Obsessive–compulsiveness and impulsivity in a non-clinical population of adolescent males and females

Abstract

Obsessive–compulsive and impulsive behaviors co-occur in certain psychiatric conditions. Some have suggested that these disturbances constitute a spectrum of altered psychologies and behaviors that share an underlying neuropathology. We investigate here whether obsessive–compulsiveness and impulsivity reflect related psychological dimensions in a non-clinical adolescent population. Out of 720 high-school students, 672 and 682 completed a questionnaire interview with a Chinese version of the Maudsley Obsessive–Compulsive Inventory (MOCI) and the Barratt Impulsiveness Scale (BIS-11), respectively. Both MOCI and BIS-11 demonstrated good overall internal consistency, each with three major factors identified with Principal Component Analysis. In the 638 participants who completed both questionnaires, the total MOCI and BIS-11 scores did not correlate with each other. However, the MOCI factor “repetitive checking and attention to details” correlated negatively with the BIS-11 factor “inability to plan and look ahead” for all participants, and for males and females separately. The same MOCI factor also correlated negatively with the BIS-11 factors “lack of perseverance and self-control” and “novelty-seeking and acting without thinking” for all participants, and for females but not for males. The MOCI factor “doubt and intrusive thoughts” correlated positively with the BIS-11 factor “lack of perseverance and self-control” for all participants, and for males but not for females. These results suggested that the relationship between obsessive–compulsiveness and impulsivity as measured by the MOCI and the BIS-11 is complicated, with gender playing an important modulatory role. We discuss the relevance of these findings to developing a conceptual scheme to characterize and study the neurobiological basis of obsessive–compulsive and impulsive behaviors.

Introduction

Obsessive–compulsiveness and impulsivity co-occur in a number of psychiatric conditions (McElroy et al., 1994). Hollander et al. viewed these altered behaviors as the two core manifestations of the obsessive–compulsive spectrum disorders (OCSDs), which are disorders that involve an inability to inhibit or delay repetitive behaviors and the share important clinical characteristics with obsessive–compulsive disorder (OCD, Hollander and Wong, 1995, Hollander et al., 1996). Compulsivity, typified by the core symptoms of OCD, is characterized by exaggerated perception of harm and excessive, over-reflective responses, and difficult-to-control harm or risk avoidance behaviors. Impulsivity, on the other hand, is characterized by underestimation of harm, non-reflective responses, and difficult-to-control desires and repetitive behaviors to obtain pleasure and gratification. Although compulsivity and impulsivity seemingly describe two opposite extremes of these altered behaviors, most OCSDs can be characterized by a combination of compulsivity and impulsivity to varying extents. In fact, both of these behavioral features are often observed simultaneously or at different times in the course of the same illness (Hollander and Wong, 1995, Skodol and Oldham, 1995, Stein et al., 1996).

Evidence supporting the continuity of symptomatology and disease categories in OCSD comes mostly from the patterns of co-morbidity observed between OCD and OCSD. For instance, studies have shown higher rates of lifetime diagnosis of body dysmorphic disorder (BDD) in OCD patients (Simeon et al., 1995) and higher rates of BDD in case probands and relatives of OCD patients compared with non-proband controls (Bienvenu et al., 2000). Other studies have found a high co-occurrence of OCD in patients with BDD (Phillips et al., 1994, Veale et al., 1996), suggesting a strong diagnostic and etiological relationship between the two disorders. On the other hand, the association of other OCSDs, particularly those presenting predominantly impulsive behaviors, with OCD has been less consistent (Swedo and Leonard, 1992, Christenson et al., 1991, Bienvenu et al., 2000, Lochner et al., 2005). For instance, while some studies showed that patients with pathological gambling report more obsessions and compulsions, other studies failed to demonstrate increased OC symptoms in these patients (Frost et al., 2001, Kim and Grant, 2001, Anholt et al., 2004). Moreover, a recent study suggested important contrasting clinical characteristics between OCD and trichotillomania, including the age of onset of illness and treatment response to serotonin reuptake inhibitors (Lochner et al., 2005). Therefore, despite the symptomatological overlaps and possible etiological relationship among the OCSDs, the differential association of various OCSDs with compulsivity and impulsivity seems to suggest the utility of using the two behavioral features as opposing concepts in characterizing these disease entities.

Hoehn-Saric and Barksdale (1983) compared OCD patients with a history of poor impulse control with those without such history in a small sample of subjects. They found that, although both groups reported comparable OC symptoms, the impulsive group rated significantly higher on disturbances during childhood. They suggested that poor impulse control was related to disturbances that had already manifested themselves during childhood, while the OCD was superimposed at a later time. A more recent study sought to determine whether OCD patients with impulsive features constituted a distinct subtype of OCD (Matsunaga et al., 2005). Of more than 150 adult OCD patients, about one third had concurrent impulse control disorder (ICD), and could be differentiated from those without ICD by younger age of onset, more severe psychopathology and poor treatment outcome. The authors suggested that these results argued against a dichotomy between compulsive and impulsive disorders and that both behavioral features could be used as orthogonal dimensions to characterize related psychiatric conditions (Lacey and Evans, 1986, Rasmussen and Eisen, 1994, McElroy et al., 1995).

Taken overall, these studies have provided a prolific body of evidence suggesting the complexity in characterizing OCSDs in terms of obsessive–compulsiveness and impulsivity. However, the question remains whether these two behavioral features should best be used as correlated, opposing (negatively correlated) or orthogonal dimensions in conceptualizing OCSDs and in clinical psychological studies in general. The current study attempts to address this question by studying the association of impulsivity and obsessive–compulsiveness in a non-clinical sample. In particular, to circumvent the issue of developmental or adaptive superimposition of personality traits as suggested by Hoehn-Saric and Barksdale (1983), we examined for this association in an adolescent population. A positive association between these two personality traits would suggest a common biological basis of impulsivity and obsessive–compulsiveness and continuity in these behavioral attributes. A negative association would suggest conceptual diagonalization of these two behavioral features and impose important constraints on their underlying neurobiology. Finally, a lack of association would suggest that impulsivity and obsessive–compulsiveness represent different psychological constructs, without ruling out the possibility that one personality trait might affect the development of the other as the full personality profiles unfold for these individuals later in their lives.

A concurrent goal of the present study is to assess the association of obsessive–compulsive and impulsive personality features in an adolescent population. Previous work has demonstrated high co-morbidity between attention deficit hyperactivity disorder, Tourette's disorder and OCD, and has suggested an etiological link between obsessive–compulsive and disinhibitory psychopathologies in adolescents and younger children (Sherman et al., 1998, Stephens and Sandor, 1999, Niehaus et al., 2000, Cath et al., 2001, Apter et al., 2003, Guerrero et al., 2003, Mathews et al., 2004). It would thus be of interest to examine whether and which specific aspects of the association between the obsessive–compulsive and impulsive symptomatologies extend to a non-clinical population.