Clinical correlates of “pure” essential tremor: the TITAN study

Background To date, there are no large studies delineating the clinical correlates of “pure” essential tremor (ET) according to its new definition. Methods From the ITAlian tremor Network (TITAN) database, we extracted data from patients with a diagnosis of “pure” ET and excluded those with other tremor classifications, including ET-plus, focal, and task-specific tremor, which were formerly considered parts of the ET spectrum. Results Out of 653 subjects recruited in the TITAN study by January 2022, the data of 208 (31.8%) “pure” ET patients (86M/122F) were analyzed. The distribution of age at onset was found to be bimodal. The proportion of familial cases by the age-at-onset class of 20 years showed significant differences, with sporadic cases representing the large majority of the class with an age at onset above 60 years. Patients with a positive family history of tremor had a younger onset and were more likely to have leg involvement than sporadic patients despite a similar disease duration. Early-onset and late-onset cases were different in terms of tremor distribution at onset and tremor severity, likely as a function of longer disease duration, yet without differences in terms of quality of life, which suggests a relatively benign progression. Treatment patterns and outcomes revealed that up to 40% of the sample was unsatisfied with the current pharmacological options. Discussion The findings reported in the study provide new insights, especially with regard to a possible inversed sex distribution, and to the genetic backgrounds of “pure” ET, given that familial cases were evenly distributed across age-at-onset classes of 20 years. Deep clinical profiling of “pure” ET, for instance, according to age at onset, might increase the clinical value of this syndrome in identifying pathogenetic hypotheses and therapeutic strategies.

Background: To date, there are no large studies delineating the clinical correlates of "pure" essential tremor (ET) according to its new definition.
Methods: From the ITAlian tremor Network (TITAN) database, we extracted data from patients with a diagnosis of "pure" ET and excluded those with other tremor classifications, including ET-plus, focal, and task-specific tremor, which were formerly considered parts of the ET spectrum.
Results: Out of subjects recruited in the TITAN study by January , the data of ( .%) "pure" ET patients ( M/ F) were analyzed.The distribution of age at onset was found to be bimodal.The proportion of familial cases by the age-at-onset class of years showed significant di erences, with sporadic cases representing the large majority of the class with an age at onset above years.Patients with a positive family history of tremor had a younger onset and were more likely to have leg involvement than sporadic patients despite a similar disease duration.Early-onset and late-onset cases were di erent in terms of tremor distribution at onset and tremor severity, likely as a function of longer disease duration, yet without di erences in terms of quality of life, which suggests a relatively benign progression.Treatment patterns and outcomes revealed that up to % of the sample was unsatisfied with the current pharmacological options.
Discussion: The findings reported in the study provide new insights, especially with regard to a possible inversed sex distribution, and to the genetic backgrounds of "pure" ET, given that familial cases were evenly distributed across age-at-onset classes of years.Deep clinical profiling of "pure" ET, for instance, according to age at onset,

Introduction
After the release of the new tremor classification by the International Parkinson and Movement Disorder Society (IPMDS) in 2018, essential tremor (ET) has been redefined as an isolated action tremor syndrome of the upper limbs with a duration of at least 3 years without other neurological signs (1).A novel construct of ET-plus has been proposed to classify those patients who, in addition to the core phenotype of ET, also manifest with rest tremor or with additional "soft signs" of uncertain relationship to the tremor syndrome (1).Moreover, new diagnostic categories of focal tremor of the head/voice and task-specific/positiondependent tremor, which were formerly included in the definition of ET, have been proposed (1).These notable departures from previous definitions of ET were aimed at improving the phenotypic homogeneity of these patients, given previous failures to identify robust pathophysiologic and etiologic correlates of ET with its loosely defined boundaries (2,3).Although the distinction between ET and ET-plus might theoretically serve the purpose of creating homogenous groups of patients for subsequent studies (4), distinguishing between ET and ET-plus by clinical distinction alone and in the absence of clear objective biomarkers is very complex (5,6).Nonetheless, the revision of the nosology of tremor syndromes by the 2018 IPMDS tremor classification, especially with regard to ET, conveys the implication that previously obtained information about the clinical features and natural history of ET needs to be tested (7).Several attempts have since been pursued to retrospectively re-classify patients according to the new classification and to provide a comparison between "pure" ET and ET-plus (7)(8)(9)(10)(11).They showed that ET-plus would be commoner, would have a higher age at onset (AAO), and would have a worse progression than ET (7)(8)(9)(10)(11), and these results mirror those obtained by a recent longitudinal study on 37 patients with "ET, " of whom more than 80% had, in fact, ET-plus at baseline (12).However, all these studies emphasized ET-plus and did not provide a detailed account of "pure" ET.In fact, no studies have attempted, to the best of our knowledge, to delve into the entity of "pure" ET diagnosed according to the new classification and to detail its clinical features.It might be possible that the removal of ET-plus as well as of focal and task-specific tremor from the group of "pure" ET, as suggested by the 2018 IPMDS consensus (1), would disprove some of the old concepts about ET in general, and this needs to be formally tested.
The ITAlian tremor network (TITAN) is a multi-center data collection platform (13), aiming to prospectively assess the phenomenology and natural history of tremor syndromes and to serve as a basis for future etiological, pathophysiological, and therapeutic research.For the current study, the aim of which is to provide a detailed account of "pure" ET, further in relation to such sensible anchors as AAO and the presence of family history (FH), we have extracted from the TITAN database the clinical data of patients with a diagnosis of "pure" ET according to the 2018 IPMDS criteria.We here present a cross-sectional analysis of the baseline data of patients with "pure" ET and describe its clinical features and its impact on the activities of daily life (ADL) and the quality of life (QoL).

Methods
The protocol and preliminary findings of the TITAN study have been published elsewhere (13).In brief, patients aged > 18 years with any tremor syndromes, but the ones combined with Parkinsonism, were eligible for recruitment (13).Patients were prospectively recruited, and the diagnosis of ET, or otherwise, was based on the current tremor classification (1).Specifically, investigators were asked to indicate the tremor diagnosis according to the 2018 IPMDS consensus for every recruited patient and to list, in the case of ET-plus, which soft signs were present.For the current study, we extracted from the TITAN platform data of patients with a diagnosis of "pure" ET, therefore excluding patients with ET-plus, focal tremor, task-specific/position-dependent tremor, or any other tremor diagnosis.
All patients were assessed with a standardized protocol (13) including a structured interview to gather the following variables: sex, age at examination, self-reported AAO, presence of FH for tremor or for other neurological diseases, tremor distribution at onset, task-specificity at onset, and presence of sensory trick at evaluation.The presence of FH was operationalized as the existence of tremor (or other neurological disorders as detailed below) in either first-or second-degree family members (14), with the percentages reported in this study representing cumulative figures.As per the protocol, the history of any type of tremor would satisfy a positive FH, regardless of its distribution or putative diagnosis (e.g., ET or other tremor diagnoses), with the exception of tremor in the context of a diagnosis of Parkinsonism, which instead qualifies for FH for "other neurological disorders".The latter information was collected to test the hypothesis of the possible clustering of different disorders.
Patients were assessed with the Essential Tremor Rating Assessment Scale (TETRAS) (15) which is composed of two parts: the performance subscale, with its total score providing an index of tremor severity, and the ADL subscale depicting tremor impact on ADL.Using individual items of the TETRAS performance subscale, we calculated the following variables: (1) "postural tremor" as the mean of the scores assigned to the items assessing upper limb tremor in the forward outstretched and lateral "wing beating" positions; (2) "kinetic tremor" as the mean of the scores assigned to the items assessing upper limb tremor during the finger-nose test and the Archimedes spirals drawing; (3) the "postural-kinetic ratio" computed as ("postural tremor" -"kinetic tremor")/("postural tremor" + "kinetic tremor"); and (4) the asymmetry index as (right-sided items -left-sided items)/(right-sided items + leftsided items).To depict the craniocervical involvement of tremor, we summed the proportion of patients with neck, face, or voice tremor.Finally, patients were asked to report on a scale from 0 (worst possible) to 100 (best possible) their perceived level of QoL.
After the release of the first protocol and the beginning of the research activities (13), an amendment was approved to gather information about the current and past pharmacological treatment.As such, the following information is available only for a subset of the recruited sample (88/208; 42.3%).Investigators were asked to record all current treatments for tremor with daily dosages and the corresponding patient-global impression of change (P-GCI) on an 11-point Likert-like scale with three explicit anchors (e.g., 0 = much better; 5 = no change; and 10 = much worse) (16).Similarly, investigators were asked to record previous treatments with the reason for withdrawal [e.g., inefficacy or presence of adverse events (AEs)].
The study has been approved by the ethics committee of the coordinator center (University of Salerno; study approval n.33_r.p.s.o._02/10/2020), and all subjects were requested to provide a written consent form to participate.

Statistical analysis
We first examined each variable by computing some descriptive statistics (mean, quartile, variance, kurtosis, and skewness) in order to check their distribution and the presence of outliers.
For numeric variables, the t-test and ANOVA tests were performed for comparing the mean between two groups or more than two groups, respectively.Furthermore, in the case of more than two groups, we used Tukey's honestly significant differences to check for which groups the differences were significantly different.For categorical variables, a Pearson's chi-square test was performed.Unless otherwise stated, data are given as mean ± standard deviation (SD), the theoretical significant level α being fixed at 5%.All analyses were implemented using R (17), RStudio (18), ggplot2 (19), and psych (20) packages.

Results
Out of 653 subjects recruited in the TITAN study by January 2022, the data of 208 (31.8%) patients (86M/122F) with "pure" ET were extracted for the current study."Pure" ET represented the second most common tremor diagnosis after ET-plus (41.3%) and before tremor combined with dystonia (13.6%) (13).
Patients with "pure" ET had a mean age of 67.49 ± 12.27 years and a mean disease duration of 20.49 ± 20.27 years.Ninetynine patients (47.6%) reported a positive FH for any neurological disorders, whereas the remaining patients were identified as sporadic cases.The distribution of familial cases by the AAO class of 20 years showed significant differences (χ2 = 13.563,p = 0.035), with sporadic cases representing the large majority of the class with AAO above 60 years and patients with a positive FH for tremor being evenly distributed across the classes although they were most represented in the class of age 21-40 years of AAO (Figure 1).The distribution of AAO was found to be bimodal, with the antimode being identified at 34 years (Figure 2).Accordingly, patients were stratified into two groups (e.g., early-onset < 34 years and late-onset ≥ 34 years) for subsequent analyses.
In the entire sample, tremor was fairly symmetric (e.g., asymmetry index = 0.01 ± 0.35, Figure 3), and the postural component was more severe than the kinetic component (i.e., with a positive postural-kinetic ratio, Table 1).Tremor involved the cranio-cervical region in ∼40% and the lower limbs in about 12% of cases.Task specificity at onset was reported by 16 patients (7.69%), and 16 additional cases (7.69%) were found to have a sensory trick at the examination.Table 1 details the demographic and clinical features of the entire sample.Treatment data were available for a subset of 88/208 (∼42%).Over 70% of this subset was on one or more anti-tremor drugs, with only ∼5% of these cases not being on any drugs because of previous failures (Figure 4A).The most commonly prescribed drug was propranolol (48.6%; mean daily dose of 64.17 ± 24.31 mg), followed by primidone (15.8%; mean daily dose of 155.21 ± 144.99 mg), clonazepam (14.5%; mean daily dose of 1.14 ± 1.15 mg), and other drugs (Figure 4B).Of patients on these treatments, 72.9, 83.3, 54.5, and 81.2%, respectively, were reported to be improved/much improved (Figure 4B).Approximately 42% of this sample further reported having previously discontinued other anti-tremor drugs either because of inefficacy (45.9%) or because of the emergence of AEs (54.1%).Figure 4C details the relative proportions of inefficacy or AEs for each anti-tremor drug.

Comparisons between sporadic and familial cases
Out of 208 patients with ET, 85 patients (40.9%) reported a positive FH for tremor, 14 patients (6.7%) reported a positive FH for other neurological diseases including Parkinsonism and dementia, whereas 109 patients (52.4%) were sporadic cases.
AAO and age at evaluation were significantly different [ANOVA F (2,205) = 3.20; p = 0.043 and F (2,205) = 3.05; p=0.049, respectively], whereas disease duration was similar between the three groups.Post-hoc tests demonstrated that patients with FH for tremor had a significantly lower AAO (p=0.032;Table 2) and age at evaluation (p = 0.045; Table 2) than sporadic cases.Approximately 22% of cases with positive FH for tremor had leg involvement which was significantly higher than sporadic cases and cases with FH for other neurological diseases (Fisher's exact, two-tailed p=0.01,Table 2).No other differences were found between the groups in terms of clinical features at onset (e.g., onset site and presence of sensory trick and/or task-specificity), tremor phenomenology (e.g., postural-kinetic ratio and asymmetry index), craniocervical distribution, TETRAS performances, ADL subscales, and QoL (Table 2).No differences were observed in terms of treatment patterns or outcomes between the two groups (for all p > 0.05).

Comparisons between early-onset and late-onset cases
Sixty-six patients (31.7%) were classified as early-onset, whereas the remaining patients (142, 68.3%) as late-onset cases.Age at evaluation was significantly lower (61.00 ± 15.49 vs. 70.51± 9.02 years; p < 0.001), and disease duration was significantly longer (44.42 ± 18.11 vs. 9.37 ± 7.71 years; p < 0.001) in the early-onset than in the late-onset group (Table 3).The site of tremor onset was significantly different between the two groups (χ2 = 18.00; p < 0.001), with more than 90% of early-onset cases and ∼65% of lateonset cases having a bilateral tremor of the upper limbs (Table 3).Tremor severity was higher in the early-onset than in the late-onset group (TETRAS performances subscale = 16.90 ± 6.12 vs. 14.85 ± 7.09; p = 0.034) as it was the postural tremor (1.49 ± 0.55 vs. 1.23 ± 0.61; p = 0.003), whereas kinetic tremor was not different (Table 3).The TETRAS-ADL score was not significantly different between the two groups, and the comparison only approached the statistical threshold (p = 0.058).The two groups were similar in terms of the remaining gathered variables (Table 3).No differences were observed in terms of treatment patterns or outcomes between the two groups (for all p > 0.05).

Discussion
To the best of our knowledge, this is the first and largest multicenter study providing a detailed account of ET based on the 2018 IPMDS classification.On the one hand, our results broadly recapitulate previous findings, but it should be noted that nearly all available literature on the subject adopted former definitions of ET so comparisons should be taken cautiously.On the other hand, we also provide novel insights on "pure" ET, which indeed might stem from its new definition, as continued below.
Previous accounts generally found ET to be more prevalent in men, as confirmed in a recent meta-analysis (21), with only a few exceptions (11,22).We observed a 1:1.5 ratio between men and women: this inverse proportion might be due to longer life survival in female elderly patients attending neurological consultation (11), but might also suggest sex-related differences in ET.The latter deserves to be tested in future worldwide epidemiological studies in view of the growing importance of understanding sex-related pathways to disease development (23).
As previously described (21,(24)(25)(26), frequency figures of ET steadily increased with advancing age, and we observed a bimodal distribution for the AAO with a peak in the second decade of life and a later increase in the sixth and seventh decades.As discussed in more detail below, this bimodal distribution seemed to reflect the effect of aging.Overall, a positive FH for tremor was reported in ∼40% of the sample, which is lower than most  previous reports of 60-75% (12,(25)(26)(27).This does not seem to be related to the difference in the criteria of operationalization for FH, as our criteria are arguably looser than those adopted in previous studies, which in most cases focused on the explicit diagnosis of ET in family members.One possibility is that the observed difference might be due to some sort of bias because of recruitment in single tertiary, university-based centers in most previous studies.In fact, our figure is in line with a study recruiting patients in three urban hospitals (25) as well as with the only available study that adopted the new IPMDS classification, yet retrospectively (11).Indeed, the possibility that the new definition of ET in itself has had an impact on these results cannot be ruled out.Interestingly, a small proportion of familial cases (∼7%) were reported to have a FH for diseases other than tremor.This might suggest a possible clustering of different phenotypes including but not limited to ET (28).However, the possibility remains that these results are, at least in part, spurious given that parkinsonism and dementia are relatively common in the general population.Genetic testing of the TITAN cohort might shed light on this issue.Importantly, FH for tremor was evenly distributed across classes of AAO and did not account only for early-onset cases, whereas the impact of aging was preferentially observed for sporadic patients.This resulted in a lower AAO in familial than sporadic cases.Patients with and without family history were otherwise comparable for all other gathered variables, except for leg involvement in patients with FH for tremor.Given that disease duration was comparable between the groups, this result would reflect an intrinsic predisposition of patients with FH for tremor to develop leg tremor.This finding is in contrast to Chen et al. ( 27) who found leg tremor to be more common in patients with craniocervical tremor, longer disease duration, and higher tremor severity.It should be noted, however, that the latter results stem from research adopting former definitions of ET, thus likely including cases that we would currently label with ETplus and that have been excluded from the current analysis which focused instead on "pure" ET.It has been shown that, applying the new tremor classification, leg tremor would be more common in ET-plus cases (8) who, in turn, also tend to have higher tremor severity (8,11,13).This might, therefore, explain the discrepancy between our and previous results.This consideration should be contemplated when examining all previously published literature as the new classification has likely had a great impact on the nature of the examined samples.Therefore, our results might set the "reference" according to the new definition of "pure" ET according to the 2018 IPMDS consensus and need to be further validated in other populations of different ethnicities.Comparisons between early-onset and late-onset cases demonstrated some significant differences which, in view of the above, might be largely attributable to the longer disease duration in the former.Early-onset cases had, in fact, higher scores in the TETRAS performances subscale and in its postural items, which suggests that this is the tremor component prone to worsening over time.This, however, did not grossly impact ADL and QoL suggesting a relatively benign progression of "pure" ET in contrast to what has been shown for ET-plus (11,12).Of note, some differences were identified between the two groups at onset, with the late-onset group reporting no or unilateral arm tremor at onset more frequently than the early-onset group.This might suggest that AAO is a useful anchor to identify clinically, and possibly biologically, different subtypes (25).
Overall, tremor phenomenology and distribution confirm most but not all previous studies (29).In our study, tremor was found to be largely symmetrical and more severe in its postural component.This is in line with what was generally assumed for ET (30) but is in contrast to Louis who suggested that the primary type of tremor in ET would be kinetic (29).Our figure of craniocervical tremor being present in ∼40% of ET might appear slightly greater than previously reported in studies using the new definition of ET (11).However, we note that this might be due to the fact that we summed the proportion of patients with face, voice, and neck tremor.The reason for our choice stands in the fact that "midline tremor" has been suggested to be a proxy of greater clinical-pathological severity (10,31).However, we did not find statistical differences in the frequency of midline tremor between patients stratified either according to the presence of FH or AAO although its frequency was higher in patients with early-onset and longer disease duration.
Importantly, we have reported the presence of some phenomenological features such as unilateral arm tremor and task-specificity at onset, and the presence of sensory trick at evaluation, which is generally assumed to be indicative of different tremor syndromes (32).These data might, therefore, turn useful in the attempt of calculating sensitivity/specificity values of these features against an alternative tremor syndrome (33), particularly of dystonia.These results further highlight one of the novelties of the new tremor classification, which is the possible diagnostic transition across tremor categories (1,2).In fact, some of the patients reported here would have been diagnosed at onset as task-specific tremor to subsequently receive an ET diagnosis.Longitudinal analysis of this cohort will clarify whether these patients will develop overt dystonia.
Finally, we also report on treatment patterns and outcomes in our sample.On the one hand, they are in line with both evidencebased (34) and expert recommendations (35) about the treatment of ET, propranolol and primidone being the two most commonly prescribed drugs.On the other hand, we also observed a relatively frequent prescription of clonazepam, which was reported to be useful by at least 50% of patients on this drug.This contrasts with a small double-blinded trial that refuted the efficacy the clonazepam in ET (36).It should be noted, however, that we collected a patientreported outcome and clonazepam might have been perceived as beneficial because of a primary effect on anxiety/embarrassment and of a secondary, indirect, effect on tremor.Approximately 10% of patients were receiving botulinum toxin injections for their tremor, ∼5% of our sample were not receiving any drugs because of previous failures, and another one-third of patients reported no change or to be worse on the current drug regimen.This sums to ∼40/45% of patients who are not satisfied with the current pharmacological options for ET and underscores the need for the development of new drugs for ET.
We acknowledge that our study has some limitations.First, we cannot entirely exclude a recruitment bias that is inherent to studies without a population-based design.However, the involvement of both secondary and tertiary movement disorder centers in the TITAN study (13) might have, at least in part, attenuated this risk, and therefore, our account of ET should be more realistic than those obtained in single-center studies.Second, we note that the diagnosis of tremor syndromes is made on a clinical basis given the lack of available biomarkers, and this might carry the risk of misdiagnosis in a proportion of patients.However, we strictly adhered to the current classification (1), and this represents the first study in which the diagnosis of ET according to the 2018 IPMDS tremor consensus has not been made retrospectively on the chart review.Moreover, the relatively long disease duration in our sample makes it unlikely that these patients have alternative conditions, including rare and/or treatable disorders (28), that usually present with combined tremor syndromes rather than with an isolated tremor syndrome such as ET.Third, we cannot exclude a certain degree of recall bias regarding the presence of FH.However, it has been demonstrated that the sensitivity of selfreported FH in patients with ET reaches ∼85% (37).Similarly, self-reported AAO might also have suffered from a recall bias.However, prior studies have indicated that it is reliably reported by ET patients (38).We also acknowledge that alcohol responsiveness was not initially included in the study protocol since the 2018 IPMDS tremor consensus considered it to be unspecific for ET (1).However, this information has been included in a protocol amendment of the TITAN study, and we plan to compare different tremor syndromes for this feature in future reports.Finally, we acknowledge that information about past/current treatment and outcomes was available only for a subset of patients.However, missing data were equally distributed across the subgroups we analyzed (e.g., early-vs.late-onset and familial vs. sporadic cases), and therefore, this should not have biased the results in a major way.
In summary, we have here reported the clinical correlates of ET according to the new tremor classification.On the one hand, they broadly recapitulate the findings obtained by previous research using former definitions of ET, which might indirectly support the argument that the term ET-plus may be artificious.It is beyond the scope of the current study to add arguments to this debate, but we note that opposite experimental evidence favoring the distinction between ET and ET-plus is increasingly being produced (11,13,33,39).On the other hand, the current study also provides novel insights with regard to a possible inversed sex distribution, the contribution of aging in late-onset cases, and the presence of some phenomenological features that might be characteristic of familial cases or have been classically attributed to different tremor syndromes.The observed differences with previous studies that adopted alternative operational criteria for ET further demonstrate the fulfillment of the primary intent of the 2018 IPMDS tremor classification to create a more homogeneous group of patients for future research.Our results would add that deep phenotyping of patients with "pure" ET might even better serve this purpose, and, in this regard, AAO seems to best identify distinct entities.Deep clinical phenotyping of ET, both of pure and plus forms, might increase the clinical value of this syndrome in identifying pathogenetic hypotheses, therapeutic strategies, and clinical research in the future.

FIGUREFIGURE
FIGUREDistribution of patients by the presence of family history in the entire sample stratified in age-at-onset classes of years.

FIGURE
FIGUREHistograms reflecting the density of frequency of the asymmetry index in the entire sample.The red curve represents the kernel density of the asymmetry index.Values close to , around which it peaks in the current cohort, indicating a symmetrical tremor between the two arms.

FIGURE
FIGURE Treatment patterns and outcome.(A) The left pie graph shows the percentage of patients being on treatment or not.(B) The central pie graph shows the prescribed anti-tremor drugs with the plain slices representing patients reporting improvement/much improvement and dotted slices representing patients reporting no change/worsening in their tremor.(C) The right pie graph shows the percentage of patients discontinuing anti-tremor drugs because of ine cacy (plain slices) or the emergence of side e ects (dotted slices, in which cases relative percentages are depicted).
TABLE Demographic and clinical features of the entire cohort.
M, male; F, female; FH, Family history; TETRAS, The Essential Tremor Rating Assessment Scale; ADL, Activity of daily living; QoL, Quality of life.Data are reported as mean ± standard deviation or percentages.
TABLE Demographic and clinical features of the our cohort stratified according to the presence of family history for tremor or other neurological diseases.TABLE Comparisons between early-onset (< years) and late-onset (≥ years) cases.
NO FH (N =) FH for tremor (N = ) FH for other neurological diseases (N = ) M, male; F, female; FH, Family history; TETRAS, The Essential Tremor Rating Assessment Scale; ADL, Activity of daily living; QoL, Quality of life.Data are reported as mean ± standard deviation or percentages.a Different from patients with no FH (p < 0.05).b Different from patients with FH for other neurological diseases (p < 0.05).M, male; F, female; FH, Family history; TETRAS, The Essential Tremor Rating Assessment Scale; ADL, Activity of daily living; QoL, Quality of life.Data are reported as mean ± standard deviation or percentages.