The definition of ADHD used in Norway, and in our work, is based on the International Classification of Diseases (ICD-10) diagnosis hyperkinetic disorders (HKD, F90). HKD usually arises in the first 5 years of life. Six or more symptoms of inattention, three or more symptoms of hyperactivity and one or more symptoms of impulsiveness are necessary for the diagnosis 1011. The American Psychiatric Association's Diagnostic and Statistic Manual (DSM IV) 12 identifies three subtypes of ADHD: ADHD-predominantly inattentive (ADHD-PI), predominantly hyperactive-impulsive (ADHD-HI), and ADHD combined (ADHD-C). Only persons with ADHD-C will meet the HKD criteria. Due to the more restrictive inclusion and exclusion criteria of the HKD diagnosis in comparison to the DSM IV ADHD diagnosis the incidence rate is much lower for HKD than for ADHD. It is about 1–2% for HKD, compared to 5–10% for ADHD 10. Lahey et al. 13 found that 26% of younger children with ADHD met the criteria for HKD. This fact makes a direct comparison of the two diagnoses difficult.

In its definition of HKD the ICD-10 states that: "Impairment of cognitive functions is common, and specific delays in motor and language development are disproportionately frequent." 11. There is currently no acknowledgment of motor problems within the differential diagnosis section for ADHD in the DSM-IV 12. It seems to be assumed that the motor problems seen in ADHD either are separate, co-morbid conditions 6 or side effects of dysfunctional attention or impulsiveness. The differential diagnosis section for the DSM-IV diagnosis 'Developmental Coordination Disorder' (DCD), states that "Individuals with Attention-Deficit-Hyperactivity Disorder may fall, bump into things, or knock things over, but this is usually due to distractibility and impulsiveness, rather than to a motor impairment" 12. While much of the "clumsiness" observed in children with ADHD may be interpreted as an expression of attention problems, or general impulsivity, it is widely recognized, however, that many children with ADHD have motor impairments 67141516171819.

Pitcher et al. 18 have demonstrated that poor fine motor ability found in children with ADHD could not be attributed to deficits in attention or concentration, but rather to factors relating to their motor ability. Similar findings are reported by Miyahara et al. 20. Denckla and Rudel 7, using discriminant function analyses for speed, rhythm, and overflow correctly classified 89% of the boys as those with "hyperactive" versus "normal" behavioral histories. They concluded that neurological examination of "hyperactive" boys does reveal developmentally immature coordination. Uslu et al. 21 found the factor 'speed of movement', which contained items related to cerebral coordination of alternate muscle groups, to differ significantly between children with ADHD only and children with learning disorder (LD) or with ADHD-LD comorbidity. Kalff and collaborators 22 found that children at risk for ADHD were in general less accurate and more variable in their movements than children with other psychopathology and controls. Children with ADHD performed jerky movements and required more time than controls to change the direction of movement 2324. Raberger and Wimmer 25 showed that children with ADHD not only have fine motor problems, but also impaired balance.

An overlap of 30–50% has been reported between ADHD and the DCD diagnosis 182627. DCD is defined as a marked impairment in the development of motor coordination that significantly interferes with activities of daily living, and is not related to a medical condition 12. The standardized motor assessment battery M-ABC test 3, which addresses the areas manual dexterity, ball skills and balance, is commonly used in the diagnosis of DCD in studies assessing motor problems in ADHD 172829. It is quite possible that ADHD and DCD might coexist as separate disorders, and that the M-ABC test in an effective way reveals cases of comorbid ADHD and DCD. However, our clinical experiences are that children with ADHD may display motor problems in natural settings that are not uncovered using this test. A child with a high tone in m. Latissimus dorsi may obtain no problem scores on the ball items in the M-ABC test, but a restriction of the shoulder movement is often revealed by the MFNU when the child throws a ball with the arm in an upwards position. Neuropsychological tests like The Grooved Pegboard 30, the Maze coordination task 30, and the Finger-tapping test 4 yielded results similar to the M-ABC test 31.

Rasmussen and Gillberg 32 found that motor problems often persist into adulthood for persons with ADHD. Meyer and Sagvolden 16 found significantly poorer performance within the 3 ADHD groups, one for each subset of the DSM-IV diagnosis, compared to a control group of normal children on the Grooved Pegboard Test and Maze Coordination Task. Problems with motor control were less noticeable in the older groups. However, the authors argue that the differences found probably were due to the effect of maturation, which made the tasks too easy for this particular age group. No effects of gender were found in this study. In a retrospective study of 73 children with ADHD age 5–17 years (62 boys and 11 girls) assessed by items of MFNU, Stray 1 found that motor problem were present both in the younger ADHD group (age 5–10) and in the older group (age 11–17). There were no significant age effects on motor performance on any of the subtests neither in the medicine responder group nor in the non responder group. Gender differences in performance were not found.

Stray 1 and Stray et al. 9 hypothesized that there may be a functional relationship between the behavioral symptoms of ADHD and motor problems, which are not accounted for by attention deficits/impulsivity or comorbidity (DCD). Children with ADHD appear to have prefrontal mediated dysfunctions resulting in difficulties with impulse/inhibition control and self-regulation associated with higher order executive functions 333435. Several studies using the stop-signal task have shown a poorer inhibition performance and longer stop-signal reaction time in children with ADHD compared with a normal control group 3637.

Through the development period of the MFNU 1 motor inhibition and balance problems showed up as the clinically most significant aspects of motor problems observed in children who were later received the ADHD diagnosis. The study performed by Stray 1 indicated a close link between positive response to central stimulants and a high total problem score on the MFNU, suggesting that problems with motor inhibition and stabilisation of the trunk were functionally linked to ADHD. A two-part study was planned to investigate this issue in a controlled manner. The present article represents the first part of this study. Our aim has been to investigate the ability of the MFNU to discriminate between children with ADHD-C/HKD and a control group without behavioral symptoms of the condition.

Our research questions were:

To what extent do children with ADHD-C/HKD show motor problems, as measured by the MFNU, and how well does the test discriminate between children with ADHD and normal controls? We hypothesized that children with ADHD-C/HKD would display consistently high problem scores, and show significantly more motor problems on all the subtests of the MFNU, used in this study, compared to children without ADHD.

Fifty-two boys, all Caucasian Norwegians, participated in the study. Informed consent was given by the parents. Twenty-five drug naïve boys, aged 8–12 (mean 10.2 years, SD 1.3), out-patients of the Child Psychiatry Department of the Regional Hospital in Kristiansand participated to this study. They were recently diagnosed as HKD F90.0 12. The boys in the ADHD group were all candidates for methylphenidate evaluation. Diagnostic assessment was carried out by a physician or a clinical psychologist, based on clinical interview and other sources of information available, including parent and school reports for the last 12 months, reports from other health professionals, and behavioral observations during the assessment period. Children were excluded if they met criteria for conduct- or oppositional defiant disorder, a depressive or anxiety disorder, Asperger's or Tourette's syndrome, or epilepsy. The group displayed full scale IQs within the normal range (Mean 97.6, SD 15.6), as assessed with the WISC-R 54.

Twenty-seven boys, all Caucasian Norwegians, aged 8–11 years (Mean age 9.5 SD 1.1) without ADHD were recruited from two local schools, one for young and one for older children, to the control group. The schools were located in a suburban area with a predominantly Caucasian, low middle class population with no known history of socio-cultural problems. Written permission was obtained from the Local Education Authority of the municipality. A letter with information about the project, and Barkley's DSM-IV rating scale of ADHD 55 was distributed by the school to parents of all boys age 8–11 years. An affirmative reply to participate was returned in a prepaid envelope addressed to the first author of this article. To exclude possible ADHD problems in the control group the children were rated with Barkley's DSM-IV rating scale of ADHD both by parents and teachers. Subjects were excluded from the control group if they met the DSM-IV criteria for ADHD at home or at school. The broader ADHD criteria (compared to HKD) were used to decrease the probability of including children with a high incidence of ADHD symptoms. No precautions were taken to exclude other clinical groups. Of at total of 35 subjects agreeing to participate, 8 were excluded from the study and referred to the local medical centre for further assessment of ADHD.

WISC-R testing was not carried out in the control group due to resource constraints. Time restrictions and the special recruitment procedures applied, involving only freshly diagnosed children, prohibited a close matching of age between the groups. This resulted in a slightly older ADHD group, difference between mean age = 0.7 years, p < .05. Matching of socio-economic background was not performed for the same reasons.

The MFNU was developed by the first author during the 1990-ies in close collaboration with well-educated and specialized personnel trained within the fields of ADHD, learning and behavioral problems. Some of the subtests were originally developed, and some are modified versions of subtests of the Danish 'Funksjonsnevrologisk Undersøgelse (FNU)' 56. The subtests were primarily chosen and designed to reveal problems with motor inhibition and increased muscle tone.

In MFNU a qualitatively based scoring system is used. The test is performed in a very "dynamic" and interactional way with no limits concerning time and number of attempts in order to focus the attention of the child 19.

Most of the subtests of the MFNU (see Table 1) are performance tests where the child is given an instruction to perform a certain task (subtest 01–12). Subtests 13–16 are "passive" tasks where the tester evaluates muscular resistance while assessing the movement of the hips and feet. Item 17 'Synkinesis' is an evaluation of the presence of synkinetic movements during the examination. The subtest 'Palpation', which is included in the standard MFNU battery, was omitted from the study, as it can not be scored on the bases of video recordings. Therefore, only 17 subtests were used in the present study (see Additional files 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16).

A Cronbach's Alpha was calculated on the Total MFNU score in both groups. An Alpha of 0.98 indicates a very high relationship (internal consistency) within the total set of subtests 57 making the use of a total score for all variables meaningful, and supporting the assumption that the sum score may be treated as a continuous variable.

The MFNU is described in detail in a user manual and an accompanying DVD 9. Table 1 presents a brief description of the subtests used, video films of the subtests are given in Additional Files 3 to 16.

MFNU is scored on a sheet applying a ranked 3-category format (0-1-2). The subtests are scored according to the criteria in Table 2. More detailed criteria for each subtest are presented in the MFNU manual and visualized in the accompanying DVD 9.

Normally when evaluating test-retest changes in performance with the MFNU a 7-category scoring system is used 9. This system yields a more detailed picture of subtle changes in performance than the 3-category system. However, for the case of simplicity in the presentation of the results the basic 3-category scoring system was applied in the comparison of the two assessment sessions.

Preliminary inter-rater reliability analyses have shown high to very high degree of agreement among raters on all subtests (Cohen's Kappa ranging from 0.67 to 1.00) 1.

The present study involved two assessments with MFNU for all children 'Assessment 1' (A1) and 'Repeated assessment' (A2), with at an interval of at least 1 day (range 1–24 days, mean 4.54 days). The repeated assessment (A2) was performed in order to investigate possible training effects due to repeated testing. To prevent possible negative test results due to distraction or to emotional reactions to an unfamiliar situation, the ADHD group was given a preliminary session introducing the test situation and the different subtests to the parents and child. The assessment of the ADHD group took place at the hospital with the parents being present.

The control group was assessed at school. Practical restraints prohibited a preliminary consultation and participation of the parents during the assessment. Practical considerations and limited resources made it necessary to test the ADHD- and control group in different environments, thus prohibiting a blinded design. The subjects were assessed individually by a physiotherapist (the first author of this article) and videotaped by the physician (the fourth author). The videotapes were rated at a later point in time by a physiotherapist with no prior experience with the children. This physiotherapist had long experience both with administering the MFNU and with video rating MFNU sessions. To evaluate differences in performance between the two sessions, two parallel monitors were used. One scoring sheet was used for both assessments.

The statistic analyses were carried out using SPSS software 15.0. Descriptive statistics were used on data from each subtest in the ADHD and control group in order to view the percent distribution of the scoring categories (0-1-2) and on the variable 'Total score' which is the sum of the scores on 17 subtests. 'Total score' ranged from 0 – 34. The non-parametric Mann-Whitney U-test was used to compare the ranked scores of the ADHD and control group on each of the 17 subtest. Since the data were not assumed to be normally distributed, the Mann-Whitney test was used to compare the ADHD- and control group on the 'Total score'. Pearsons Chi square test was used, due to the small group sizes, when examining possible effects of age on motor performance. A Cronbach alpha analysis was performed to establish the internal consistency of the total set of subtests. The Wilcoxon Signed Rank Test for related samples was used to compare changes in performance on repeated measurements.

The study was approved by The Norwegian Data Inspectorate, The Regional Committee for Medical Research Ethics, the research committee and the director of medicine at the Department of Child Psychiatry, Sørlandet Hospital, Kristiansand, Norway, and the school authorities at the municipality of Songdalen, Norway, where the control group was recruited.