–

0.459

CI 95%: confidence interval at 95%. AIC: Akaike's Information Criterion. BIC: Bayesian Information Criterion.

BIC

973.821

 89.038

979.564

 87.512

978.864

 87.862

1052.523

 52.378

Fig. 2. Boys bivariate correlation factors model path diagram.

4. Discussion This is the first investigation of genetic and environmental influences on perfectionism in a Spanish adolescents sample.

Please cite this article as: Iranzo-Tatay, C., et al., Genetic and environmental contributions to perfectionism and its common factors. Psychiatry Research (2015), http://dx.doi.org/10.1016/j.psychres.2015.11.020i

6

C. Iranzo-Tatay et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Moreover, this is the first study to explore the influential nexus between socially prescribed and self-oriented perfectionism. Briefly, the specific aims of our work were: (1) to determine the relative magnitude of environmental and genetic effects among two components of perfectionism (socially prescribed and selforiented) in an adolescents twin sample; and (2) to determine whether those effects are shared between both components of perfectionism. The equal environments assumption, which holds that traitrelevant environments are equally correlated among monozygotic (MZ) and dizygotic (DZ) twin pairs, is essential in twin designs. Violations of this assumption could lead to biased parameter estimates in twin models (Mitchell et al., 2007). If the EEA is violated (i.e. when more similar treatment of monozygotic (MZ) compared to dizygotic (DZ) twins has an effect on twin similarity for the phenotype), estimates of heritability will be over-estimated (Wade et al., 2003). The validity of the EEA is trait-specific and must be examined for each diagnosis of interest (Mazzeo et al., 2010), just as we did in our study. The EEA has been questioned, and even a higher environment similarity among MZ twins compared to DZ twins may not hold an EEA violation. This means that causal effects may go from the twins to their environment. In this case, known as an active gene  environment correlation (Plomin et al., 1977) what we interpret as environmental similarity could be an outcome of genetic effects rather than a cause of bias (Mitchell et al., 2007). We investigate the EEA with respect to both constructs of perfectionism, self-oriented and socially prescribed. Our test of the EEA in the two measures of perfectionism, with respect to similarity of childhood treatment and co-socialization, revealed no violations of the EEA. Results demonstrated that mean scores obtained in the total perfectionism scale are relatively low when compared to those observed in clinical samples, regardless of the mental pathology (eating disorders or others) (Gustafsson et al., 2010; Clausen et al., 2011; Giovanni et al., 2011; Hamel et al., 2012; Lee et al., 2012). Regarding its genetic and environmental contributions, we observed that heritability estimates are similar in girls and boys (44% and 41% respectively), being the rest of the variance due to specific environmental factors (E). We found no contribution of common environmental influences (C). We have already mentioned the Kamakura's study that also used the EDI Perfectionism subscale among a female twin sample. Their results defend an AE model, with genetic contribution to perfectionism accounting for 37% of the variance. Even though our findings are consistent with theirs, it is difficult to compare both studies given the methodological differences; sample characteristics (mean age of 21 years) and the different cultural environment where the sample was recruited (Japan) (Kamakura et al., 2003). It is interesting that, despite the cultural differences, heritability estimates for perfectionism are similar in both studies. As we know, the degree of perfectionism is significantly stronger in Western societies. This line of evidence suggests that differences in the expression of the trait are not only due to genetic differences but probably due to environmental factors, which could be related to cultural differences in the importance of the sense of accomplishment (Tachi et al., 2007). Furthermore, we examined genetic and environmental contributions to both constructs of perfectionism. We found moderate heritability estimates for both socially prescribed and self-oriented perfectionism. Given the confidence intervals, the true heritability for socially prescribed perfectionism appears to be somewhere between 11% and 56% in girls and boys. For self-oriented perfectionism, heritability estimate in girls would be somewhere between 11% and 46%. Although in boys the AE was chosen as the best model, the true heritability estimate for self-oriented perfectionism ranged from 0% to 44%. This could mean that this construct could only be influenced by environmental factors in

boys. Some differences were seen between boys and girls, especially for self-oriented perfectionism which showed higher genetic variance among girls. As we described previously self-oriented perfectionism focuses on excessively high standards, including setting exacting standards for oneself, and striving to attain perfection in one's own endeavors as well as striving to avoid failure. This sex differences could be due in part to variations in the manifestation of gene-environment interactions. Our occidental culture is probably more demanding in terms of achievement and performance among women. With the feminist movement of the late 1960s, societal attention has become more firmly focused on the position and status of women (Gilbert, 1985). This environment could lead to genetic interactions, and could explain a higher heritability for the expression of perfectionism in girls. In short, perfectionism constructs´ expression is in part due to the cumulative impact of many individual genes each of small effect that act additively (Bulik et al., 2000). In all cases, the rest of the variance was attributable to unique or specific environmental influences to which one member of a twin pair is exposed but not the other (Bulik et al., 2000). Common environmental effects did not show any influence in the expression of self-oriented and socially prescribed perfectionism. Our univariate results are difficult to compare to previous studies, which have examined dimensions of perfectionism by means of the MPS-F scale (Frost Perfectionism Scale) (Tozzi et al., 2004; Wade and Bulik, 2007). This scale includes constructs related to perfectionism clinical phenomenology (such as concern over mistakes, personal standards, and doubts about actions), that do not match with the variables that we measured in the current investigation. An original aspect of our study relies on the bivariate equation model analysis that allows an approximation to the association between the genetic and environmental components of self-oriented and socially prescribed perfectionism. Our findings evidenced an extremely high association between genetic determinants of self-oriented and socially prescribed perfectionism. That is to say, they do not only have similar heritability estimates, but they also share a similar genetic contribution (genetic correlation in girls of 0.7 and 0.98 in boys), which may mean that there is an underlying common biologic substrate. Findings regarding unique environmental experiences showed a lower association (around 0.25 in both sexes), suggesting that there is a wide scope of action for individual experiences that can facilitate the expression of one trait or the other. This is not surprising given the AE model that was found in self-oriented and socially prescribed perfectionism, which indicates that differential unique experiences have an important influence in the expression of one psychological trait or the other. Therefore, we cannot expect a high coincidence between them. This line of evidence paves the way to future investigation, and suggests that there are certain circumstances in the individual specific environment that turn out to favor the expression of selforiented perfectionism, socially prescribed perfectionism or both. The overall pattern of results does not support the notion of two independent constructs of perfectionism at least regarding its causal elements. On the contrary, both traits seem to share the same genetic influences and just a few of the environmental nonshared determinants. Those are specific of each person and do not seem related to any aspect of perfectionism. Our results indicate that perfectionism has an important genetic substrate, but its expression requires specific experiences which would probably determine the social and/or self-oriented expression of the trait. In light of these findings, we believe that both constructs of perfectionism are in fact the same dimension (given the nature of their high genetic correlation), that is expressed differently depending on the environmental circumstances (given the nature of the low non-shared environment correlation), at least in our Spanish adolescents sample. Our results may not be generalizable to other

Please cite this article as: Iranzo-Tatay, C., et al., Genetic and environmental contributions to perfectionism and its common factors. Psychiatry Research (2015), http://dx.doi.org/10.1016/j.psychres.2015.11.020i

C. Iranzo-Tatay et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

components of perfectionism measured throughout other perfectionism scales. Despite the fact that we did not measure perfectionism constructs using the Hewitt and Flett Multidimensional perfectionism scale, as Sherry et al. described one half of the EDI-Perfectionism scale represents intrapersonal themes embodying what Hewitt and Flett have termed self-oriented perfectionism, while the other half of the EDI-Perfectionism scale reflects interpersonal themes approximating what Hewitt and Flett have called socially prescribed perfectionism (i.e., perceiving that others are demanding perfection of oneself) (Hewitt and Flett, 1991; Sherry et al., 2004). Even though it is difficult to equate our methodology to Hewitt and Flets's multidimensionality, our results do not support a distinction of two components of perfectionism, at least regarding its genetic and environmental dimensions. These findings can contribute to the planning of specific prevention strategies among the different mental illnesses where perfectionism has been established as a risk factor, given its noticeable environmental influence. There are several problems that limit generalizability of twin studies. First of all, twins may not be representative of general population. The representativeness assumption holds that twins do not exhibit different means or variances from the general population for psychological characteristics (Klump et al., 1999). This was not tested in our study. Moreover, sources of bias are potentially many and include unusual circumstances of twin pregnancy, birth and development or the possibility that the environments of MZ twins may be more alike than those of DZ twins (Eaves et al., 1978). Finally, twin studies also enclose selection bias. In our study twins were identified by matching on family name, date of birth, school and year of the survey. As zygosity was established from children's principals or tutors information, there was no selection bias attributable to parents. The present study has several limitations that should be recognized. First of all, we must point out that only twin pairs that participated in DITCA-CV eating disorders prevention program were included in the study. Therefore, recruitment was carried out in two steps, what usually entail more losses (Tambs et al., 2009). Secondly, twins may not be representative of general population and, therefore, results may not be generalizable to non-twin samples. Thirdly, in the ACE model, the E component includes unique environmental experiences but also includes measurement error, which contributes to generate differences in perfectionism measures in a twin pair. That hinders making interpretation of unique environment contribution (van den Berg et al., 2010). Fourthly, our sample was relatively small for a twin study. We included 516 participants (258 twin pairs; 149 MZ twin pairs and 109 DZ twin pairs). Thus, this sample issue has the potential to underpower to discern C effects. Finally, we measured two factors of perfectionism using the EDI perfectionism subscale, composed by very few items in each measure and so, the internal reliability is likely to be poor. As we pointed out, the DITCA-CV program is a study of risk factors, early detection and prevention of eating disorders which includes a large list of psychometric measures, but did not include a specific perfectionism measure such as the Frost Perfectionism Scale. The primary goal of the study is to prevent eating disorders, and not to measure their psychological traits. For this reason, we found it important to evaluate the genetic and environmental components of perfectionism in order to provide the best preventive intervention. This intervention would prioritize working on social or intrapersonal aspects of perfectionism or could avoid this distinction. 5. Conclusions Genetic influence among perfectionism and its two constructs,

7

self-oriented and socially prescribed, is moderate during adolescence. The expression of this personality trait is predominantly due to unique environmental factors/influences. Results revealed that these unique experiences appear to be specific for each construct, while they share, to a large extent, the same genetic influences. Perfectionism is related to different psychiatric disorders (Shafran and Mansell, 2001; Moser et al., 2012), and specially to eating disorders such us anorexia nervosa (Ruggiero et al., 2003). As Shafran and Mansell described, the past decade of research using a multidimensional approach to perfectionism has yielded few advances in the theoretical understanding or clinical treatment of specific psychiatric disorders. They concluded that, for that reason, in order to investigate it, a clearly defined notion is needed and not a group of constructs (Shafran and Mansell, 2001). Our results suggest that there is one genotype for perfectionism, which expresses differently depending on specific environmental experiences. So, these non-shared experiences determine the expression of one phenotype of perfectionism or the other. Future research will need to characterize the nature of the relationship between perfectionism and psychiatric disorders, where one or other phenotype of perfectionism would probably be expressed depending on the shared environmental influential factors of both the disorder and the temperamental factor.

Declaration of interest The authors have not transmitted any conflicts of interest based on business relationships of their own or of immediate family members.

Acknowledgements Amelia Granda Piñán participated in the translation of the manuscript from our native Spanish language to English.

References Akaike, H., 1987. Factor analysis and AIC. Psychometrika 52 (3), 317–332. Antony, M.M., Purdon., C.L., Huta, V., Swinson, R.P., 1998. Dimensions of perfectionism across the anxiety disorders. Behav. Res. Ther. 36 (12), 1143–1154 (PubMed: 9745799). Boomsma, D., Busjahn, A., Peltonen, L., 2002. Classical twin studies and beyond. Nat. Rev. Genet. 3 (11), 872–882 (PubMed: 12415317). Bulik, C.M., Sullivan, P.F., Wade, T.D., Kendler, K.S., 2000. Twin studies of eating disorders: a review. Int. J. Eat. Disord. 27 (1), 1–20 (PubMed: 10590444). Burns, D.D., 1980. The perfectionist's script for self-defeat. Psychol. Today 14 (6), 34–51. Christiansen, L., Frederiksen, H., Schousboe, K., Skytthe, A., von Wurmb-Schwark, N., Christensen, K., Kyvik, K., 2003. Age- and sex-differences in the validity of questionnaire-based zygosity in twins. Twin Res. 6 (4), 275–278 (PubMed: 14511432). Clausen, L., Rosenvinge, J.H., Friborg, O., Rokkedal, K., 2011. Validating the Eating Disorder Inventory-3 (EDI-3): a comparison between 561 female eating disorders patients and 878 females from the general population. J. Psychopathol. Behav. Assess. 33 (11), 101–110 (PubMed: 21472023). Eaves, L.J., Last, K.A., Young, P.A., Martin, N.G., 1978. Model-fitting approaches to the analysis of human behaviour. Heredity 41 (3), 249–320. Frost, R.O., Marten, P., Lahart, C., Rosenblate, R., 1990. The dimensions of perfectionism. Cogn. Ther. Res. 14, 449–468. Garner, D.M., Olmstead, M.P., Polivy, J., 1983. Development and validation of a multidimensional eating disorder inventory for anorexia nervosa and bulimia. Int. J. Eat. Disord. 2 (2), 15–34. Gilbert, L., 1985. Dimensions of same-gender student-faculty role-model relationships. Sex Roles 12, 111–123. http://dx.doi.org/10.1007/BF00288041. Giovanni, A.D., Carla, G., Enrica, M., Federico, A., Maria, Z., Secondo, F., 2011. Eating disorders and major depression: role of anger and personality. Depression Res. Treat. 2011, 194732 (PubMed: 21977317). Guimerá, E., Torrubia, R., 1987. Adaptación española del Eating Disorder Inventory (EDI) en una muestra de pacientes anoréxicas. An. Psiquiatr. 3, 185–190.

Please cite this article as: Iranzo-Tatay, C., et al., Genetic and environmental contributions to perfectionism and its common factors. Psychiatry Research (2015), http://dx.doi.org/10.1016/j.psychres.2015.11.020i

8

C. Iranzo-Tatay et al. / Psychiatry Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Gustafsson, S.A., Edlund, B., Kjellin, L., Norring, C., 2010. Characteristics measured by the Eating Disorder Inventory for children at risk and protective factors for disordered eating in adolescent girls. Int. J. Women's Health 2, 375–379 (PubMed: 21151684). Hamel, A.E., Zaitsoff, S.L., Taylor, A., Menna, R., Le Grange, D., 2012. Body-related social comparison and disordered eating among adolescent females with an eating disorder, depressive disorder, and healthy controls. Nutrients 4 (9), 1260–1272 (PubMed: 23112914). Hewitt, P.L., Flett, G.L., 1991. Perfectionism in the self and social contexts: conceptualization, assessment, and association with psychopathology. J. Pers. Soc. Psychol. 60 (3), 456–470 (PubMed: 2027080). Hewitt, P.L., Mittelstaedt, W., Wollert, R., 1989. Validation of a measure of perfectionism. J. Personal. Assess. 53 (1), 133–144 (PubMed: 2918450). Hollender, M.H., 1965. Perfectionism. Compr. Psychiatry 6, 94–103 (PubMed: 14298909). Kamakura, T., Ando, J., Ono, Y., Maekawa, H., 2003. A twin study of genetic and environmental influences on psychological traits of eating disorders in a Japanese female sample. Twin Res. 6 (4), 292–296 (PubMed: 14511435). Klump, K.L., Keel, P.K., Leon, G.R., Fulkerson, J.A., 1999. Risk for eating disorders in a school-based twin simple: are twins representative of the general population for eating disorder behaviour? Eat. Disord. 7, 33–41. Lee, J.H., Shin, M.Y., Jo, H.H., Jung, Y.C., Kim, J.K., Kim, K.R., 2012. Validation of the Korean version of the Eating Disorder Inventory-2: psychometric properties and cross-cultural comparison. Yonsei Med. J. 53 (6), 1099–1106 (PubMed: 23074108). Markon, K.E., Krueger, R.F., 2004. An empirical comparison of information-theoretic selection criteria for multivariate behavior genetic models. Behav. Genet. 34 (869), 593–610 (PubMed: 15520516). Mazzeo, S.E., Mitchell, K.S., Bulik, C.M., Aggen, S.H., Kendler, K.S., Neale, M.C., 2010. A twin study of specific bulimia nervosa symptoms. Psychol. Med. 40 (7), 1203–1213 (PubMed: 19818201). Mitchell, K.S., Mazzeo, S.E., Bulik, C.M., Aggen, S.H., Kendler, K.S., Neale, M.C., 2007. An investigation of a measure of twins' equal environments. Twin Res. Hum. Genet. 10 (6), 840–847 (PubMed: 18179396). Moser, J.S., Slane, J.D., Burt, A.S., Klump, K.L., 2012. Etiologic relationships between anxiety and dimensions of maladaptive perfectionism in young adult female twins. Depression Anxiety 29 (1), 47–53. Neale, M.C., Boker, S.M., Xie, G., Maes, H.H., 2006. MX: Statistical Modeling, 7th. Ed, VCU Box 900126, Richmond, VA 23298, Department of Psychiatry. Neale, M., Cardon, L., 1992. Methodology for Genetic Studies of Twins and Families. Kluwer Academic Publishers, Dordrecht. Patch, A.R., 1984. Reflections of perfection. Am. Psychol. 39 (4), 386–390. Plomin, R., Defries, J.C., Loehlin, J.C., 1977. Genotype-environment interaction and correlation in the analysis of human behavior. Psychol. Bull. 84 (2), 309–322 (PubMed: 557211). Posada, D., Buckley, T.R., 2004. Model selection and model averaging in phylogenetics: advantages of akaike information criterion and bayesian approaches

over likelihood ratio tests. Syst. Biol. 53 (5), 793–808 (PubMed: 15545256). Rojo-Moreno, L., Rubio, T., Plumed, J., Barbera, M., Serrano, M., Gimeno, N., Conesa, L., Ruiz, E., Rojo-Bofill, L., Beato, L., Livianos, L., 2013. Teasing and disordered eating behaviors in spanish adolescents. Eat. Disord. 21 (1), 53–69 [PubMed: 23241090]. Ruggiero., G.M., Levi, D., Ciuna, A., Sassaroli, S., 2003. Stress situation reveals an association between perfectionism and drive for thinness. Int. J. Eat. Disord. 34 (2), 220–226 (PubMed: 12898558). Shafran, R., Cooper, Z., Fairburn, C.G., 2002. Clinical perfectionism: a cognitive-behavioural analysis. Behav. Res. Ther. 40 (7), 773–791 (PubMed: 12074372). Shafran, R., Cooper, Z., Fairburn, C.G., 2003. “Clinical perfectionism” is not “multidimensional perfectionism”: a reply to Hewitt, Flett, Besser, Sherry & McGee. Behav. Res. Ther. 41 (10), 1217–1220 (PubMed: 12971941). Shafran, R., Lee, M., Payne, E., Fairburn, C.G., 2006. The impact of manipulating personal standards on eating attitudes and behaviour. Behav. Res. Ther. 44 (6), 897–906 (PubMed: 16257388). Shafran, R., Mansell, W., 2001. Perfectionism and psychopathology: a review of research and treatment. Clin. Psychol. Rev. 21 (6), 879–906 (PubMed: 11497211). Sherry, S.B., Hewitt, P.L., Besser, A., McGee, B.J., Flett, G.L., 2004. Self-oriented and socially prescribed perfectionism in the Eating Disorder Inventory Perfectionism subscale. Int. J. Eat. Disord. 35 (1), 69–79 (PubMed: 14705159). Shore, R.A., Porter, J.E., 1990. Normative and reliability data for 11 to 18 year olds on the eating disorder inventory. Int. J. Eat. Disord. 9 (2), 201–207. Tachi, T., Kusano-Schwarz, M., Murakami, K., Kobayashi, Y., Miyake, Y., 2007. Usefulness of the Eating Disorder Inventory-2 Japanese version in patients with eating disorders. Tokai J. Exp. Clin. Med. 32 (3), 78–82 (PubMed: 21318941). Tambs, K., Czajkowsky, N., Roysamb, E., Neale, M.C., Reichborn-Kjennerud, T., Aggen, S.H., Harris, J.R., Orstavik, R.E., Kendler, K.S., 2009. Structure of genetic and environmental risk factors for dimensional representations of DSM-IV anxiety disorders. Br. J. Psychiatry 195 (4), 301–307 (PubMed: 19794197). Tozzi, F., Aggen, S.H., Neale, B.M., Anderson, C.B., Mazzeo, S.E., Neale, M.C., Bulik, C. M., 2004. The structure of perfectionism: a twin study. Behav. Genet. 34 (5), 483–494 (PubMed: 15319571). van den Berg, P.A., Mond, J., Eisenberg, M., Ackard, D., Neumark-Sztainer, D., 2010. The link between body dissatisfaction and self-esteem in adolescents: similarities across gender, age, weight status, race/ethnicity, and socioeconomic status. J. Adolesc. Health 47 (3), 290–296 (PubMed: 20708569). Wade, T.D., Bulik, C.M., 2007. Shared genetic and environmental risk factors between undue influence of body shape and weight on self-evaluation and dimensions of perfectionism. Psychol. Med. 37 (5), 635–644 (PubMed: 17176504). Wade, T.D., Wilkinson, J., Ben-Tovim, D., 2003. The genetic epidemiology of body attitudes, the attitudinal component of body image in women. Psychol. Med. 33 (8), 1395–1405 (PubMed: 14672248). Webbink, D., Roeleveld, J., Visscher, P.M., 2006. Identification of twin pairs from large population-based samples. Twin Res. Hum. Genet. 9 (4), 496–500 (PubMed: 16899156).

Please cite this article as: Iranzo-Tatay, C., et al., Genetic and environmental contributions to perfectionism and its common factors. Psychiatry Research (2015), http://dx.doi.org/10.1016/j.psychres.2015.11.020i