The present study included an estimation of ADHD prevalence among adult male longer-term prison inmates from a high-security Swedish prison. Further, we evaluated ADHD and executive functions among prison inmates and then compared results with ADHD psychiatric outpatients and healthy controls. We estimated a prevalence rate as high as 40% among these prison inmates. Further, those inmates we later confirmed with ADHD were severely affected and disabled from ADHD and coexisting disorders, such as SUD, ASD, personality disorders, mood- and anxiety disorders. Previous studies reported increased frequencies of major mental disorders, personality disorders, and early adjustment problems among prison inmates, regardless of ADHD . The present study confirms these observations. In addition, educational level and executive functions were poorer among ADHD inmates compared with ADHD psychiatric outpatients and controls. These findings remained after controlling for IQ. Thus, our findings imply prison inmates with ADHD to present a severely affected group of ADHD.
Although ADHD is common among prison inmates, prevalence rates are inconsistent, probably because of different used criteria among different prison populations [5–9]. Further, symptoms of ADHD, such as hyperactivity and impulsivity have shown to decline by age, whereas inattention and executive dysfunction continue . Besides, most prevalence studies on male prison inmates have been conducted among younger inmates . Further, knowledge is sparse on clinical features and executive functions among adult male prison inmates confirmed with ADHD [6–10] compared with adult ADHD among other groups and controls.
To our best knowledge, this study is the first to report a screening survey for ADHD, followed by extensive evaluations of ADHD and coexisting disorders among adult male longer-term prison inmates. The evaluations incorporated both self-reports and confirming collateral information from parents, medical records and school reports. Additionally, evaluations included a physical examination and neuropsychological assessments. Further, we compared ADHD prison inmates with ADHD psychiatric outpatients and controls for ADHD symptom load, coexisting disorders and executive functions.
Prevalence of ADHD among prison inmates
As hypothesized, ADHD was prevalent among these adult male longer-term prison inmates with a median age of 31 years. We estimated the prevalence as high as 40%, compared with previous findings by Rösler et al  who reported a prevalence of 45%, though among younger inmates (mean age 19). Thus, our results suggest ADHD to be comparably present among older and younger inmates. Our finding contradicts the common view of ADHD to decline by age. Thus, this symptom reduction by age might not held true for ADHD prison inmates. Further, the total survey response rate was 62%, which we view as acceptable, considering a common mistrust against authorities among prison inmates. However, we have to consider the attrition rate and its impact on the results. We imply that we not exaggerated the ADHD prevalence, as we did not approach inmates who we considered too psychiatric affected to take part. In some of these cases, ADHD might contribute to their psychiatric symptoms. On the other hand, we can not exclude some selection bias at the end of the study period when the study was more commonly known in the Swedish prison and probation service. It might be that some inmates recognised themselves as having ADHD and therefore applied for serving conviction at Norrtälje Prison in hope for treatment. However, as we screened the majority at the beginning of the study period, we imply this potential bias to be of minor importance. In summary, when considering the specificity of the screening procedure, we suggest a 40% ADHD prevalence rate among adult male longer-term inmates from a high-security prison.
Clinical characteristics of ADHD
This study only partially supported our hypothesis that ADHD prison inmates would present more severe ADHD symptoms across the lifespan, compared with ADHD psychiatric outpatients. The ADHD-prison group reported more ADHD symptoms and behaviours during both childhood and adulthood. However, collateral information from parents on childhood symptoms did not reveal any differences between groups. As a result, subjects rated more childhood symptoms retrospectively compared with parental ratings. This observation contradicts previous findings by Barkley  who displayed adults with ADHD to underreport their symptoms compared with parents. Thus, when considering the negative trajectory of these prison inmates and continuing ADHD symptoms, you would predict symptoms to be obvious during childhood, consistent with self-reports. Further, most subjects reported previous need of health services and educational support during childhood, pointing to obvious difficulties, although not recognised as ADHD. Notably, prison inmates showed a remarkably lower educational level compared with both other groups. Lower IQ levels among these inmates might partially explain these findings. Further, executive dysfunctions may contribute to lower school attendances and performances. In fact, we expect educational underachievement among ADHD also with normal IQ . Besides, more hindering symptoms from ADHD and coexisting learning disabilities, including dyslexia and externalising symptoms such as ODD and CD, possibly contribute to poorer educational achievements and early dropouts from school. Another explanation might be prison inmates exaggerating their symptoms in hope for methylphenidate treatment. However, parents of both ADHD groups rated similarly on Conners' Hyperactivity Index. This index reflects externalising symptoms besides ADHD, which is notable considering the negative trajectory of our ADHD-prison group. Therefore, self-reported childhood symptoms by prison inmates seem more in line with their negative trajectories across time. Further, symptoms of substance abuse, depression and anxiety could mimic ADHD. However, our inmates were kept from drugs for more than three months, in some cases for years. Further, all coexisting disorders were stable and treated at the assessment, thus implying present symptoms to be ADHD related.
To summarise, our findings imply the importance of recognising ADHD early and offering effective treatment immediately. Prospective studies should evaluate if treatment will reduce the risk for serious outcomes.
As hypothesized, coexisting disorders were common among our prison inmates. In fact, all subjects reported a lifetime history of SUD, with amphetamine as the most preferred drug of choice. Besides, abuse and antisocial behaviour had an early onset, consistent with previous findings . Additionally, anxiety disorders and depression were common, and half of inmates received treatment at the assessment. Further, all but one subject displayed CD before APD. Notably, psychopathy was present among only one tenth, which was fewer than we expected, as all but one subject displayed APD. However, previous studies reported that most psychopaths fulfil the criteria for APD, whereas the opposite is true for only a minority of inmates. These findings signal that psychopathy would be a more homogeneous disorder than APD . In addition, Soderstrom used a 3-factor model of PCL-R among forensic subjects for distinguishing psychopathy traits and evaluating if certain traits reflected ADHD . By this model, he showed that total PCL-R scores, as well as Factor 2 (unemotionality) and Factor 3 (behavioural dyscontrol), reflected ADHD. However, Factor 1 defining exaggerated self-opinion towards others and dishonesty did not reflect ADHD. In fact, the literature considers these interpersonal traits of Factor 1 to be most specific of psychopathy. Besides, we confirmed ASD among almost one fourth of ADHD prison inmates, mainly PDD-NOS. We are not aware of any previous reports estimating the prevalence of ASD among prison inmates. However, Anckarsater  showed that ASD was more common among forensic subjects than among the general population. In that study , PDD-NOS presented the most common ASD, paralleling our findings. In summary, we suggest that ASD is common also among prison inmates. However, studies comprising larger samples need to confirm these preliminary findings. If ASD is common among prison inmates, we need to consider this for successfully meeting the specific needs of these inmates.
Previous studies reported that personality disorders are common among different ADHD populations, such as prison inmates . Recently, Rydén et al observed that personality disorders were common among adults with "pure" ADHD, ADHD combined with bipolar disorder, and bipolar disorder only, although most prevalent among "pure" ADHD (Rydén E, and collaborators, personal communication). For defining personality disorders, they used the same procedure as in the present study. By comparing those, "pure" ADHD with our ADHD prison inmates, most personality disorders implied more common among inmates. However, histrionic, depressive, and schizoid personality disorder implied more common among "pure" ADHD subjects (Rydén E, and collaborators, personal communication).
The present study supported our hypothesis that ADHD prison inmates would present poorer cognitive abilities compared with ADHD psychiatric outpatients and healthy controls. As expected, the ADHD-prison group showed lower estimated IQ. However, different inclusion criteria could not explain the observed IQ differences between groups, as differences remained when excluding prison inmates with IQ < 85. As presented, both ADHD groups displayed poorer executive functions compared with controls, also when adjusting for IQ. Working memory functions were similar between ADHD groups when adjusting for IQ. Considering the CCPT results overall, controls and the ADHD-psychiatry group showed similar results. Further, at least one of them outperformed ADHD prison inmates on all accuracy dependent measures, and on several variability dependent measures, respectively. On the other hand, reaction time was comparable between groups, thus implying slow reaction time not to be a concern among adult ADHD. Summarising, these findings are in line with theories of ADHD as an executive disorder . In addition, these findings parallel recent reports by Wood et al [48, 49] who suggested that lower IQ does not account for the key cognitive problems noted among ADHD. Further, one striking notion of the present study, was the increased levels of Perseverations on the CCPT, which reflects difficulties in holding back or adjusting non-proper behaviours. Previous studies reported response perseveration among ADHD subjects suffering from CD [50–52], as well as among pathological gamblers . However, researchers interpreted response perseveration among ADHD in different ways. Quay [50, 51] suggested increased thriving for rewards among CD, because of a more actively working behavioural activation (reward) system (BAS) compared with the behavioural inhibition system (BIS). Reverse, he suggested less active BIS compared with BAS among ADHD. Beauchaine interpreted the opposite way , as he suggested less active BAS among CD, resulting in a reward-seeking behaviour as a stimulation seeking. Finally, Seguin , Newman and Wallace [56, 57] respectively, suggested deficits in attending for peripheral information, which usually directs the subject changing for a more effective behaviour. Therefore, future studies should explore the cognitive underpinnings of response perseveration, as they remain elusive.
We have to consider several limitations of this study. As the attrition rate of the screening survey was 38%, we must interpret the results with caution. However, we imply that we not exaggerated the prevalence rate of ADHD. Further, both rating scales used for screening lack Swedish validations. Nevertheless, these scales are used as standard tools in clinical practice. Besides, this study included only male longer-term prison inmates, why results can not extend to female inmates or to inmates serving shorter-term convicts. Further, there might have been selection bias when recruiting for the screening survey, mainly at the end of the study period when the study was commonly known in the Swedish Prison and Probation service. It might be that some inmates recognised themselves as suffering from ADHD and therefore applied for serving conviction at Norrtälje Prison in hope for treatment. However, as we screened the majority at the beginning of the study period, we imply this potential bias as minor important. Further, there might have been selection bias because of different inclusion criteria between groups. Therefore, it implies a selection of subjects among ADHD psychiatric outpatients, functioning better than average, as treatment for coexisting disorders excluded for the present study. Actually, in clinical practice, adults with ADHD often receive treatment for common coexisting disorders. On the other hand, the ADHD-psychiatry group may better reflect ADHD among the general population, as considered presenting less severe symptoms and severities compared with psychiatric outpatients. Additionally, there may have been a selection bias when recruiting prison inmates for diagnostic assessments. We noticed a few prison inmates denied taking part in the study in lack of motivation for changing their behaviour, or resistance to stay at the ADHD ward. The ward was apart from other wards for reducing the risk of exposing inmates to illicit drugs. As a result, study subjects received less time for physical exercise and restricted access to some prison programmes, as long as they stayed at the ADHD ward. Therefore, a selection bias towards more motivated prison inmates could have been present. If so, the bias probably worked towards better performances and higher functioning, than the reverse. Finally, the study samples were small. However, results were statistically significant despite small sample sizes. Notably, the strength of this study was the extensive clinical description of ADHD, coexisting disorders and executive functioning among prison inmates, as well as comparisons with ADHD psychiatric outpatients and controls. The extensive diagnostic evaluation included self-reported information, collection of collateral information, physical examination, structured diagnostic interviews and neuropsychological assessments. To our best knowledge, such extensive evaluations of longer-term prison inmates have not previously been reported. We infer our reported findings of ADHD symptom severity, coexisting disorders and executive functioning among prison inmates, are clinical important and relevant. We need to consider these severities when adjusting existing, or designing new ADHD treatment programmes for prison inmates. Further, these extensive evaluations might provide helpful insight for addressing future research on ADHD endophenotypes. Knowledge on endophenotypes may promote individually tailored treatments by identifying who will benefit from treatment. Finally, we will report effects of methylphenidate treatment among these ADHD prison inmates in another paper (Ginsberg and Lindefors, unpublished data).