2b
Susan C. South and Amber M. Jarnecke
In their chapter, Jang and Choi (this volume) highlight the challenges of using behavior and molecular genetics methods to understand the etiology of personality disorders (PDs). We resonate with much of what they said, including the measurement difficulties that plague the PD field. We would argue, and have previously argued, that the bulk of findings from behavior genetics research support a dimensional conceptualization of PDs (South & DeYoung, 2013). We assert here that continued refinement of the phenotype (PDs) is necessary to maximize the utility of genetics methods. Further, we argue in line with Jang and Choi (this volume), that for PDs the best use of genetics methods might be to understand environmental influences. Examining how the environment shapes the expression of genes presents a much greater opportunity for translating findings into efficacious prevention and intervention efforts.
Personality Disorder: What Is the Phenotype?
As Jang and Choi (this volume) briefly review, the PDs classified in the current, Fifth Edition of the Diagnostic and Statistical Manual (DSM-5; American Psychiatric Association, 2013) are exactly the same as they were in the Fourth Edition of the manual (American Psychiatric Association, 2000). They are plagued by multiple problems that make the study of these phenotypes difficult and therefore produce estimates in genetics research that might not be reliable or valid. Each PD, as outlined in DSM-5 Section II, is diagnosed according to a group of heterogeneous criteria that delineate maladaptive behaviors, thoughts, and feelings. The threshold needed for diagnosis is not based on scientific understanding and the criteria across the PDs are often overlapping if not identical (e.g., Widiger & Trull, 2007). Dissatisfaction with the multitude of problems with the DSM-IV PDs led to a proposed shift toward a dimensional conceptualization, but this proposal was abandoned (Krueger & Markon, 2014) and left as an alternative model in DSM-5 Section III (i.e., an appendix for future study).
The Section III DSM alternative model of personality disorders (AMPD) defines personality dysfunction using self- and interpersonal problems (Criterion A) that are associated with pathological personality traits (Criterion B; American Psychiatric Association, 2013). A dimensional trait model of five higher-order PD domains (and 25 lower-order facets) builds organically on the decades of research demonstrating associations between the categorical DSM PDs and the Five Factor Model (FFM) facets and domains (e.g., Widiger, 2011). Indeed, this trait model looks much like the maladaptive ends of four of the five domains of the FFM. There is substantial overlap between the AMPD domains and the FFM domains (Thomas et al., 2013). Neuroticism from the FFM maps on to Negative Affectivity, extreme Introversion on to Detachment, low Agreeableness on to Antagonism, and low Conscientiousness on to Disinhibition; Psychoticism in the DSM-5 model does not seem to align with the Openness domain of the FFM.
One study demonstrated that all three conceptualizations of PDs (i.e., Section II DSM, AMPD, and FFM) can be captured by five common latent factors/domains (Negative Affect, Detachment, Psychoticism, Disinhibition, Antagonism) in a structural model (Wright & Simms, 2014). Given the difficulties of defining and measuring PDs, these refined phenotypes may represent constructs of interest for geneticists. In fact, there is now a consortium of scientists dedicated to revising the conceptualization and classification of psychopathology broadly, along the lines of latent domains of psychopathology that capture covariance among different forms of pathology (Kotov et al., 2017). There is a growing interest in including PDs in these models but this has extended to the genetics literature more slowly.
Much of the extant literature supports a dimensional conceptualization of PDs, and therefore a ripe place for genetics research moving forward is to study the AMPD. The AMPD can be operationalized using the Personality Inventory for DSM-5 (PID-5; Krueger, Derringer, Markon, Watson, & Skodol, 2012). To our knowledge, only two studies have examined the heritability of the AMPD domains. Genetic variance on a shortened, Norwegian version of the PID-5 ranged from .25 (Psychoticism) to .37 (Detachment; South et al., 2017). In an adult US twin sample, Wright and colleagues (Wright, Pahlen, & Krueger, 2017) found that additive genetic influences on the facet scales from the full PID-5 ranged from .25 (Callousness, Irresponsibility) to .48 (Distractibility).
These heritability estimates suggest that more of the “action” is at the lower-order facet level. This makes sense, as the facets are more likely to be homogeneous and thus “purer” constructs (Smith, McCarthy, & Zapolski, 2009). Two individuals with the same score on a measure of neuroticism may have very different profiles along the neuroticism facets. As Jang and Choi (this volume) state, “Perhaps the way forward is to cease to do research on broad concepts.” If we take this argument to its logical conclusion, we might suggest that all work on genetics be done at the item (i.e., “nuance”) level. This may be too fine-grained an analysis, as the complexity of endorsing a particular item/behavior/nuance on any given day may render interpretation of resulting genetic and environmental influences near impossible.
Instead, we need a better sense of what constitutes “disorder” when considering PDs. For instance, narcissistic PD has been included in every version of the DSM, but there are still healthy debates in the field about the core features of the disorder (Miller, Lynam, & Campbell, 2016; Wright, 2016). If we are to move to a system that looks like the AMPD in DSM-5 Section III, then we will need to elucidate what makes a person’s thoughts, feelings, and behaviors disordered (rather than eclectic, quirky, or idiosyncratic). Some might argue that disorder is inherent in demonstrating extreme levels of maladaptive traits; whereas others might suggest that the way forward for PD research is to focus on Criterion A – identifying when a collection or profile of maladaptive traits becomes maladaptive according to the presence of a deficit or functional impairment (see Widiger et al., 2019). This is not simply a thought exercise, but a necessity for understanding the roots of personality pathology. Until we clarify the phenotype, genetics researchers will not be able to maximize all of their tools to understand the etiology of PDs.
Genetic Methods to Understand Environmental Influences
As Jang and Choi (this volume) mention, the genetics of PDs are understudied relative to normal personality or other forms of psychopathology. The extant literature suggests that PDs, although heritable, tend to show lower heritability than normal personality traits, and estimates range substantially depending on the PD in question or whether dimensional or categorical conceptualizations and measurements are used (Jang, Livesley, Vernon, & Jackson, 1996; Kendler et al., 2008; Reichborn-Kjennerud, 2010). The few candidate, association, and genome-wide association studies (GWAS) of PDs have identified specific genes or single nucleotide polymorphisms (SNPs) associated with PDs; however, in general, findings from studies using these methods fail to replicate unless very large sample sizes are used (Marigorta, Rodríguez, Gibson, & Navarro, 2018). The genetics field as a whole has recently turned to polygenic risk score or genome-based restricted maximum likelihood (GREML) methods. To date, there are no studies that use these methods to examine PDs. It is likely that GREML studies will find lower heritability of PDs than twin and adoptions studies because the heritability estimates using GREML are derived from a subset of genes; however, these GREML estimates may provide a lower bound of genetic variance that is not conflated with the assumptions of behavioral genetics studies (e.g., assumption of equal environments).
If nothing else, genetics research has showcased the complex relationship between genes and mental health disorders. This has moved the field forward in terms of generating new methodologies that hope to get us closer to understanding how genes relate to phenotypes. Although valuable, uncovering the relationship between genes and PDs using sample-specific, point estimates captured at a cross-section of time is limited in its utility. Thus, our position is aligned with what Jang and Choi (this volume) argue: maybe the story behind genetics research is less about finding all of the genes that contribute to PDs and more about focusing on the role of the environment.
One way to focus on the environment is by using twin and other family studies. Some may question the continued need for these types of quantitative behavior genetic methods when molecular genetic research is becoming more affordable, efficient, and widely available. We would argue that there are many reasons to continue using these methods. Use of twin and family studies provides estimates of genetic variance, common environmental variance (i.e., contexts that make twins similar to one another), and unique environmental variance (i.e., contexts that make twins different from each other). Examining the environmental variance components offers a starting point for understanding what type of contexts give way to PDs and helps to generate hypotheses about environments that may contribute to the expression of a PD (e.g., childhood socioeconomic status, which may make twins more similar, versus unique trauma exposure, which may make twins less similar).
In addition, researchers can examine how PDs share genetic and environmental variance with other phenotypes, such as normal personality or other forms of psychopathology. This may help tease apart the nature of comorbidity, for instance, along the lines of quantitative structural modeling of phenotypes as described above. Finding that the patterns of shared genetic variance between PDs and other forms of psychopathology differ from patterns of environmental variance (e.g., Kendler et al., 2011) will highlight the relative genetic and environmental contributions that give way to the comorbidity. Modeling of twin and family data is also flexible enough to make use of longitudinal designs so that both genetic and environmental variance and covariance can be examined over time (e.g., Bornovalova et al., 2013).
Another important way that biometric modeling can elucidate environmental effects on PDs is through study of gene–environment (G×E) interaction. These models allow researchers to investigate how a putatively environmental context, such as childhood abuse or dating violence, moderates genetic and environmental variance, in essence exploring how the environment impacts the expression of PDs (at one point in time and across the life course). G×Es have been explored broadly in the field of genetics, and particularly in behavior genetics investigations, over the last few decades. However, there have been relatively few behavior genetics studies of PDs that examine G×E and even fewer molecular genetics studies of PDs that include G×E. The lack of research in this area may be due, in part, to the fact that very few genetics studies include measures of PDs.
In the absence of direct measures of PDs, we can use G×E to examine how environmental variables impact the expression of endophenotypes presumed to be associated with PDs (e.g., production of proteins, neurotransmitter pathways, normal personality traits, affective responses). Moving forward this work will be important regardless of where future genetic studies of PDs take us. Uncovering the role of environment on endophenotypes may have even greater clinical utility than examining how environment impacts the genetic expression of the PD itself. That is, knowing how environment impacts the expression of a neurotransmitter pathway or affective responding may improve our ability to detect which populations respond best to treatment.
Summary
Behavior genetics research can offer much more than finding the genes that contribute to PDs. The heritability of personality, normal or maladaptive, will never reach 100 percent; the environment does have an impact on the variation in individual differences. Further, the environment most undoubtedly has a role on how genetic influences on PDs are expressed. Moving forward, we must challenge our conceptualization of what genetics research is and what it can do. We also argue for a refocusing of behavior and molecular genetics: back to the phenotype. By bringing our focus to the phenotype under study, greater strides will be made in understanding both genetic and environmental contributions to dysfunction in the characteristic patterns of how we think, feel, and behave.
References
American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders (4th ed., Text Revision). Washington, DC: American Psychiatric Association.
American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.
Bornovalova, M. A., Hicks, B. M., Iacono, W. G., & McGue, M. (2013). Longitudinal twin study of borderline personality disorder traits and substance use in adolescence: Developmental change, reciprocal effects, and genetic and environmental influences. Personality Disorders, 4(1), 23–32.
Jang, K. L., Livesley, W. J., Vernon, P. A., & Jackson, D. N. (1996). Heritability of personality disorder traits: A twin study. Acta Psychiatrica Scandinavica, 94(6), 438–444.
Kendler, K. S., Aggen, S. H., Czajkowski, N., Røysamb, E., Tambs, K., Torgersen, S., … Reichborn-Kjennerud, T. (2008). The structure of genetic and environmental risk factors for DSM-IV personality disorders: A multivariate twin study. Archives of General Psychiatry, 65(12), 1438–1446.
Kendler, K. S., Myers, J. M., & Keyes, C. L. (2011). The relationship between the genetic and environmental influences on common externalizing psychopathology and mental wellbeing. Twin Research and Human Genetics: The Official Journal of the International Society for Twin Studies, 14(6), 516–523.
Kotov, R., Krueger, R. F., Watson, D., Achenbach, T. M., Althoff, R. R., Bagby, R. M., … Zimmerman, M. (2017). The hierarchical taxonomy of psychopathology: A dimensional alternative to traditional nosologies. Journal of Abnormal Psychology, 126, 454–477.
Krueger, R. F., Derringer, J., Markon, K. E., Watson, D., & Skodol, A. E. (2012). Initial construction of a maladaptive personality trait model and inventory for DSM-5. Psychological Medicine, 42, 1879–1890.
Krueger, R. F., & Markon, K. E. (2014). The role of the DSM-5 personality trait model in moving toward a quantitative and empirically based approach to classifying personality and psychopathology. Annual Review of Clinical Psychology, 10(1), 477–501.
Marigorta, U. M., Rodríguez, J. A., Gibson, G., & Navarro, A. (2018). Replicability and prediction: Lessons and challenges from GWAS. Trends in Genetics, 34(7), 504–517.
Miller, J. D., Lynam, D. R., & Campbell, W. K. (2016). Measures of narcissism and their relations to DSM-5 pathological traits: A critical reappraisal. Assessment, 23(1), 3–9.
Reichborn-Kjennerud, T. (2010). The genetic epidemiology of personality disorders. Dialogues in Clinical Neuroscience, 12(1), 103–114.
Smith, G. T., McCarthy, D. M., & Zapolski, T. C. B. (2009). On the value of homogenous constructs for construct validation, theory testing, and the description of psychopathology. Psychological Assessment, 21(3), 272–284.
South, S. C., & DeYoung, N. J. (2013). Behavior genetics of personality disorders: Informing classification and conceptualization in DSM-5. Personality Disorders: Theory, Research, and Treatment, 4(3), 270–283.
South, S. C., Krueger, R. F., Knudsen, G. P., Ystrom, E., Czajkowski, N., Aggen, S. H., … Reichborn-Kjennerud, T. (2017). A population based twin study of DSM-5 maladaptive personality domains. Personality Disorders: Theory, Research, and Treatment, 8(4), 366–375.
Thomas, K. M., Yalch, M. M., Krueger, R. F., Wright, A. G. C., Markon, K. E., & Hopwood, C. J. (2013). The convergent structure of DSM-5 personality trait facets and five-factor model trait domains. Assessment, 20(3), 308–311.
Widiger, T. A. (2011). Integrating normal and abnormal personality structure: A proposal for DSM-V. Journal of Personality Disorders, 25(3), 338–363.
Widiger, T. A., Bach, B., Chmielewski, M., Clark, L. A., DeYoung, C., Hopwood, C. J., … Thomas, K. M. (2019). Criterion A of the AMPD in HiTOP. Journal of Personality Assessment, 101(4), 345–355.
Widiger, T. A., & Trull, T. J. (2007). Plate tectonics in the classification of personality disorder: Shifting to a dimensional model. American Psychologist, 62(2), 71–83.
Wright, A. G. C. (2016). On the measure and mismeasure of narcissism: A response to ‘Measures of narcissism and their relations to DSM-5 pathological traits: a critical reappraisal’. Assessment, 23(1), 10–17.
Wright, A. G. C., & Simms, L. J. (2014). On the structure of personality disorder traits: Conjoint analyses of the CAT-PD, PID-5, and NEO-PI-3 trait models. Personality Disorders: Theory, Research, and Treatment, 5(1), 43–54.
Wright, Z. E., Pahlen, S., & Krueger, R. F. (2017). Genetic and environmental influences on Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition (DSM-5) maladaptive personality traits and their connections with normative personality traits. Journal of Abnormal Psychology, 126(4), 416–428.
Both authors contributed equally to this work.