Abstract
Aim
We investigated relationships between hand function and genotype and aspects of phenotype in Rett syndrome.
Method
Video assessment in naturalistic settings was supplemented by parent-reported data in a cross-sectional study of 144 females with a mean age of 14 years 10 months (SD 7y 10mo; range 2y to–31y 10mo), 110 of whom had a mutation of the methyl CpG binding protein 2 (MECP2) gene. Ordinal logistic regression was used to assess relationships between hand function and MECP2 mutation, age, a modified Kerr score, Functional Independence Measure for children (WeeFIM), ambulation level, and frequency of hand stereotypies.
Results
Approximately two-thirds of participants demonstrated purposeful hand function, ranging from simple grasping skills to picking up and manipulating small objects. In participants with a confirmed MECP2 mutation, those with the p.R168X mutation had the poorest hand function on multivariate analysis with C-terminal deletion as the baseline (odds ratio [OR] 0.19; 95% confidence interval [CI] 0.04–0.95), whereas those with the p.R133C or p.R294X mutation had better hand function. Participants aged 19 years or older had lower hand function than those aged less than 8 years (OR 0.36; 95% CI 0.14–0.92). Factors that were associated with better hand function were lower Kerr scores for a 1-point increase in score (OR 0.77; 95% CI 0.69–0.86), higher WeeFIM scores for a 1-point increase in score (OR 1.08; 95% CI 1.04–1.12), and greater ambulation than those completely dependent on carers for mobility (OR 22.64; 95% CI 7.02–73.08). The results for participants with a confirmed pathogenic mutation were similar to results obtained when participants without a mutation were also included.
Interpretation
Our novel assessment of hand function in Rett syndrome correlated well with known profiles of common MECP2 mutations and overall clinical severity. This promising assessment could measure clinical responses to therapy.
Rett syndrome, which mainly affects females, is characterized by severe intellectual and physical disability and is associated with mutations in the methyl CpG binding protein 2 (MECP2) gene.1,2 Loss of hand function and hand stereotypies are core diagnostic features which emerge at the regressive period of the disorder.3,4
There is considerable clinical variability in Rett syndrome, depending in part on genotype5,6 and age.7 However, the full extent of variability within the domain of hand function is unknown. Total absence of hand function has been reported in association with some severe mutations but is less likely with milder mutations.5,6 The ability to grasp has been reported in 80% of females, to hold an object in 70%, and to self-feed in 25% to 43%.8,9 Although hand stereotypies are a hallmark of Rett syndrome, their relationship with hand function is unknown.
To date, items describing hand function have been relatively blunt; hand function has had little individual focus and has generally been assessed as one component of overall severity. A detailed analysis of hand function and its relationship with genotype and other aspects of phenotypic severity could inform a better understanding of Rett syndrome and methods for measuring hand function. Using a video assessment tool supplemented by parental reports,10 we examined hand function in naturalistic settings to create a novel measure. We also evaluated the relationships between hand function and genotype, age group, and clinical severity.
METHOD
The Australian Rett Syndrome Database (ARSD) is a population-based register of confirmed cases of Rett syndrome, with data collected from family and clinician questionnaires.11 In 2004 and 2007, families participating in the ARSD were invited to participate in the video study.10 Families who agreed to participate were sent a structured filming protocol, a demonstration video featuring four individuals with Rett syndrome of varying ages and functional ability for guidance, a blank video, and a parent-report checklist for additional description of their daughter’s functional skills. The structured protocol asked them to film sequences which demonstrated aspects of oromotor function, mobility, hand function and movements, personal care, and communication, broadly based on activities of daily living as conceptualized in the Functional Independence Measure for Children (WeeFIM).12 Families were asked to film these activities in the individual’s usual environment.10
The protocol for describing hand function was based on the Hand Apraxia Scale, a 10-item measure of fundamental hand skills.13 Participants were encouraged to pick up and hold a selection of large objects (toy, small ball, cup, utensil) and a small object (e.g. sultana or sweet, often demonstrated with a small piece of sandwich). For all examples of hand function viewed, the best ability demonstrated was coded to represent the skill level. Abilities to grasp, pick up, and hold the large and small objects were coded as ‘able to do independently’ or ‘not able to do independently’. Grasping a small object was coded as ‘using a raking grasp’ or ‘using the radial side of the hand involving the thumb’. A radial hand grasp included scissor, inferior pincer, and superior pincer grasps. Pre-shaping skills of the hand were observed as the hand approached each object, with hand orientation and size recognition coded as ‘close approximation’ or ‘not close approximation’. The ability to transfer an object from one hand to another was coded as ‘present’ or ‘absent’. The videos were scored by an experienced physiotherapist and a trained research assistant who were blind to MECP2 genotype; kappa scores ranging from 0.75 to 1.00 indicated substantial to excellent reliability.
Overview of the composite of videos suggested that the ability to manipulate large objects was a potential mid-point in the range of skill levels. Frequencies of combinations of skills that were more or less complex than this mid-point were determined, and levels were defined with minimal overlap (Table I).14 The level for each participant was identified. Questionnaire data were used to identify the level if video had been provided without footage of hand function.
Table I.
Description of skills observed for each level of hand function
| Level of hand function |
Skills observed |
|---|---|
| 1 | No observed hand function |
| 2 | Able to hold at least one large object (cup, spoon, small ball or toy) >2s |
| 3 | Assistance to grasp but able to pick up and hold at least one large object >2s |
| 4 | Able to grasp, pick up and hold at least one large object >2s |
| 5 | Able to grasp, pick up and hold at least one large object >2s AND use a raking grasp to grasp, pick up and hold a small object (e.g. sultana, sweet, or small piece of sandwich) >2s |
| 6 | Able to grasp, pick up and hold at least one large object >2s AND use the radial side of the hand to grasp, pick up and hold a small object >2s. Can be a scissors, inferior pincer, or superior pincer grasp |
| 7 | Skills for level 6 AND able to transfer an object from one hand to the other. Accurate pre-shaping of the hand is NOT seen |
| 8 | Skills for level 7 AND, when hand is approaching an object, hand orientation and size recognition closely approximates the position and size of the object |
Cases were categorized by MECP2 mutation, including C-terminal deletion, large deletion, early-truncating, p.R106W, p.R133C, p.T158M, p.R168X, p.R255X, p.R270X, p.R294X, and p.R306C mutations; a final group included all other pathogenic mutations. Age was grouped into four categories representing the pre-school and early school years (<8y), primary school (8y to <13y), adolescence (13y to <19y), and adult years (≥19y). The ARSD 2004 and 2006 follow-up questionnaires completed in 2004 and 2006/2007 were used as additional data sources15 to describe clinical severity (Kerr16 and WeeFIM12 scores) and frequency of hand stereotypies. The original Kerr measure comprised 19 items representing a range of clinical features such as ambulation, scoliosis, and epilepsy, each item being rated on a 3-point scale with higher scores representing greater severity.16 A modified Kerr score using 14 of the items1 was calculated for this study, and the item for hand function was omitted to avoid circularity. WeeFIM scores were used to describe function in activities of daily living, higher scores representing greater independence.12 Mobility was described on a 4-point scale with categories of ‘totally dependent on carer’, ‘able to support weight briefly for transfer activities’, ‘severely restricted walking’, and ‘mildly restricted or normal levels of walking’.
Ethical approval was provided by the ethics committee of the Princess Margaret Hospital in Western Australia. Parents gave informed consent for their children to participate in this study and for the publication of the results.
Analysis
Median and range of hand function levels were reported for each of the categorical dependent variables. Ordinal logistic regression was used to analyse univariate relationships between level of hand function and common mutations, age group, Kerr and WeeFIM scores, and mobility. We used the Stata procedure ologit (Stata Corp., College Station, TX, USA), which performs ordinal logistic regression, incorporating the proportional odds model, to estimate the effect of age and mutation type on hand function. The model outputs a common odds ratio (OR) for each predictor variable, which is interpreted as the relative odds of a participant being higher on the hand function variable across the entire range of hand function levels. Pseudo R2 values, which assess the fit of the ordinal logistic models used to predict the odds of higher levels of hand function, were assessed for each model. These measure the ratio of the log estimated likelihoods from the model with and without predictors, and have values between 0 and 1 such that higher values indicate better model fit. All analyses were undertaken using Stata 9 (Stata Corp., College Station, TX, USA).
RESULTS
Video data were provided by 144 families. All participants had a clinical diagnosis of Rett syndrome, their mean age being 14 years 10 months (SD 7y 10mo; range 2y –31y 10mo). Thirty-six participants were younger than 8 years, 25 were aged between 8 and under 13 years, 40 were aged 13 to under 19 years, and 43 were aged 19 years or older. A MECP2 mutation was identified in 110 (80.9%) of the 136 participants tested. Mutation status was unknown or not tested in eight participants. The early-truncating, p.R106W, p.R306C, and large-deletion mutations were identified in approximately 5% of participants; the C-terminal, p.R133C, p.R168X, p.R255X, p.R270X, p.R294X, p.T158M, and other mutations were identified in approximately 10% of participants (Table II).
Table II.
Level of hand function by mutation type, age, and mobility
| Participants |
|||||||
|---|---|---|---|---|---|---|---|
| Total (n=144) | With a mutation (n=110) |
Level of hand function | |||||
| n | (%) | n | (%) | Median | (range) | ||
| Mutation | C-terminal | 12 | (8.3) | 12 | (10.9) | 5 | (1–8) |
| Early-truncating | 4 | (2.8) | 4 | (3.6) | 2.5 | (1–4) | |
| p.R106W | 5 | (3.5) | 5 | (4.6) | 3 | (2–5) | |
| p.R133C | 12 | (8.3) | 12 | (10.9) | 6.5 | (1–8) | |
| p.R168X | 11 | (7.6) | 11 | (10.0) | 1 | (1–5) | |
| p.R255X | 9 | (6.2) | 9 | (8.2) | 2 | (1–8) | |
| p.R270X | 10 | (6.9) | 10 | (9.1) | 2 | (1–7) | |
| p.R294X | 12 | (8.3) | 12 | (10.9) | 6 | (1–6) | |
| p.R306C | 5 | (3.5) | 5 | (4.6) | 5 | (4–6) | |
| p.T158M | 9 | (6.2) | 9 | (8.2) | 2 | (1–6) | |
| Large deletions | 6 | (4.2) | 6 | (5.4) | 3.5 | (1–8) | |
| Other mutations | 14 | (9.7) | 14 | (12.7) | 5.5 | (1–8) | |
| No confirmed pathogenic mutation | 34 | (23.6) | NA | NA | 5 | (1–8) | |
| Age, y | <8 | 36 | (25.0) | 30 | (27.3) | 4 | (1–8) |
| 8–<13 | 25 | (17.4) | 19 | (17.3) | 5 | (1–7) | |
| 13–<19 | 40 | (27.8) | 27 | (24.5) | 6 | (1–8) | |
| ≥19 | 43 | (29.9) | 34 | (30.9) | 2 | (1–8) | |
| Level of mobilitya | Dependent on carer for all mobility | 41 | (31.8) | 32 | (32.3) | 1 4 | (1–6) |
| Able to support weight during transfers | 20 | (15.5) | 13 | (13.1) | 4 | (1–8)b | |
| Severely restricted walking ability | 29 | (22.5) | 20 | (20.2) | 5 | (1–7) | |
| Mildly or not restricted walking ability | 39 | (30.2) | 34 | (34.3) | 6 | (1–8) | |
n=129 for total participants, n=99 for participants with a mutation.
Range 1–6 for patients with mutations. NA, not applicable.
The video package was sent to the families and carers of 246 females with Rett syndrome, and 144 families returned a video on one or two occasions, giving an overall response of 58.5%. To assess the representativeness of our sample, the age group at the time of video for the 144 participants was compared with the age group mid-way through 2004 for the 153 females known to the ARSD for whom a video had not been provided, irrespective of whether families had received a video package (some families were not sent a video package because they declined to participate or because the person with Rett syndrome subsequently died). Participants for whom a video was provided were a little more likely to be younger than 8 years (25.0% vs 17.6%) and less likely to be aged 8 to less than 13 years (17.4 vs 32.0%; p=0.366). There were 110 participants (80.9% of 136 tested) with positive pathogenic mutations in the video study, and 113 (82.5% of 137 tested) with positive pathogenic mutations for whom a video was not provided. The group without a video included 21 females who had died since 1991, to account for survival bias. The distribution of types of mutation was similar for both groups (χ211=8.9, p=0.631).
Approximately one-third of participants were totally dependent on a carer for mobility, and a similar proportion was able to walk with mild or no restriction (Table II). For all participants, the mean Kerr score was 16.6 (SD 4.1, range 7.0–24.0) out of a total possible score of 28, and the mean WeeFIM score was 31.6 (SD 11.6, range 18.0–85.0) out of a total possible score of 126.
Hand function was coded from video observations in 129 participants, and questionnaire data were used to identify the level of hand function in 15 participants who did not provide footage to illustrate hand function. The number and age of participants at each level of hand function is shown in Table III. The distribution of hand function levels was bimodal, with the highest proportions for level 1 (indicating no hand function) and level 6 (indicating the ability to grasp, pick up, and hold a small object using the radial side of the hand). More specifically, nearly one-third of participants did not demonstrate hand function (n=43), a further 17% (n=25) could perform one or only a few single skills with a large object, approximately 12% could manipulate a large object (n=18), and the remainder (n=58) could manipulate a small object. Approximately 13% (n=18) could transfer an object, with 5% (n=8) demonstrating close approximation during their manipulation of small objects.
Table III.
Number of participants and severity scores for each category of hand function
| Level of hand function | Participants | Kerr scorea | WeeFIM scoreb | |||
|---|---|---|---|---|---|---|
| n | (%) | Mean | (SD) | Mean | (SD) | |
| All participants (n=144)c | ||||||
| 1 | 43 | (29.9) | 18.2 | (3.5) | 28.1 | (9.2) |
| 2 | 16 | (11.1) | 19.5 | (2.9) | 28.0 | (9.5) |
| 3 | 9 | (6.2) | 18.4 | (2.7) | 27.5 | (5.0) |
| 4 | 18 | (12.5) | 16.3 | (4.0) | 25.9 | (5.8) |
| 5 | 16 | (11.1) | 13.4 | (2.9) | 36.8 | (12.9) |
| 6 | 24 | (16.7) | 15.3 | (3.8) | 36.0 | (10.8) |
| 7 | 10 | (6.9) | 13.1 | (4.5) | 39.1 | (14.2) |
| 8 | 8 | (5.6) | 13.7 | (4.2) | 41.0 | (24.8) |
| Participants with a pathogenic MECP2 mutation (n=110)d | ||||||
| 1 | 34 | (30.9) | 18.6 | (3.5) | 27.1 | (9.3) |
| 2 | 11 | (10.0) | 19.0 | (3.1) | 29.1 | (9.8) |
| 3 | 6 | (5.5) | 18.0 | (3.0) | 27.2 | (5.8) |
| 4 | 10 | (9.1) | 16.8 | (4.6) | 28.4 | (6.2) |
| 5 | 13 | (11.8) | 13.1 | (3.0) | 35.7 | (13.4) |
| 6 | 23 | (20.9) | 15.0 | (3.6) | 36.3 | (11.0) |
| 7 | 8 | (7.3) | 11.5 | (3.6) | 43.5 | (12.8) |
| 8 | 5 | (4.5) | 12.3 | (5.1) | 49.3 | (31.1) |
Total possible Kerr score 28 (higher scores represent greater severity).
Total possible Functional Independence Measure for children (WeeFIM) score 126 (higher scores represent greater independence).
n=128 for Kerr scores, n=118 for WeeFIM scores.
n=98 for Kerr scores, n=91 for WeeFIM scores.
MECP2, methyl CpG binding protein 2 gene.
The direction of effects and size of ORs for the analyses of each dependent variable in those with a positive pathogenic mutation were similar to the findings in all participants (Tables III and IV). Findings for participants with a positive pathogenic mutation are described below.
Table IV.
Multivariate analysis of the influence of mutation, age-group, and severity on the level of hand function
| Factor | Category | All participants (n=144) | Participants with a pathogenic mutation (n=110) |
||
|---|---|---|---|---|---|
| Odds ratioa (95% CI) | p value | Odds ratioa (95% CI) | p value | ||
| Mutationb | C-terminal | Baseline | Baseline | ||
| Early-truncating | 0.24 (0.03–1.77) | 0.16 | 0.40 (0.06–2.48) | 0.32 | |
| p.R106W | 0.57 (0.11–2.93) | 0.50 | 0.59 (0.12–3.07) | 0.53 | |
| p.R133C | 2.54 (0.58–11.19) | 0.22 | 2.65 (0.57–12.26) | 0.21 | |
| p.R168X | 0.22 (0.05–1.02) | 0.05 | 0.19 (0.04–0.95) | 0.04 | |
| p.R255X | 0.37 (0.08–1.80) | 0.22 | 0.39 (0.08–1.87) | 0.24 | |
| p.R270X | 0.36 (0.08–1.64) | 0.18 | 0.34 (0.07–1.65) | 0.18 | |
| p.R294X | 1.70 (0.41–7.04) | 0.46 | 1.76 (0.41–7.57) | 0.45 | |
| p.R306C | 1.35 (025–7.18) | 0.72 | 1.24 (0.23–6.77) | 0.80 | |
| p.T158M | 0.41 (0.09–1.90) | 0.25 | 0.46 (0.10–2.18) | 0.33 | |
| Large deletions | 0.68 (0.12–3.89) | 0.67 | 0.65 (0.10–2.18) | 0.63 | |
| Other mutations | 1.04 (0.26–4.10) | 0.96 | 0.91 (0.22–3.80) | 0.90 | |
| No confirmed pathogenic mutation | 0.58 (0.17–2.03) | 0.40 | NA | ||
| Age,c, y | <8 | Baseline | Baseline | ||
| 8–<13 | 0.70 (0.26–1.87) | 0.48 | 0.90 (0.30–2.68) | 0.86 | |
| 13–<19 | 1.22 (0.52–2.82) | 0.65 | 1.62 (0.62–4.20) | 0.33 | |
| ≥19 | 0.44 (0.19–1.03) | 0.06 | 0.36 (0.14–0.92) | 0.03 | |
| Severity score d | Kerre | 0.79 (0.71–0.87) | <0.001 | 0.77 (0.69–0.86) | <0.001 |
| WeeFIMf | 1.06 (1.03–1.10) | <0.001 | 1.08 (1.04–1.12) | <0.001 | |
| Level of mobilityg | Dependent on carer for all mobility | Baseline | Baseline | ||
| Able to support weight during transfers | 5.62 (1.82–17.29) | 0.003 | 6.82 (1.94–23.97) | 0.003 | |
| Severely restricted walking ability | 8.89 (3.09–25.59) | <0.001 | 11.63 (3.35–40.38) | <0.001 | |
| Mildly or not restricted walking ability | 15.06 (5.15–44.04) | <0.001 | 22.64 (7.02–73.08) | <0.001 | |
Odds ratio refers to a unit increase in the predicting variable.
Controlling for age.
Controlling for mutation.
Controlling for age and mutation.
Total n=122; participants with a mutation n=98.
Total n=113, participants with a mutation n=91.
Controlling for age and mutation; total n=122, participants with a mutation n=99.
CI, confidence interval; NA, not applicable; WeeFIM, Functional Independence Measure for Children.
Participants with a p.R168X mutation had the lowest median level of hand function (level 1), and those with a p.R255X, p.R270X, or T158M mutation had a median level of 2. Participants with a p.R294X or p.R133C mutation had the highest median levels of hand function (level 6 and 6.5; Table II; Fig. 1). The hand function of those with a p.R168X mutation was on average two levels lower than those with a C-terminal mutation, which was used for baseline comparison, this difference persisting after controlling for age group (adjusted OR 0.19, 95% confidence interval [CI] 0.04–0.95; Table IV). Participants who were 19 years or older had lower levels of hand function than those younger than 8 years, with a decrease of approximately two levels of hand function between the age groups after controlling for pathogenic mutation (adjusted OR 0.36, 95% CI 0.14–0.92; Table IV).
Figure 1.
Levels of hand function associated with each of the common mutations (n=110). ET, early-truncating mutation; LD, large deletions; CT, Cterminal mutation.
The Kerr score decreased and WeeFIM score increased as hand function improved for all participants and for those with a pathogenic mutation (Table III). Higher Kerr scores were associated with poorer hand function levels (OR 0.77, 95% CI 0.70–0.85), and this effect persisted after controlling for age group and mutation (adjusted OR 0.77, 95% CI 0.69–0.86; Table IV). Higher WeeFIM scores were associated with better hand function levels (OR 1.08, 95% CI 1.05–1.12), although the effect size was small, and this persisted after controlling for age group and mutation (adjusted OR 1.08, 95% CI 1.04–1.12; Table IV).
Participants with the lowest level of mobility had a median hand function level of 1 (range 1–6), compared with 6 (range 1–8) for those with the greatest level of mobility (Table II), with higher ORs for each successive increase in mobility level (range of ORs 4.82–18.04). After controlling for age group and mutation, the strong relationships between mobility and hand function persisted (range of adjusted ORs 6.82–22.64; Table IV). In all models, the pseudo R2 was between 0.05 and 0.15. Therefore, the values (which are always smaller than a conventional R2 value) were considered appropriate for the descriptive purposes of these models.
DISCUSSION
Using a structured video assessment tool in naturalistic settings, eight levels of hand function were described in a representative population of females with Rett syndrome. Genotype, age, and clinical severity were related to the level of hand function, with poorer function in those with the p.R168X mutation, those with greater clinical severity, and those aged 19 years or older with mutations.
Participants with the p.R168X mutation had the poorest hand function, whereas those with the p.R133C or p.R294X mutation had generally better hand function. These findings were based on detailed assessment of current hand function in a sample that was representative of the Australian population with Rett syndrome, and such population-based information is unique in the literature. Using data from the international database InterRett, the p.R168X and p.R270X mutations were also found to be associated with the lowest levels of hand function and the p.R133C and p.R294X mutation were associated with better levels.6 In that study, hand function was assessed with five much broader categories (never acquired hand skills; lost all skills; maintained grasping; maintained manipulation; and maintained all skills). In a large US clinic-based study, p.R168X was found to be the most severe mutation overall.5 Hand skills were also measured on a 5-point scale representing similar types of categories to those used in the international study6 (i.e. never acquired hand skills; acquired but lost skills; acquired late, partially conserved; acquired on time, partially conserved; and conserved). Neither of those studies was population-based. However, the composite of using different samples and methodologies has identified a consistent pattern of association between MECP2 mutations and hand function, supporting the construct validity of the new scale. What our current study adds is a new measure with greater increments and one which assesses current hand function independently from developmental history. Therefore, this measure is more likely to be sensitive to smaller amounts of change when evaluating clinical interventions.
Among participants with a pathogenic mutation, hand function was poorest in those aged 19 years and over, consistent with findings that functional skills in general deteriorate with age.3,7 We acknowledge that additional longitudinal study is required, but, for our sample, decline of hand function after the initial regression period appeared to have been mild. General gross motor abilities, such as standing, walking, and transfer skills, also decline with age.9 In contrast to hand function, these gross motor changes are highly visible and can have consequences such as dependence on a wheelchair for mobility. The mechanism of the decline in motor skills with age is not yet clear but probably reflects genetic determinants.5,6 Environmental factors could also play an important role in maintaining function, as has been illustrated in animal studies.17 An enriched, sustained, motivating environment has the potential to be a modifier of hand function in Rett syndrome.
As might be expected, milder symptoms of clinical severity, greater independence in activities of daily living, and better levels of mobility were associated with better hand function, even after controlling for genotype and age. The Kerr measure of clinical severity16 encompassed many of the features of Rett syndrome,3 but we excluded the hand function item to avoid circularity during analysis. We confirmed limited variability in daily living skills due to a floor effect and the previously observed variability in weight-bearing skills.18 Therefore, the severity of other clinical domains was reflected in current hand function levels, also supporting the construct validity of the hand function scale.
We have described a more sensitive measure of hand function in Rett syndrome than ever previously available. The participants were videotaped in familiar settings while presented with a simple series of tasks, and professional assessment of the level of hand function was a straightforward procedure. Skills were based on the Hand Apraxia Scale,13 and the eight levels of hand function were sequential in complexity, consistent with the early development of hand skills.14 We acknowledge the difficulty of measuring hand function in children and adults with low cognitive function but believe that these similarities support the content validity of the scale. Abilities to use the basic skills of grasping and holding a large object, which were captured in levels 2 and 3, were performed with apparent satisfaction by many severely affected participants. At the other end of the scale, those who could manipulate and transfer a small object could be graded as level 7 or 8. Previously, hand function in Rett syndrome has been measured by rating agreement with a broad statement on a Likert scale. For example, hand clumsiness is coded on a 5-point scale in the Rett Syndrome Motor-Behavioural Assessment.19 The hand function items in previously reported global severity scales have also included broad categories and have not been limited to current status (i.e. they have also included information about regression).1 Thus, the hand function items in these scales contribute to an understanding of larger concepts such as behaviour and general severity in Rett syndrome. In contrast, our 8-point scale provides greater characterization of hand function and focuses on current function, which may be amenable to change with therapeutic interventions.
Other scales of early hand function skills can be difficult to apply to Rett syndrome. The Erhardt Developmental Prehension Assessment20 and the Quality of Upper Extremity Skills Test21 were developed for children with cerebral palsy, and the Peabody Developmental Fine Motor Scale22 for developmental delay. One example of a 2 to 3-month item in the Peabody scale with uncertain validity for those with profound intellectual disability requires the infant to hold and shake a rattle for 30 seconds, a task that needs cognitive engagement as well as hand function skills. Our scale relates specifically to the variability seen in our sample with Rett syndrome, and testing was feasible within the context of the severe level of intellectual disability.
Our observations of naturalistic function within familiar settings provide ecological validity to the findings, and we have provided some evidence to support reliability and construct validity of the scale. Additionally, our sample was representative of the participants known to the ARSD in terms of age and genotype, thus reducing the likelihood of selection bias, which may be a shortcoming of the non-population-based clinic samples often used in Rett syndrome research. As these factors are important, the similarities are gratifying, but there may be other less measurable factors such as family proactivity that differentiate our sample from the total population. Therefore, we acknowledge that there could be a bias (although difficult to quantify) towards families who are working more energetically with their children, and in this way we could be presenting a slightly more optimistic picture of these children responding positively to their environment. Future researchers should certainly make a point of measuring intensity and duration of any interventions that are being undertaken. Finally, we acknowledge that, unlike other studies that we have undertaken to investigate the relationship between genotype and clinical severity in Rett syndrome, in which deceased individuals with the syndrome were not excluded, a study of this nature is limited to living individuals and for this reason the issue of survival bias has to be considered. We know that our previous Australian population-based study has provided a slightly different pattern of severity from the US clinic-based studies.5 We attribute that to the survival bias associated with the latter. Interestingly, the p.R168X mutation, which was the most severe in the clinic-based study, was also the one associated with the poorest hand function in this study. This would be in keeping with the effect that we would hypothesise that surviving participants with p.R270X (the mutation that we found to be most severe and associated with increased mortality)23 may not be totally representative of this mutation and hence not necessarily have the most severe hand function.
It is important to recognize the burden on families in providing the video, and we acknowledge their efforts that have allowed us to develop an assessment scale for potential use in clinical and research settings. Although the scale can be used in clinical settings to rate direct observations, families could provide video taken in familiar settings to their clinicians on an ongoing basis as a part of their working relationship. This process would now be more feasible and considerably less burdensome than in 2004 and 2007 because the technology available to create and send video clips is evolving rapidly and becoming conventional.
Finally, this study was cross-sectional and we recommend longitudinal analysis of hand function, the subject of ongoing research in our group. Nevertheless, we believe that further assessment of the hand function scale is indicated, and this promising assessment could have considerable applicability in clinical and research settings.
WHAT THIS PAPER ADDS.
We have developed a measure of hand function that is specific to Rett syndrome.
Using this measure, we have documented previously unknown variability in hand function.
The level of hand function relates to age, mutation type, and disease severity.
This novel measure has a potential role for use in clinical trials.
ACKNOWLEDGEMENTS
We acknowledge the work of Carol Philippe from the Telethon Institute for Child Health Research in Perth who was responsible for much of the video data collection. We also acknowledge the molecular work of Linda Weaving and Sarah Williamson (under the guidance of Professor John Christodoulou and Dr Bruce Bennetts) in Sydney and Mark Davis in Perth. We express our sincere gratitude to all of the families who have contributed to the study and the Australian Paediatric Surveillance Unit (APSU) and the Rett Syndrome Association of Australia who facilitated case ascertainment in Australia. The APSU is a unit of the Division of Paediatrics, Royal Australasian College of Physicians, and is funded by the Department of Health and Ageing and the Faculty of Medicine of the University of Sydney. The video component of the Australian Rett Syndrome programme was funded by the National Medical and Health Research Council (NHMRC) under project grant 303189, and major aspects of the research programme were funded by the National Institutes of Health (1 R01 HD43100-01A1). Helen Leonard was previously funded by NHMRC grant (no. 353514) and currently by an NHMRC Senior Research Fellowship (572568). Walter Kaufmann is funded by NIH grant P01 HD24448.
LIST OF ABBREVIATIONS
- ARSD
Australian Rett Syndrome Database
- MECP2
methyl CpG binding protein 2 gene
- WeeFIM
Functional Independence Measure for children
REFERENCES
- 1.Colvin L, Fyfe S, Leonard S, et al. Describing the phenotype in Rett syndrome using a population database. Arch Dis Child. 2003;88:38–43. doi: 10.1136/adc.88.1.38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Amir RE, Van den Veyver IB, Wan M, Tran CQ, Francke U, Zoghbi HY. Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat Genet. 1999;23:185–188. doi: 10.1038/13810. [DOI] [PubMed] [Google Scholar]
- 3.Hagberg B, Hanefeld F, Percy A, Skjeldal O. An update on clinically applicable diagnostic criteria in Rett syndrome. Comments to Rett Syndrome Clinical Criteria Consensus Panel Satellite to European Paediatric Neurology Society Meeting, Baden Baden, Germany, 11 September 2001. Eur J Pediatr Neurol. 2002;6:293–297. doi: 10.1053/ejpn.2002.0612. [DOI] [PubMed] [Google Scholar]
- 4.Mount RH, Charman T, Hastings RP, Reilly S, Cass H. The Rett Syndrome Behaviour Questionnaire (RSBQ): refining the behavioural phenotype of Rett syndrome. J Child Psychol Psychiatry. 2002;43:1099–1110. doi: 10.1111/1469-7610.00236. [DOI] [PubMed] [Google Scholar]
- 5.Neul JL, Fang P, Barrish J, et al. Specific mutations in methyl-CpG-binding protein 2 confer different severity in Rett syndrome. Neurology. 2008;70:1313–1321. doi: 10.1212/01.wnl.0000291011.54508.aa. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bebbington A, Anderson A, Ravine D, et al. Investigating genotype–phenotype relationships in Rett syndrome using an international dataset. Neurology. 2008;70:868–875. doi: 10.1212/01.wnl.0000304752.50773.ec. [DOI] [PubMed] [Google Scholar]
- 7.Hagberg B. Clinical manifestations and stages of Rett syndrome. Ment Retard Dev Disabil Res Rev. 2002;8:61–65. doi: 10.1002/mrdd.10020. [DOI] [PubMed] [Google Scholar]
- 8.Larsson G, Lindström B, Engerström IW. Rett syndrome from a family perspective: The Swedish Rett Center survey. Brain Dev. 2005;27(Suppl 1):S14–S19. doi: 10.1016/j.braindev.2005.03.015. [DOI] [PubMed] [Google Scholar]
- 9.Cass H, Reilly S, Owen L, et al. Findings from a multidisciplinary clinical case series of females with Rett syndrome. Dev Med Child Neurol. 2003;45:325–337. doi: 10.1017/s0012162203000616. [DOI] [PubMed] [Google Scholar]
- 10.Fyfe S, Downs J, McIlroy O, et al. Development of a video-based evaluation tool in Rett syndrome. J Autism Dev Disord. 2007;37:1636–1646. doi: 10.1007/s10803-006-0293-9. [DOI] [PubMed] [Google Scholar]
- 11.Leonard H, Bower C. Is the girl with Rett syndrome normal at birth? Dev Med Child Neurol. 1998;40:115–121. [PubMed] [Google Scholar]
- 12.Msall ME, DiGaudio K, Rogers BT, et al. The Functional Independence Measure for Children (WeeFIM). Conceptual basis and pilot use in children with developmental disabilities. Clin Pediatr (Phila) 1994;33:421–430. doi: 10.1177/000992289403300708. [DOI] [PubMed] [Google Scholar]
- 13.Burd L, Cook J, Randall T. The Hand Apraxia Scale. Percept Mot Skills. 1990;70:219–224. doi: 10.2466/pms.1990.70.1.219. [DOI] [PubMed] [Google Scholar]
- 14.Schneck CM, Henderson A. Descriptive analysis of the developmental progression of grip position for pencil and crayon control in nondysfunctional children. Am J Occup Ther. 1990;44:893–900. doi: 10.5014/ajot.44.10.893. [DOI] [PubMed] [Google Scholar]
- 15.Jian L, Nagarajan L, de Klerk N, et al. Predictors of seizure onset in Rett syndrome. J Pediatr. 2006;149:542–547. doi: 10.1016/j.jpeds.2006.06.015. [DOI] [PubMed] [Google Scholar]
- 16.Kerr AM, Nomura Y, Armstrong D, et al. Guidelines for reporting clinical features in cases with MECP2 mutations. Brain Dev. 2001;23:208–211. doi: 10.1016/s0387-7604(01)00193-0. [DOI] [PubMed] [Google Scholar]
- 17.Kondo M, Gray LJ, Pelka GJ, Christodoulou J, Tam PP, Hannan AJ. Environmental enrichment ameliorates a motor coordination deficit in a mouse model of Rett syndrome - Mecp2 gene dosage effects and BDNF expression. Eur J Neurosci. 2008;27:3342–3350. doi: 10.1111/j.1460-9568.2008.06305.x. [DOI] [PubMed] [Google Scholar]
- 18.Downs JA, Bebbington A, Jacoby P, et al. Gross motor profile in Rett syndrome as determined by video analysis. Neuropediatrics. 2008;39:205–210. doi: 10.1055/s-0028-1104575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.FitzGerald PM, Jankovic J, Percy AK. Rett syndrome and associated movement disorders. Mov Disord. 1990;5:195–202. doi: 10.1002/mds.870050303. [DOI] [PubMed] [Google Scholar]
- 20.Erhardt RP, Beatty PA, Hertsgaard DM. A developmental prehension assessment for handicapped children. Am J Occup Ther. 1981;35:237–242. doi: 10.5014/ajot.35.4.237. [DOI] [PubMed] [Google Scholar]
- 21.DeMatteo C, Law M, Russell D, Pollock N, Rosenbaum WS. The reliability and validity of the Quality of Upper Extremity Skills Test. Phys Occup Ther Pediatr. 1993;13:1–18. [Google Scholar]
- 22.Folio MR, Fewell RR. Peabody developmental motor scales and activity cards. Allen, TX, USA: DLM Teaching Resources; 1983. [Google Scholar]
- 23.Jian L, Archer HL, Ravine D, et al. p.R270X MECP2 mutation and mortality in Rett syndrome. Eur J Hum Genet. 2005;13:1235–1238. doi: 10.1038/sj.ejhg.5201479. [DOI] [PubMed] [Google Scholar]