Introduction

Attention-deficit and hyperactivity disorder (ADHD) is characterized by a persistent pattern of inattention and/or hyperactivity–impulsivity and essentially begins before the age of 12. A comprehensive meta-analysis all over the world indicated that the prevalence of ADHD in childhood was 5.3 % (Polanczyk et al. 2007). Two-thirds of the cases persist into adulthood either in residual symptoms or a thoroughgoing clinical syndrome (Biederman et al. 2000; Kessler et al. 2006). In one of the two epidemiological studies with the largest samples, the prevalence of ADHD was found to be 4.4 % among 3199 adults in the USA (Kessler et al. 2006). The other study screening even a higher number of adults (n = 11,422) in 10 countries in the Americas, Europe and the Middle East revealed an ADHD prevalence of 3.4 % (Fayyad et al. 2007). A meta-analysis of the studies on the prevalence of adult ADHD estimated that 2.5 % of adults suffered from ADHD (Simon et al. 2009).

A frequent comorbid condition in children with ADHD is obsessive–compulsive disorder (OCD). OCD is characterized by the presence of obsessions or compulsions or both of them. Zero to 7.5 % of children or adolescents with ADHD have OCD (Biederman et al. 2006; Carter et al. 2010; Geller et al. 2007; Ghanizadeh 2010; Lee et al. 2012; Moll et al. 2000; Skirbekk et al. 2011; Zohar et al. 1992). On the other hand, the frequency of ADHD in children or adolescents with OCD is much more common, ranging between 0 and 60 and literally above 20 % in 17 out of 29 studies (Canavera et al. 2010; Coskun et al. 2012; de Mathis et al. 2013; Farrell and Barrett 2006, 2012; Geller et al. 1996, 2000, 2001, 2003, 2008; Hanna 1995; Heyman et al. 2001; Ivarsson et al. 2008; Jaisoorya et al. 2003; Joshi et al. 2010; Lack et al. 2009; Langley et al. 2010; Lewin et al. 2010, 2011; Mancebo et al. 2008; Masi et al. 2006, 2010; Riddle et al. 1990; Toro et al. 1992; Storch et al. 2007, 2008, 2010; Walitza et al. 2008; Zohar et al. 1992). Various studies investigating the prevalence of OCD in adult ADHD have found it 1–13 % (Biederman et al. 2004; Kessler et al. 2006; Mannuzza et al. 1998; Millstein et al. 1997; Shekim et al. 1990; Wilens et al. 2009). It seems that OCD in ADHD is frequent despite one study (Mannuzza et al. 1998) contradicting this. On the other hand, ADHD in adults with OCD has been found from zero to as high as 23 % (Anholt et al. 2010; Brakoulias et al. 2011; Frost et al. 2011; de Mathis et al. 2013; Jaisoorya et al. 2003; Mancebo et al. 2008; Ruscio et al. 2010; Sheppard et al. 2010). Differences in methodology and diagnostic criteria and whether samples were selected from community or clinics may account for this great variability.

Recognizing OCD–ADHD comorbidity in children as well as adults is important in that individuals with both disorders at the same time have more unfavorable clinical course, lower social and educational functioning, greater susceptibility to substance use, poorer response to treatment, higher frequency of perinatal adverse experiences and different dimensions of obsessive–compulsive symptoms (Anholt et al. 2010; de Mathis et al. 2013; Farrell et al. 2012; Geller et al. 2003, 2008; Sheppard et al. 2010; Sukhodolsky et al. 2005).

To the best of our knowledge, the history of childhood ADHD in adults with OCD is yet to be searched. Investigating this may be important in a developmental perspective and in detecting the outcomes of ADHD and risk factors for OCD, since ADHD is always childhood-onset and OCD is usually adult-onset. We explored childhood symptoms of ADHD in an adult sample of OCD patients who had never been diagnosed with ADHD during their childhood. We hypothesized that childhood ADHD symptoms are common in adults with OCD and OCD has an earlier onset in those having a history of ADHD symptoms. We also compared the levels of impulsiveness between obsessive–compulsive adults having and not having childhood ADHD symptoms.

Methods

Subjects

Between January and September of 2014, 83 consecutive OCD patients (who identified OCD as the major mental problem of their life) between 18 and 72 years visiting Uskudar University Neuropsychiatry Health Practice and Research Center were enrolled in the study (43 women, 51.8 %; 40 men, 48.2 %). The project was approved by the Ethics Committee of the University, and all participants gave informed written consent. We excluded patients who had been diagnosed with ADHD during their childhood and those having the history of tic disorder, traumatic head injury, any other psychotic disorder, bipolar I or II disorders, mental retardation or any neurological disorder (such as epilepsy or multiple sclerosis). The diagnoses of the above-mentioned psychiatric disorders and adult ADHD were performed using the Structured Clinical Interview for DSM-IV (American Psychiatric Association 1994) Axis I Disorders.

Measures

Using the 25 item form of the Wender Utah Rating Scale (WURS) (Ward et al. 1993), we searched the symptoms of ADHD in childhood. The WURS is an instrument developed to assess the symptoms and signs of ADHD during childhood retrospectively and has the ability to differentiate adults with and without childhood ADHD. It is a Likert-type scale and its total score may vary 0 to 100. Its Turkish version (Oncu et al. 2005) has an excellent internal consistency (Cronbach’s alpha = 0.93) and a test–retest coefficient of 0.81. Its cutoff point is 36.

The severity of OCD was assessed by the Yale–Brown Obsessive Compulsive Scale (Y-BOCS) (Goodman et al. 1989). To measure the severity of depression and anxiety, we used the Hamilton Depression Rating Scale-17 (HDRS-17) (Williams 1978) and the Beck Anxiety Inventory (BAI) (Beck et al. 1988), respectively. Impulsiveness was measured by the Barratt Impulsiveness Scale-11 (BIS-11) (Patton et al. 1995) assessing the cognitive or attentional impulsiveness (CI), motor or behavioral impulsiveness (MI) and non-planning (NP).

Statistics

Statistical analyses were performed using SPSS version 17.0. To compare groups for continuous and categorical variables, independent samples t test and Chi-square test were used, respectively. Correlation analyses were performed using Pearson correlation coefficient.

Results

Table 1 shows the demographic and clinical characteristics of OCD patients with and without childhood ADHD symptoms. It is seen that 40.9 % (n = 34) of the whole sample, 41.9 (n = 18) of females and 40 % (n = 16) of males had childhood ADHD symptoms. The frequency of childhood ADHD symptoms did not significantly differ between the females and males. At the onset of OCD, the patients with childhood ADHD symptoms were significantly at younger ages (t[81] = 2.22, p = 0.03). The scores of the Y-BOCS and its obsession and compulsion subscales, HDRS-17, CI and NP, were similar between those having and not having childhood ADHD symptoms, whereas the BAI (t[81]) = 2.44, p = 0.02), BIS-11 total (t[81]) = 2.18, p = 0.03) and MI (t[819] = 2.69, p = 0.009) scores were greater in childhood ADHD group. The prevalence of current ADHD was 16.9 % (n = 14, five females, nine males).

Table 1 Demographic and clinical characteristics of OCD patients with and without childhood OCD
Full size table

Table 2 shows the correlations of the WURS with other measures. The scores of the WURS were significantly correlated only with the scores of the BAI, BIS-11 total and MI. The scores of the WURS did not show any significant correlation with scores of the Y-BOCS and its obsession and compulsion subscales, HDRS-17, CI and NP. The WURS scores were negatively correlated with age at the onset of OCD (r = −0.28, p = 0.01), that is, the higher the WURS scores, the earlier OCD begins.

Table 2 Correlations of Wender Utah Rating Scale with the severity of OCD, depressive symptoms, anxiety and impulsiveness
Full size table

Discussion

We found that the symptoms of childhood ADHD were quite frequent (40.9 %) in adults with OCD. However, 16.9 % of our patients had adult ADHD, that is, 41.2 % of those with childhood ADHD had current ongoing ADHD. There were no significant sex differences in obsessive–compulsive females and males with and without childhood ADHD symptoms. Patients having and not having childhood ADHD symptoms differed in severity of neither OCD nor depressive symptoms. On the contrary, the severity of anxiety was significantly higher in those with childhood ADHD symptoms. Subjects with the history of childhood ADHD symptoms when compared to the rest of the sample had significantly higher total and motor impulsiveness but similar levels of cognitive impulsiveness and non-planning. The onset of OCD was significantly higher in those showing childhood ADHD symptoms. ADHD scores showed significant correlation with the severity of anxiety, but not with the severity of OCD and depressive scores.

Unlike the prevalence of ADHD being 5.3 % in all children (Polanczyk et al. 2007), a prevalence of 40.9 % in the history of individuals with OCD is remarkably high, despite the exclusion of patients who had been diagnosed with ADHD during their childhood. This finding is consistent with the literature. Several previous studies done in children with OCD have found the prevalence of ADHD above 40:60 % in a child and adolescent psychiatry clinic (Coskun et al. 2012), 59 % within a sample of pediatric psychopharmacology clinic (Geller et al. 2000), 51 % in a pediatric OCD clinic (Geller et al. 2001), 44 % in a sample of child and adolescent OCD program (Geller et al. 2000), 43 % among the sample of a family genetic study (Joshi et al. 2010) and 42 % in an OCD specialty clinic (Geller et al. 2008). Prevalence of ADHD among those suffering from pediatric OCD has been estimated between 21 and 33 % in some other studies (Farrell et al. 2012; Geller et al. 1996; Ivarsson et al. 2008; Lack et al. 2009; Lewin et al. 2011 Mancebo et al. 2008; Storch et al. 2007, 2010; Walitza et al. 2008). There are also studies showing the frequency of ADHD in children and adolescents diagnosed with OCD between 9 and 19 % (Canavera et al. 2010; de Mathis et al. 2013; Farrell and Barrett 2006; Hanna 1995; Geller et al. 2003; Jaisoorya et al. 2003; Langley et al. 2010; Lewin et al. 2010; Masi et al. 2010; Riddle et al. 1990; Zohar et al. 1992, 1999). Out of 29 studies, only two found the rate of coexistence of ADHD with OCD below nine: 0 % in the study by Heyman et al. (2001) and 5.5 % by Toro et al. (1992).

We chose to use the term “childhood ADHD symptoms” rather than the “diagnosis of childhood ADHD,” because it is certain that making a diagnosis of ADHD by utilizing the WURS is not proper. Preferring such a methodological approach including the employment of self-report measure may bring about some vagueness in the evaluation of our findings. However, the WURS has a high Cronbach’s coefficient and largely been used in adult ADHD research. Furthermore, the fact that patients have been clinically interviewed has not yielded satisfactory information in the diagnosis of adult ADHD because the diagnosis of childhood ADHD in adults is inevitably blurred by recall bias, as is the case with self-report measures, since ADHD is a chronic illness essentially beginning in early life (Mannuzza et al. 2002).

The prevalence of current ongoing ADHD according to the DSM-IV in our sample was 16.9 %. Different studies investigating the prevalence of ADHD in adults with OCD revealed the figures of 22.9 % (Anholt et al. 2010), 19 % (Ruscio et al. 2010), 13.7 % (de Mathis et al. 2013), 12 % (Sheppard et al. 2010), 10 % (Mancebo et al. 2008), 8 % (Frost et al. 2011), 2 % (Jaisoorya et al. 2003) and 0 % (Brakoulias et al. 2011). It seems that, consistent with the majority of studies, adult ADHD is at least four times as common in patients with OCD as in general population (2.5–4.4 %). A study investigating coexisting ADHD in obsessive–compulsive adults who have been under clinical follow-up since their childhood would certainly yield more definite results. However, a prospective design is not appropriate for our aims or would require an epidemiological study recruiting a greater sample, because ADHD has to begin in childhood in definition, whereas OCD may begin at every age and prospective design would exclude adult-onset OCD cases. In an epidemiological study having a prospective design, Peterson et al. (2001) who randomly selected 976 children from families living in New York and reassessed 776 of them after 8, 10 and 15 years found that ADHD symptoms in adolescence predicted more obsessive–compulsive symptoms in adulthood and OCD in adolescence predicted more ADHD symptoms in adulthood. That study, having quite a larger sample but fewer patients with OCD when compared to the current one, supports our finding of the association between childhood ADHD and adult OCD.

Another finding of our study was that the ratio of obsessive–compulsive females and males with childhood ADHD symptoms did not differ significantly. In children, males have been diagnosed with ADHD two to three times in epidemiological samples and two to nine times in clinical samples (Nussbaum 2012). It seems that girls with ADHD might have a higher risk to develop OCD in later years than boys. This may be accounted for differences in the neuropathology of ADHD between boys and girls (Nussbaum 2012). On the other hand, it has been reported that in adulthood similar numbers of females and males apply to clinics with the complaints of ADHD in search for diagnosis and treatment, contrary to male dominance in childhood (Biederman et al. 2004). Different prevalence of ADHD in boys and girls might result from referral bias caused by the underdiagnosis of ADHD in girls. It should also be noted that our sample is not representative as we excluded those who had previously been diagnosed with ADHD and this might have also accounted for equal gender distribution.

How do OCD and ADHD or its symptoms so frequently occur at the same or different times in the same individuals although they have so distinctive, even opposite clinical phenomenology? Carlsson explained this apparent paradox suggesting that OCD and ADHD are the opposite diseases of the similar (prefrontal) brain regions: OCD is a hyperglutamatergic condition, whereas ADHD is a hypoglutamatergic one (Carlsson 2001). The frequent comorbidity of these antithetical conditions was further explained by the same author by the fact that: (1) Prefrontal glutamate systems are not stable, producing OCD symptoms when hyperactive and ADHD symptoms when hypoactive; (2) selected prefrontal glutamate neurons are hyperactive when others are hypoactive; and (3) in OCD, prefrontal glutamate neurons are exhausted due to severe hyperactivity periods and become hypoactive intermittently. It was also reported that as brain activity increases OCD gets more severe, while ADHD alleviates. These observations might partly explain the finding of ours that the severity of OCD did not differ between obsessive–compulsive patients with and without childhood ADHD symptoms.

We found no significant differences in severity of depressive symptoms between OCD patients with and without childhood ADHD symptoms. On the contrary, individuals with childhood ADHD symptoms had significantly more severe anxiety. Furthermore, ADHD scores measured with WURS were correlated with the severity of neither OCD nor depressive scores, but were significantly correlated with the severity of anxiety. More prominent relationship of ADHD with anxiety rather than depression has been supported by some studies. In a literature review, Biederman et al. (1991) detected that 25–35 % of children with ADHD had an anxiety disorder, whereas 9–32 had a depressive disorder. More recently, Yüce et al. (2013) found anxiety disorders in 49 % in children with ADHD, but depressive disorders in only 9 %. Kessler et al. (2006) estimated that, in adults with ADHD, the odds ratio of all anxiety disorders (with the exception of OCD) was higher than that of depression, but lower than that of dysthymia and bipolar disorder.

Levy proposed a behavioral inhibition concept related to mesolimbic dopamine systems to explain the high comorbidity between ADHD and anxiety disorders (Levy 2004). Reward and delay of reinforcement are determined by tonic/phasic dopamine effects, an impairment of which leads to fearless impulsive responses. A breakdown in the synaptic processes gating selectively fear or aggressive responses from the amygdala at the accumbens also plays a role in anxiety coexisting in ADHD. Prefrontocortical, hippocampal and amygdala projection neurons converge at the accumbens forming a monosynaptic intersection and allowing a synaptic gating mechanism to operate among all these structures. If the prefrontal cortical inhibition decreases, this mechanism is impaired and greater amygdala input becomes possible, and thus, anxiety-related processes produce more impact over the accumbens. In short, long-term tonic/phasic dopamine relationships and impaired prefrontocortical inputs to the synaptic gating of anxiety at the accumbens may account for the anxiety–ADHD comorbidity. On the other hand, some studies report that the coexistence of OCD and ADHD does not have a prominent effect on clinical features (Geller et al. 2002, 2003; Arnold et al. 2005; Sukhodolsky et al. 2005).

Our finding that impulsiveness was higher in patients with childhood ADHD symptoms than in others is not unexpected since impulsiveness is a remarkable characteristic of ADHD. High impulsiveness in these patients is caused by higher motor impulsiveness, whereas cognitive impulsiveness and non-planning did not differ between two groups. Researchers have usually found that OCD patients when compared to the controls have higher total and cognitive impulsiveness (Ettelt et al. 2007; Roh et al. 2009; Boisseau et al. 2012; Sohn et al. 2014) similar levels of motor impulsiveness and non-planning (Ettelt et al. 2007; Boisseau et al. 2012; Benatti et al. 2014; Sohn et al. 2014). Abramovitch et al. (2015) also reported that high impulsiveness in OCD results from cognitive impulsiveness. Two studies (Roh et al. 2009; Sohn et al. 2014) found higher motor impulsiveness in OCD patients than in controls, while one (Roh et al. 2009) did not reach statistical significance, possibly due to a small sample.

Although various studies have compared children with OCD having and not having coexisting ADHD, there are few studies done in adults. Educational, familial and psychosocial aspects of functioning have been found lower in children with comorbidity (Geller et al. 2003; Masi et al. 2006; Sukhodolsky et al. 2005). Children or adolescents with both disorders have a more frequent history of adverse perinatal experiences, which is also associated with other adversities (Geller et al. 2008) and an earlier onset of OCD (Masi et al. 2006, 2010; Walitza et al. 2008). The co-occurrence of pediatric OCD with ADHD is a predictor of poorer response to medical or psychological treatment (Farrell et al. 2012; Masi et al. 2010; Storch et al. 2008; Walitza et al. 2008). Walitza et al. (2008) who followed up children and adolescents having OCD with ADHD and having OCD but not ADHD for approximately 5 years observed that the severity of OCD did not significantly differ at the very beginning and at the end of an inpatient treatment between the two groups; nevertheless, those with coexisting ADHD had higher Y-BOCS scores (almost twice as much as OCD without ADHD) at the end of 5 years. Masi et al. (2006) found that OCD with ADHD was accompanied by a higher number of comorbid disorders including bipolar disorder, tic disorder and defiant disorder/conduct disorder. On the other hand, Walitza et al. (2008) found that OCD without ADHD coexisted with higher number of additional disorders (anxiety disorder, tic disorder, depressive disorders, specific phobia, adjustment disorder, anorexia nervosa, dyslexia, conduct disorder, enuresis) than OCD with ADHD (tic disorder, dyslexia, personality disorder adjustment disorder, enuresis). In the study by Walitza et al. (2008), it is remarkable that no patient with OCD plus ADHD had depressive or anxiety disorders. In one of the few studies recruiting adults, Anholt et al. (2010) found that individuals with comorbid OCD and ADHD were characterized by higher scores in an autism questionnaire and higher frequencies of cleaning and aggression-checking symptoms when compared to those with OCD not accompanied by ADHD. Anxiety disorders, depressive and bipolar disorders, eating disorders and substance abuse/disorders were equally prevalent in two samples consisting of adults as well as children with OCD having and not having ADHD; the only feature distinguishing the two groups was hoarding that was more prevalent in those with comorbidity (Sheppard et al. 2010). de Mathis et al. (2013), in an attempt to fathom the trajectory of OCD comorbidities, divided their sample comprising children and adults into groups according to the first diagnosed disorder. Those having been first diagnosed with ADHD and then OCD had higher frequencies of substance abuse and a progressive worsening course of OCD compared to those having been diagnosed first with OCD and then ADHD. Symptom dimensions of OCD, global impairment and the severity of OCD, depression and anxiety did not differ between two groups. Among our sample of adults who had a diagnosis of OCD but had never received a diagnosis of ADHD, those reporting childhood ADHD symptoms were characterized by an earlier age of OCD, more severe anxiety and higher impulsiveness than in the rest of the sample.

The main shortcoming of our study is the fact that our sample is not the representative of all patients with comorbid OCD and childhood symptoms of ADHD since we only recruited the patients who had never been diagnosed as having ADHD. Second, the current study is sensitive to recall bias. The good psychometric properties of the WURS may somewhat balance this limitation. Another shortcoming is related to the size of the sample. Although our sample is great enough to allow the comparison of obsessive–compulsive patients with and without childhood ADHD, it is too small to include those having current ADHD in comparisons. Prospective studies in epidemiological samples larger than that of Peterson et al. (2001) will surely be extremely useful although following up a huge number of non-clinical subjects is quite difficult. Future studies must research OCD patients with remitted and ongoing ADHD.

In conclusion, OCD and ADHD have a high a rate of comorbidity. Some children suffering the symptoms of undiagnosed ADHD will have a high risk of OCD when they grow up even if their ADHD remits. The history of childhood OCD in adults with OCD, though not associated with a more severe OCD, is associated with some features impairing the general clinical picture including higher levels of anxiety and impulsiveness and an earlier onset of OCD reflecting a more chronic illness.