Prodromal Markers of Upper Limb Deficits in FMR1 Premutation Carriers and Quantitative Outcome Measures for Future Clinical Trials in Fragile X‐associated Tremor/Ataxia Syndrome

ABSTRACT

Background

Fragile X‐associated Tremor/Ataxia Syndrome (FXTAS) is a rare, late‐onset neurodegenerative disorder characterized by tremor and cerebellar gait ataxia, affecting premutation carriers (PMC) of CGG expansions (range, 55–200) in the fragile X mental retardation 1 (FMR1) gene. Discovery of early predictors for FXTAS and quantitative characterization of motor deficits are critical for identifying disease onset, monitoring disease progression, and determining efficacy of interventions.

Methods

A total of 39 PMC with FXTAS, 20 PMC without FXTAS, and 27 healthy controls performed a series of upper extremity (UE) motor tasks assessing tremor, bradykinesia, and rapid alternating movements that were quantified using an inertial‐based sensor system (Kinesia One; Great Lakes NeuroTechnologies, Cleveland, OH, USA). Sub‐scores from the clinician‐rated FXTAS Rating Scale were correlated with the severity scores generated by the sensor system to determine its validity in FXTAS.

Results

PMC with FXTAS had significantly worse postural and kinetic tremor compared with PMC without FXTAS (P = 0.02, 0.03) and controls (P = 0.001, 0.0001), respectively, and slower finger tap (P = 0.001), hand movement (P = 0.0001), and rapid alternating movement speed (P = 0.003) and amplitude (P = 0.04) than controls. PMC without FXTAS had significantly worse right finger tap (P = 0.004), hand movement (P = 0.01), and rapid alternating movement speed (P = 0.003) and amplitude (P = 0.02) than controls. FXTAS Rating Scale subscores significantly correlated with all tremorography scores except for finger taps and left rapid alternating movement.

Conclusions

These findings support the use of inertial sensor quantification systems as promising measures for preclinical FXTAS symptom detection in PMC, characterization of the natural history of FXTAS, assessment of medication responses, and outcome assessment in clinical trials.

Fragile X‐associated tremor/ataxia syndrome (FXTAS) is a late‐onset, progressive neurodegenerative disorder affecting approximately 40% to 70% of men and 16% to 20% of women carrying a “premutation” range allele of 55 to 200 CGG repeats in the fragile X mental retardation 1 (FMR1) gene located on the X chromosome. ¹ , ² , ³ , ⁴ The mechanisms underlying the neurodegeneration of FXTAS are thought to be attributed to a toxic gain of function as a result of (1) increased FMR1 mRNA transcription and sequestration of RNA binding proteins contributing to protein aggregation and inclusion formation in neurons and astrocytes and (2) CGG repeat‐associated non‐AUG translation, which generates toxic polyglycine peptides also found in the inclusions in neurons. ⁵ , ⁶ , ⁷ Mitochondrial dysfunction and antisense transcripts to FMR1 may also contribute to the pathogenesis of FXTAS. ⁷ , ⁸ , ⁹ Clinical features include kinetic tremor, cerebellar gait ataxia, psychiatric issues, and executive dysfunction/cognitive decline in premutation carriers (PMC) with onset generally beginning after 50 years of age. ¹⁰ , ¹¹ Kinetic tremor and cerebellar ataxia are the core motor features of the disease, although parkinsonism may be present in a significant number of individuals with FXTAS. ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ To date, there are no disease‐modifying therapies available for FXTAS, which causes significant disability and increased morbidity and mortality. ¹⁹ However, as treatments become available, early symptom detection will be critical for earlier intervention and timely management.

Prodromal symptoms of FXTAS have been previously reported in PMC without FXTAS including manual slowing of movement and prolonged reaction times, ²⁰ deficits in postural response latencies and the sensorimotor control of balance, ²¹ increased gait variability, ²² and reduced stride length during gait ²³ and delayed step reaction times ²⁴ when dual tasking.

The Kinesia One inertial sensor‐based system (Great Lakes NeuroTechnologies, Cleveland, OH, USA) has been shown to be a valid and reliable method of measuring UE tremor and bradykinesia in Parkinson's disease (PD) and essential tremor. Our group recently demonstrated its ability to distinguish the tremor and bradykinesia profiles between patients with FXTAS, PD, and essential tremor. ²⁵ Therefore, the primary objectives of this study were to use the Kinesia Oneinertial sensor system to (1) characterize the severity of tremor subtypes and bradykinesia in PMC with FXTAS and (2) determine whether this technology can detect early signs of motor dysfunction in PMC without FXTAS.

Methods

Study Participants

PMC participants with and without FXTAS were recruited from the FXTAS and Fragile X Clinics at Rush University Medical Center and from recruitment advertisements from the National Fragile X Foundation. Inclusion criteria for FXTAS participants were (1) FMR1 CGG repeat size of 55 to 200; (2) a diagnosis of possible, probable, or definite FXTAS according to well‐established clinical and radiological criteria ¹ , ¹³ ; and (3) mild to severe tremor. Inclusion criteria for PMC participants without FXTAS were (1) FMR1 CGG repeat size of 55 to 200 and (2) a diagnosis of no FXTAS made by a movement disorder specialist (D.A.H.) after administration of a neurological examination. None of the PMC participants without FXTAS had magnetic resonance imaging. Exclusion criteria were (1) a history of stroke or traumatic brain injury with focal neurological deficit or any other neurological or muscular disease, (2) a significant seizure disorder, or (3) clinical diagnosis of dementia. Healthy control participants were recruited from Rush University Medical Center or the community or were friends or family members of the enrolled PMC participants. Inclusion criteria for controls were (1) a normal neurological examination and (2) <55 CGG repeats on both FMR1 alleles. Exclusion criteria were (1) a history of tremor, balance problems, falls, or dizziness in addition to the same exclusion criteria as the PMC participants. All participants signed an informed consent that was approved by the institutional review board at Rush University Medical Center (no. 15030606).

Tremorography

The ETSense (Great Lakes NeuroTechnologies, Cleveland, OH, USA) inertial sensor with the Kinesia One systemwas used to assess tremor and bradykinesia in both upper extremities (UEs). The inertial sensor has the capacity to capture motion in 6 degrees of freedom via the combination of three accelerometers and 3 gyroscopes. The system has been validated for the quantitative measurement of kinetic, postural, and rest tremor and bradykinesia of the UEs in PD and essential tremor. ²⁶ Scores from this system demonstrated a high correlation with those of the clinician‐rated Unified Parkinson's Disease Rating Scale and the Modified Bradykinesia Rating Scale ²⁶ , ²⁷ and have proved useful in monitoring medication efficacy in PD. ²⁸

If medicated for their movement disorder symptoms, all participants with FXTAS were tested (whenever possible) in their optimal medicated state. The sensor was strapped around the participant's index finger while performing 6 different postures or motor tasks: (1) arms resting on thighs (rest tremor), (2) arms extended to 90° (postural tremor), (3) finger to nose (kinetic tremor), (4) finger taps of index finger to thumb (bradykinesia), (5) hand grasps via hand opening and closing (bradykinesia), and (6) rapid alternating pronation and supination with elbows extended (dysdiadochokinesia and bradykinesia). Each trial was performed bilaterally for 8 seconds. Motion data were wirelessly transferred in real time to the Kinesia iPad and processed into quantitative variables for tremor, bradykinesia, and dysdiadochokinesia. Severity scores of 0 to 4 for each measure were generated by algorithms developed and validated by the manufacturer with 0.1 resolution, with 0 indicating no symptoms and 4 indicating the most severe symptoms. ²⁶ The severity scores for rest, postural, and kinetic tremor represent tremor amplitude, not frequency.

FXTAS Rating Scale

The FXTAS Rating Scale (FXTAS‐RS) was administered to all participants and scored by a movement disorder specialist (D.A.H. or G.P.). The rating scale is composed of 44 different rating items from the Unified Parkinson Disease Rating Scale, the International Cooperative Ataxia Rating Scale, and the Unified Huntington's Disease Rating Scale and assesses the severity of tremor, postural sway, gait deficits, parkinsonism, dystonia, speech, and oculomotor deficits. ²⁹ , ³⁰ Seven items from the FXTAS‐RS were extracted for this study to compare with the corresponding Kinesia‐generated variables. These were kinetic (ie, intention), postural and rest tremor and finger tap, hand movement, and rapid alternating movement (RAM) speed and amplitude. The amplitude and speed variables from Kinesia outputs were first averaged before correlating with the corresponding subscores of the FXTAS‐RS for finger tap, hand movement, and RAM variables because the FXTAS‐RS does not score amplitude and speed separately.

Molecular Analysis

Buccal swabs or peripheral blood leukocytes of all participants were obtained, and DNA were extracted using a Qiagen (Hilden, Germany) blood and tissue DNA isolation kit at the Rush University Molecular Diagnostic Laboratory. A commercially available internally primed Asuragen AmplideX FMR1 polymerase chain reaction kit (Asuragen Inc., Austin, TX) was used to determine allele‐specific CGG repeat length, and FMR1 activation ratios were determined in women using the Asuragen Amplidex FMR1 multiplex polymerase chain reaction kit (Asuragen Inc.) as previously described. ³¹ , ³²

Statistical Analysis

Demographic variables were compared between the 3 groups using a 1‐way analysis of variance followed by the Bonferroni post hoc multiple comparisons test (for parametric, normally distributed variables) or the Kruskal‐Wallis test (for nonparametric, nonnormally distributed variables). Paired t tests and signed‐rank tests were first performed to compare the right and left hand tremorography measures within groups. When no significant differences were found, then the mean (combined value) between the right and left UE were used for subsequent analyses. A univariate analysis was then performed to examine group differences in tremorography scores using a 1‐way analysis of variance followed by the Bonferroni post hoc multiple comparisons test (for normally distributed variables) or the Kruskal‐Wallis test followed by Dwass, Steel, and Critchlow‐Fligner multiple comparison tests (for nonnormally distributed variables). Scores that were found to be significantly different between the 3 groups were then entered into a multivariate regression analyses controlling for age and sex with group as the main predictor variable to assess differences in tremorography measures between controls, PMC without FXTAS, and PMC with FXTAS. Another regression analysis was performed to determine the association between CGG repeat size, age, and sex on tremorography measures in PMC participants with and without FXTAS. Interactions between the 2 groups and CGG repeat size were also examined, and those having significant interactions on the tremorography outcome variables were then entered into a separate regression analysis for the 2 groups with CGG repeat size, age, and sex as the main predictor variables. The assumptions of multiple regression (ie, independence of residuals, normality of residuals, linearity, and homoscedasticity) were met for these analyses. Spearman correlations were performed between activation ratio and tremorography variables in women PMC with and without FXTAS. Correlations between FXTAS‐RS subscores and the corresponding items on tremorography‐derived scores were also assessed by the Spearman rank correlation coefficient. The significance level was set at P < 0.05. SAS version 9.3 was used for the statistical analyses.

Results

Participant Characteristics

A total of 39 PMC with FXTAS, 20 PMC without FXTAS, and 27 controls were enrolled in this study. Of these, 25 participants with FXTAS and 20 controls were also enrolled in a separate study previously conducted by our group. ²⁵ Demographic and clinical characteristics of study participants are summarized in Table 1. The PMC without FXTAS group was significantly younger than the control and FXTAS groups (P ≤ 0.001) and had more woman participants than men (P = 0.001). As expected, the FXTAS group scored significantly higher than both PMC without FXTAS and controls on the FXTAS‐RS (P ≤ 0.05). Medications that participants with FXTAS were taking for their tremor symptoms are reported in Table 2.

TABLE 1.

Participant demographic and clinical characteristics

Variable	Controls, n = 27	PMC Without FXTAS, n = 20	PMC with FXTAS, n = 38
Age, mean ± SD	65.37 ± 9.07	53.0 ± 9.04a***,b**	68.05 ± 8.9
Sex	12 men, 15 women (44% men)	2 men, 18 women (10% men)	23 men, 16 women (59% men)
FXTAS diagnosis (%, n)
Possible			13%, 5
Probable			45%, 17
Definite			42%, 16
FMR1 CGG repeat size, mean ± SD	32.36 ± 5.3	96 ± 25.5	86.69 ± 17.12
FXTAS Rating Scale, mean ± SD
Total scores, mean ± SD	12.5 ± 7.16	6.2 ± 4.8	37.91 ± 25.65a*,c*
Disease duration, mean ± SD	N/A	N/A	7.7 ± 5.2 years
FMR1 activation ratio, mean ± SD	N/A	46.15 ± 24.9	44.6 ± 17.0

^aSignificantly different from controls.

^bSignificantly different from FXTAS.

^cSignificantly different from PMC without FXTAS.

*P < 0.05; **P < 0.01; ***P < 0.001.

Abbreviations: PMC, premutation carriers; FXTAS, fragile X‐associated tremor/ataxia syndrome; SD, standard deviation.

TABLE 2.

List of the tremor medications participants with FXTAS were taking at the time of testing

Medication	PMC with FXTAS, n (%)
Carbidopa–levodopa	2 (5)
Clonazepam	2 (5)
Propranolol	8 (20)
Metoprolol	2 (5)
Amantadine	7 (18)
Topamax	1 (3)
Not on medication	22 (56)

Abbreviations: PMC, premutation carriers; FXTAS, fragile X‐associated tremor/ataxia syndrome.

Tremorography

There were no significant differences between right and left hand severity scores for any tremorography measures except for finger tap speed and rhythm in the control group (P = 0.04), finger tap speed in the FXTAS group (P = 0.04), and postural tremor in PMC with FXTAS (P = 0.008) and PMC without FXTAS (P = 0.04). Participants with FXTAS had significantly worse combined kinetic (P = 0.0001) and right and left postural tremors (right, P = 0.01; left, P = 0.001), right and left finger tap speeds (right, P = 0.001; left, P = 0.006;), combined hand movement speed (P = 0.0001), and combined RAM speed (P = 0.003) and amplitude (P = 0.04) than control participants (Table 3; Fig. 1). Combined kinetic tremor (P = 0.03) and left postural tremor (P = 0.04) were significantly more severe in participants with FXTAS than PMC without FXTAS. PMC without FXTAS had significantly worse bradykinesia for right finger tap speed (P = 0.004), combined hand movement speed (P = 0.01), and both combined RAM speed (P = 0.003) and amplitude (P = 0.02) than control participants, and these severity scores were not different from participants with FXTAS (Table 3; Fig. 1).

TABLE 3.

Sex‐adjusted and age‐adjusted regression analysis comparing the mean group differences in tremorography severity scores between control, PMC without FXTAS, and FXTAS groups

	Tremorography Severity Scores: Mean (SD) or Median (IQR)
Variable	Controls, n = 27		PMC Without FXTAS, n = 20			PMC with FXTAS, n = 39
Combined kinetic tremor	0.95 (0.32)		0.97 (0.24)			1.32 (0.41)
Right postural tremor	0 (0.04)		0 (0.05)			0.1 (0.51)
Left postural tremor	0 (0.04)		0 (0)			0.21 (0.84)
Combined rest tremor*	0.05 (0.09)		0.03 (0.04)			0.08 (0.17)
Right finger taps speed	0.75 (0.55)		1.11 (0.05)			1.25 (0.59)
Left finger taps speed	0.94 (0.66)		1.0 (0.49)			1.49 (0.88)
Combined HMS	1.16 (0.38)		1.39 (0.37)			1.59 (0.49)
Combined RAMS	1.22 (0.48)		1.53 (0.37)			1.57 (0.42)
Combined finger taps amplitude*	1.34 (0.9)		1.23 (0.8)			1.62 (1.04)
Combined HMA*	0.65 (0.95)		0.47 (0.72)			0.95 (1.03)
Combined RAMA	0.33 (0.33)		0.43 (0.26)			0.55 (0.41)
Right finger taps rhythm*	0.44 (0.32)		0.61 (0.4)			0.64 (0.46)
Left finger taps rhythm*	0.66 (0.52)		0.56 (0.42)			0.89 (0.84)
Combined HMR*	0.39 (0.26)		0.38 (0.31)			0.42 (0.44)
Combined RAMR*	0.17 (0.24)		0.26 (0.4)			0.25 (0.5)

	Regression Analysis
	FXTAS vs Controls	PMC Without FXTAS vs Controls			FXTAS vs PMC Without FXTAS
Variable	β (SE)	P Value	β (SE)	P Value	β (SE)	P Value
Combined kinetic tremor	0.35 (0.09)	0.0001	0.09 (0.11)	0.42	0.26 (0.11)	0.03
Right postural tremor	0.3 (0.12)	0.01	0.09 (0.16)	0.55	0.21 (0.16)	0.18
Left postural tremor	0.49 (0.14)	0.001	0.05 (0.18)	0.79	0.44 (0.18)	0.02
Right finger taps speed	0.47 (0.14)	0.001	0.53 (0.18)	0.004	−0.06 (0.18)	0.75
Left finger taps speed	0.5 (0.18)	0.006	0.38 (0.23)	0.10	0.11 (0.23)	0.62
Combined HMS	0.41 (0.1)	0.0001	0.37 (0.13)	0.01	0.04 (0.13)	0.78
Combined RAMS	0.32 (0.11)	0.003	0.43 (0.14)	0.003	−0.11 (0.14)	0.45
Combined RAMA	0.18 (0.08)	0.04	0.25 (0.11)	0.02	−0.07 (0.11)	0.51

	Age			Sex
Variable	β (SE)	P Value		β (SE)	P Value
Combined kinetic tremor	0.003 (0.004)	0.53		−0.14 (0.08)	0.09
Right postural tremor	0.004 (0.006)	0.53		−0.14 (0.11)	0.21
Left postural tremor	−0.0005 (0.007)	0.95		−0.23 (0.13)	0.09
Right finger taps speed	0.02 (0.007)	0.006		0.16 (0.13)	0.21
Left finger taps speed	0.03 (0.008)	0.001		0.15 (0.16)	0.35
Combined HMS	0.02 (0.005)	0.001		0.19 (0.1)	0.049
Combined RAMS	0.01 (0.005)	0.04		0.01 (0.1)	0.89
Combined RAMA	0.007 (0.004)	0.09		−0.24 (0.08)	0.003

Tremorography scores are reported as mean (SD) for normally distributed variables and in italics as median (IQR) for nonnormally distributed variables. Statistically significant values are bold.

^*No significant differences in the univariate analysis between groups; therefore, these variables were not entered into the subsequent regression analysis.

Abbreviations: PMC, premutation carriers; FXTAS, fragile X‐associated tremor/ataxia syndrome; SD, standard deviation; IQR, interquartile; HMS, hand movements speed; RAMS, rapid alternating movement speed; HMA, hand movements amplitude; RAMA, rapid alternating movement amplitude; HMR, hand movements rhythm; RAMR, rapid alternating movement rhythm; SE, standard error.

FIG. 1.

Tremorography differences between premutation carriers with FXTAS, premutation carriers without FXTAS, and controls. FXTAS, fragile X‐associated tremor/ataxia syndrome; RAMS, rapid alternating movement speed; RAMA, rapid alternating movement amplitude. *P < 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001.

Higher CGG repeat size was significantly associated with worse combined kinetic tremor (P = 0.001) and right and left postural tremors (P = 0.0006; P = 0.0003) in PMC participants with FXTAS (Table 4; Fig. 2). Older age was significantly associated with worse kinetic tremor (P = 0.001) in participants with FXTAS. Men with FXTAS had significantly greater kinetic tremor severity scores than women (men, 1.45 ± 0.41; women, 1.14 ± 0.27; P = 0.045). In the PMC group as a whole, advancing age was associated with significantly worse combined kinetic tremor (P = 0.02) and slower right and left finger tap speeds (P = 0.048 and P = 0.005, respectively), and women had significantly slower right finger tap (P = 0.047) and combined hand movement speed (P = 0.049) than men (Table 4). No significant association between CGG repeat size, age, or sex and tremorography severity scores was found in the PMC without FXTAS group alone. Activation ratio did not correlate with any tremorography variables in the women PMC group as a whole.

TABLE 4.

Regression analysis to determine the association between CGG repeat size, age and sex, and tremorography measures in PMC with or without FXTAS

	CGG Repeat Size		Age		Sex
Variable	β (SE)	P Value	β (SE)	P Value	β (SE)	P Value
Combined kinetic tremor	0.01 (0.003)	0.001	0.01 (0.005)	0.02	−0.19 (0.1)	0.06
Right postural tremor	0.02 (0.005)	0.0001	0.01 (0.008)	0.12	−0.15 (0.17)	0.37
Left postural tremor	0.03 (0.006)	<0.0001	0.007 (0.009)	0.46	−0.28 (0.19)	0.15
Right finger taps speed	−0.002 (0.004)	0.62	0.02 (0.009)	0.048	0.36 (0.18)	0.047
Left finger taps speed	−0.003 (0.006)	0.57	0.03 (0.01)	0.005	0.34 (0.24)	0.17
Combined HMS	−0.0007 (0.003)	0.83	0.01 (0.007)	0.06	0.28 (0.14)	0.049
Combined RAMS	0.001 (0.003)	0.70	0.008 (0.006)	0.21	0.09 (0.13)	0.50
Combined RAMA	−0.001 (0.003)	0.64	0.01 (0.006)	0.08	−0.12 (0.11)	0.30

	Group (PMC Without FXTAS vs FXTAS)			CGG × Group Interaction (PMC Without FXTAS vs FXTAS)
Variable	β (SE)		P Value	β (SE)		P Value
Combined kinetic tremor	0.73 (0.44)		0.10	−0.01 (0.005)		0.04
Right postural tremor	1.82 (0.74)		0.02	−0.02 (0.008)		0.01
Left postural tremor	2.25 (0.84)		0.01	−0.03 (0.009)		0.002
Right finger taps speed	−0.13 (0.21)		0.53
Left finger taps speed	−0.17 (0.28)		0.55
Combined HMS	−0.24 (0.16)		0.14
Combined RAMS	−0.05 (0.15)		0.72
Combined RAMA	0.12 (0.13)		0.35

Regression Analysis
FXTAS
	Combined Kinetic Tremor		Right postural tremor		Left postural tremor
Variable	β (SE)	P Value	β (SE)	P Value	β (SE)	P Value
CGG repeat size	0.01 (0.003)	0.001	0.02 (0.006)	0.0006	0.03 (0.007)	0.0003
Age	0.02 (0.006)	0.009	0.02 (0.01)	0.07	0.01 (0.01)	0.32
Sex	−0.23 (0.11)	0.045	−0.19 (0.21)	0.37	−0.33 (0.25)	0.21

PMC Without FXTAS
	Combined Kinetic Yremor		Right Postural Tremor		Left Postural Tremor
Variable	β (SE)	P Value	β (SE)	P Value	β (SE)	P Value
CGG repeat size	0.0005 (0.003)	0.87	0.0006 (0.003)	0.84	−0.0009 (0.001)	0.45
Age	0.004 (0.008)	0.62	−0.003 (0.008)	0.74	−0.005 (0.003)	0.11
Sex	0.05 (0.22)	0.83	0.1 (0.22)	0.64	0.01 (0.08)	0.89

For measures with significant interaction between group and CGG repeat size, separate regression analyses were performed for the PMC with FXTAS and without FXTAS groups. Statistically significant values are bold.

Abbreviations: PMC, premutation carriers; FXTAS, fragile X‐associated tremor/ataxia syndrome; SE, standard error; HMS, hand movements speed; RAMS, rapid alternating movement speed; RAMA, rapid alternating movement amplitude.

FIG. 2.

Impact of CGG repeat size and age on postural and kinetic tremor in participants with FXTAS. The regression line and 95% confidence intervals are displayed. FXTAS, fragile X‐associated tremor/ataxia syndrome.

All FXTAS‐RS subset scores except finger taps and left RAM were significantly correlated with the severity scores on the corresponding tremorography variables (Table 5). Postural tremor (right, r = 0.62, P < 0.001; left, r = 0.76, P < 0.001) and kinetic tremor (right, r = 0.61, P < 0.001; left, r = 0.67, P < 0.001) had strong correlations. Moderate correlations were found for rest tremor (right hand, r = 0.37, P = 0.009; left hand, r = 0.34, P = 0.02), hand movements (right, r = 0.32, P = 0.022; left, r = 0.44, P = 0.001), and right RAM (right, r = 0.37, P = 0.009).

TABLE 5.

Correlations between FXTAS Rating Scale and tremorography severity scores in all FMR1 premutation carriers

Variable	Right, ρ	P Value	Left, ρ	P Value
Rest tremor	0.365	0.009	0.340	0.015
Postural tremor	0.620	0.0001	0.761	0.0001
Kinetic tremor	0.609	0.0001	0.665	0.0001
Finger taps speed and amplitude	0.064	0.657	0.166	0.245
Hand movement speed and amplitude	0.319	0.022	0.439	0.001
Rapid alternating movement speed and amplitude	0.367	0.009	0.025	0.865

Spearman correlation coefficient values (ρ) with significant values in bold.

Abbreviations: FXTAS, fragile X‐associated tremor/ataxia syndrome.

Discussion

Inertial sensor based tremorography was able to identify and quantify key UE motor deficits in FXTAS, namely, kinetic and postural tremors, coordination deficits, and bradykinesia. Not surprisingly, we found significantly worse kinetic tremor and bradykinesia in participants with FXTAS than PMC without FXTAS or healthy controls. Previous studies reported congruent findings, especially in kinetic and postural tremors. ¹³ , ¹⁵ , ¹⁶ As anticipated, kinetic tremor and postural tremor were the most abnormal and severe of all variables, followed by slowing of UE movements in all 3 of the speed variables: finger taps, hand movements, and RAM. All of these findings in FXTAS are congruent with the characteristic clinical manifestations of the disease. ¹ , ⁴ , ³³ Furthermore, the tremorography system separately quantifies different components of UE tremor and bradykinesia, including the speed, amplitude, and rhythm of movement, adding more specific information than that provided by the FXTAS‐RS.

The presence of significantly worse bradykinesia and dyscoordination in PMC participants without FXTAS in a number of speed variables than healthy controls were noteworthy in our study, as these findings could potentially represent prodromal signs of FXTAS. Our findings of significant bradykinesia in finger tap, hand movement, and RAM speed in PMC without FXTAS are also in alignment with a report that showed significant slowing in performance of fine motor pointing and manual dexterity tasks in PMC men without FXTAS compared with healthy controls. ²⁰ We importantly identified a new correlate of more specific signs of upper limb cerebellar deficits, namely, deficits in RAM. Our PMC without FXTAS had slower RAM speeds and greater RAM amplitude deficits than healthy controls. Abnormal RAM, also known as dysdiadochokinesia, is a classic sign of cerebellar pathology, and this finding may be a specific indicator of the onset of cerebellar degeneration. Prior imaging studies have demonstrated reduced functional connectivity in the middle cerebellar peduncles and reduced cerebellar volume in PMC without FXTAS, which could potentially form the basis for our findings. ³⁴ , ³⁵

Our findings of the potential for preclinical disease identification in PMC is supported by a number of prior studies from our group and others. We previously found that PMC without FXTAS demonstrated significantly delayed response latencies and disrupted sensory weighting for balance control. ²¹ Moreover, advancing age, male sex, increased CGG repeat size, and reduced X activation of the normal allele in PMC women predicted balance dysfunction. In a dual‐task gait paradigm to investigate the interaction between motor and cognitive function in PMC women without FXTAS, longer CGG repeat length was observed to predict the significantly greater dual‐task costs for gait speed compared with controls. ²³ Shickman and colleagues ²⁰ found a significant interaction between CGG repeat and age with higher CGG repeat numbers associated with the slowest movement times in older PMC men without FXTAS. Our findings warrant a deeper investigation into the complex pathology of the clinical onset of FXTAS in PMC by performing longitudinal studies with a larger sample size and simultaneous neuroimaging studies that will determine the specific cerebellar, cortical, and white matter motor tract involvement and their potential association with the development of tremor, bradykinesian and incoordination.

Higher CGG repeat size was significantly associated with worse kinetic and postural tremors and age was associated with worse kinetic tremor in participants with FXTAS, consistent with previous studies. ²⁹ Men with FXTAS had more severe kinetic tremor than women, which is also in alignment with prior findings. ⁴ , ⁵ , ²⁹ Neither CGG repeat size, age, nor sex showed a significant association with tremor or bradykinesia severity in PMC without FXTAS. This was not surprising given the low severity scores for these variables, especially those for tremor in this group.

Tremor, bradykinesia, and dysdiadokinesia scores derived from the Kinesia tremorography system were significantly correlated with the same measures on the clinician rated FXTAS‐RS, except for finger tap measures. This demonstrates validity of using Kinesia‐derived tremorography in PMC/FXTAS studies. However, kinetic and postural tremor scores had strong correlations while moderate correlations were found for rest tremor and bradykinesia measures. This could be attributed to the fact that kinetic tremor is a hallmark of the disease, whereas parkinsonian signs such as rest tremor and bradykinesia were likely present in a lower percentage of our participants. Indeed, rest tremor and finger tap amplitude were not found to be significantly different than controls in the FXTAS group consistent with our prior study. ²⁵ Other studies reported 12% to 26% of FXTAS patients displayed rest tremor and 30% to 67% had bradykinesia, ¹³ , ¹⁵ , ¹⁶ but parkinsonian symptoms are milder in FXTAS than in PD, ⁴ , ¹⁶ findings consistent with our study. Future validation with larger subject numbers and a wider range of clinical severities is recommended in future studies.

The limitations of this study include a relatively small sample size in the PMC without FXTAS group. Participant recruitment is challenging because of the rare nature of the disease. In addition, it is much harder to identify and recruit PMC men without FXTAS because these men do not come into the clinic, whereas PMC women are typically ascertained in fragile X clinics when they have a child with fragile X syndrome. Although the PMC participants without FXTAS in our study were predominantly women, who are much less likely to get FXTAS and tend to have less severe symptoms than men, ¹¹ , ³⁶ we posit that the significant fine motor coordination and speed deficits we observed in this group strengthens our findings of early markers of FXTAS. Future longitudinal testing with larger participant numbers is needed to confirm whether our findings are indeed predictive for the development of the disease. Lastly, the study is limited by the effects of any medication the participants with FXTAS were taking at the time of testing. This may have altered the clinical presentation of the disease in FXTAS participants, perhaps narrowing the comparison of the deficits with controls and PMC without FXTAS. Indeed, the severity scores on almost all tremorography variables were fairly low for participants with FXTAS. However, because 56% of participants with FXTAS were not on any medications for tremor or bradykinesia, it is likely that we had a cohort with relatively mild UE symptoms. Finally, test‐retest reliability studies using the Kinesia tremorography system in FXTAS has not yet been established and warrants further investigation.

In summary, tremorography discriminated subtle coordination deficits in PMC who did not meet the criteria for FXTAS and more fully characterized the nature of tremor deficits in FXTAS patients. We found that higher CGG repeat size was a significant predictor of greater kinetic and postural tremors and advancing age with greater kinetic tremor in participants with FXTAS. Identification of preclinical signs of fine motor speed and coordination deficits and potentially early signs of dysdiadochokinesia in PMC is vital for the early provision of disease‐modifying therapies as they become available in FXTAS. These sensitive markers not only allow for early detection of UE motor coordination and speed deficits but also have the potential to be used as outcome measures to evaluate the benefits of pharmacological and rehabilitative interventions, document disease progression, and measure efficacy in clinical trials for individuals with FXTAS.

Author Roles

(1) Research Project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript: A. Writing of the First Draft, B. Review and Critique.

J.O.:1A, 1B, 1C, 2A, 2C, 3A, 3B

D.B.: 1B, 2B, 2C, 3A, 3B

E.E.R.: 1B, 1C, 3B

A.B.: 1B, 2B, 2C, 3B

B.O.: 2A, 2 B, 2C, 3B

Y.L.: 2A, 2 B, 2C, 3B

G.P.: 1C, 3B

E.B.‐K.: 3B

D.A.H.: 1A, 1C, 3B

Disclosures

Ethical Compliance Statement

We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Funding Sources and Conflict of Interest

This work was supported by a National Institute of Health K01HD088762 (J.O.) and a National Fragile X Foundation Summer Fellowship Award (E.E.R.). The authors declare that there are no conflicts of interest relevant to this work.

Financial Disclosures for the Previous 12 Months

None of the authors has or has had any financial connection to the manufacturers of the Kinesia tremorography unit or analyzing software (Great Lakes Neurotechnologies, Inc., Cleveland, OH) and have not received any form of compensation for the conduct of the present study. Joan A. O'Keefe receives research support from the NIH (K01 HD088762). Deborah Bang, Erin Robertson, Alexandras Biskis, Bichun Ouyang, and Yuanqing Liu report no disclosures related to this manuscript. Gian Pal receives research support from National Institute of Neurological Disorders and Stroke (NINDS)/NIH (1K23NS097625), NINDS/Loan Repayment Program (LRP), Bakalar Family Foundation, Parkinson's Foundation Consulting/Advisory Board: Asim, Huron, AbbVie, Boston Scientific, KeyQuest, Qessential, SeaGrove, Medtronic, and Abbott. Elizabeth Berry‐Kravis has received funding from Seaside Therapeutics, Novartis, Roche, Alcobra, Neuren, Cydan, Fulcrum, Neurotrope, BioMarin, GW, Marinus, Zynerba, Ovid, Tetra, Retrophin, Acadia, Anavex, Biomarin, Neurogene, Orphazyme, GeneTx, Ionis, Vtesse, and Mallinckrodt Pharmaceuticals to consult on trial design or development strategies and/or conduct clinical trials in Fragile X Syndrome (FXS) and other genetic neurodevelopmental and neurodegenerative disorders, and from Asuragen, Inc. to develop testing standards and resources for FMR1 testing as well as research support from National Institute of Child Health and Human Development (NICHD), NINDS, NIMH, Center for Disease Control (CDC), The National Center for Advancing Translational Sciences (NCATS), Federal Drug Administration (FDA), FRAXA Research Foundation and the John Merck Fund. Deborah A. Hall has received research support from the NIH (R01 HD082110), CHDI Foundation, Fujifilm, Biohaven, and Neurocrine.