Abstract
Background: Action tremor may be more prevalent in relatives of patients with Parkinson’s disease (PD) than in relatives of control subjects. This tremor could represent mild PD or essential tremor. An estimate of the risk of this condition in families of patients with PD is important when studying the genetics of PD.
Objectives: To determine the risk of action tremor in first-degree relatives of probands with tremor-dominant PD (TD-PD) and postural instability gait disorder PD (PIGD-PD) compared with first-degree relatives of control probands.
Methods: PD and control probands participated in a familial aggregation study of PD. The presence of action tremor in their relatives was ascertained from reports of one or more informants. Relatives who met diagnostic criteria for PD were excluded. Cox proportional hazards models adjusting for gender, education, race, and vital status (dead vs alive) of the relatives were used to assess the relative risk (RR) of action tremor in first-degree relatives of PD probands vs first-degree relatives of control probands.
Results: There were 487 PD probands, 409 control probands, and 5,563 relatives. The risk of action tremor was higher in the relatives of TD-PD probands than in the relatives of control probands (RR = 2.14; 95% CI = 1.53 to 2.98) but not in the relatives of PIGD-PD probands compared with the relatives of control probands (RR = 1.81; 95% CI = 0.66 to 5.02).
Conclusion: The risk of action tremor was increased in the relatives of PD probands, particularly when they had TD-PD. Whether the tremor in these relatives represents essential tremor or an isolated manifestation of PD requires further investigation.
Parkinson’s disease (PD) is clearly more prevalent in the relatives of patients with PD than in the relatives of control subjects.1,2⇓ In addition, tremor in the absence of a diagnosis of PD may be more prevalent in the relatives of patients with PD than in the relatives of control subjects.3,4⇓ Whether the tremor in these relatives represents a limited manifestation of PD or whether it represents essential tremor (ET) is not well understood. Therefore, an estimate of the frequency with which this condition occurs in families of patients with PD is critical when studying the genetics and especially the familial aggregation of PD. The incidence and prevalence of ET5,6⇓ and PD7 increase markedly with age, and the risk of these diseases in relatives is expected to increase with age, although studies of the familial aggregation of tremor in PD have not adjusted for the ages of the relatives.3,4⇓
The type of PD in the proband also may influence the familial risk of tremor. In several studies,4,8⇓ patients with PD with prominent tremor were particularly prone to have a relative with tremor, although results did not reach significance.
The goal of the present study was to determine risk of action tremor in first-degree relatives of PD probands, including probands with tremor-dominant PD (TD-PD) and postural instability gait disorder PD (PIGD-PD) compared with first-degree relatives of control probands, adjusting for the ages of the relatives and other potential confounding factors.
Methods.
Selection of PD probands and control probands.
In this family study of PD, we enrolled probands with PD and control probands, and ascertained disease status of their respective first-degree relatives. Probands with PD were recruited during a 4-year period from the Center for Parkinson’s Disease and Other Movement Disorders at Columbia-Presbyterian Medical Center (a tertiary referral center for the New York, New Jersey, and Connecticut region) and from a neurology clinic at Columbia-Presbyterian Medical Center that primarily serves patients from the Washington Heights community in northern Manhattan, New York. For the PD probands from the Center for Parkinson’s Disease, control probands from the New York, New Jersey, and Connecticut region were recruited during the same period by random-digit dialing and were frequency matched to the PD probands by 5-year age strata, gender, ethnicity, and telephone area code and exchange. For the PD probands from the neurology clinic, control probands were recruited using a 50% sample of names and addresses of Medicare recipients aged ≥65 years living in the Washington Heights community in northern Manhattan, New York. The Health Care Finance Association (HCFA) provided this information. For 18 patients who were aged <65 years, a suitable control subject could not be identified using this method, and these 18 control subjects were recruited from the control series of the Northern Manhattan Stroke Study (NOMASS) in the Washington Heights community, which used random digit dialing for control selection.9 Control probands were frequency matched to the PD probands from the neurology clinic by 5-year age strata, gender, and ethnicity.
Evaluation of PD probands and control probands.
All PD and control probands underwent an evaluation that included a medical history and a neurologic examination, which included the Unified PD Rating Scale10 and an ET examination (performance of six tasks with each arm to evaluate action tremor).11 Each of the 12 action items on the ET examination was rated (E.D.L.) using a 0 to 3 scale, and a total tremor score (range = 0 to 36) was assigned.11 The modified Mini-Mental State Examination (MMSE; range = 0 to 57)12 was administered to all PD and control probands, and to reduce the possibility of collecting family history information of questionable validity, those who scored <40 (indicating possible cognitive impairment) were excluded. In addition to the MMSE, Spanish-speaking PD probands from the neurology clinic were administered a core neuropsychological battery used in previous studies.13,14⇓ The Spanish-speaking PD probands who scored <40 on the MMSE but who did not meet criteria for dementia based on the neuropsychological battery were included.
For PD probands, criteria for idiopathic PD15-17⇓⇓ included two of four cardinal signs: bradykinesia, rest tremor, rigidity, and postural instability, one of which had to be bradykinesia or rest tremor. PD probands were excluded if they had other forms of parkinsonism (e.g., progressive supranuclear palsy, multisystem atrophy, drug-induced parkinsonism) or if they had a history of dementia before the onset of parkinsonism. PD probands were categorized by whether they had TD-PD or PIGD-PD and by the age at onset of motor signs of PD (≤50 or >50 years). Criteria for TD-PD and PIGD-PD were derived from the DATATOP study, which used a ratio of mean tremor score to mean PIGD score. When the ratio was ≥1.5, PD probands were classified as TD-PD, and when the ratio was ≤1, they were classified as PIGD.18 Eighty PD probands could not be classified as either TD-PD or PIGD-PD because they had a ratio that was >1 and <1.5.
Ascertainment of PD in first-degree relatives.
To ascertain PD and other neurologic disease in first-degree relatives of PD or control probands, a structured family history interview was administered to the proband. Each first-degree relative was enumerated, and demographic information and six screening questions about the cardinal motor manifestations of PD were collected. An affirmative answer to any one of the six screening questions (i.e., a positive screen) prompted a set of follow-up questions about PD in that relative. To increase the sensitivity of the interview, we tried to obtain a second interview, by telephone, with the first-degree relative or with another informant if the first-degree relative had died or was unavailable. The Telephone Interview for Cognitive Status (TICS)19 was administered to each first-degree relative or the person who served as an informant; data provided by informants who scored <30 (indicating cognitive impairment) were not used. To resolve any ambiguities, supplemental information was collected from a review of the first-degree relative’s medical records. For first-degree relatives who lived within 2 hours of Columbia-Presbyterian Medical Center or who had symptoms of PD, an in-person interview and physical examination was also used to screen for the diagnosis of PD. A diagnosis of possible, probable, or definite PD was assigned to a first-degree relative if, based on all available information (interviews, medical record review, and physical examination, if applicable), at least three of the following appeared to be present: 1) tremor at rest; 2) shuffling gait; 3) stooped posture or decreased arm swing; and 4) complaint of performing daily activities stiffly or slowly, or, alternatively, if at least one of these was present along with a previous diagnosis of PD or treatment with levodopa.20 The validity of these diagnostic criteria has been examined and was high (sensitivity = 95.5%; specificity = 96.2%).20 Ninety-six relatives (71 patient relatives and 25 control relatives) were excluded from the main analyses because they were diagnosed with PD.
Ascertainment of action tremor in first-degree relatives.
For each first-degree relative, the PD and control probands also answered a set of 12 validated screening questions for action tremor (e.g., “Does your hand usually tremble when you drink or pour from a cup or a glass?”).11 These included the question, “has a doctor diagnosed you as having familial tremor or benign essential tremor?” This question incorporated the specific term “essential tremor” rather than the more general term “tremor.” It also required a physician diagnosis of ET rather than self-diagnosis. In the Washington Heights-Inwood Genetic Study of Essential Tremor,21 this question elicited the smallest proportion of false-positive responses (0.4%).
For the current study, action tremor was defined alternatively according to liberal criteria (an affirmative answer to any one of the 12 screening questions for action tremor) or conservative criteria (an affirmative answer to the one question, “has a doctor diagnosed you as having familial tremor or benign essential tremor?”).
Statistical analyses.
Student’s t-tests were used to assess continuous variables, and chi-square tests were used for categorical variables. We applied Cox proportional hazards models with double-censoring techniques for missing information22 to model the relative risk (RR) for action tremor among first-degree relatives of PD and control probands. Left censoring was applied when a relative was known to have action tremor, but the onset date was unknown. Right censoring was applied if action tremor was not reported in the relative. The current age of the relative was used when right censoring was applied except when the relative was deceased, and in that case, age at death was used to right censor. Using this maximum likelihood method, the entire sample of relatives was used to estimate the RR of action tremor in first-degree relatives of patient probands compared with first-degree relatives of control probands. The model was adjusted for the independent effects of gender, education, race (white vs nonwhite), and vital status (dead vs alive) of the relatives. In additional analyses, the PD probands were stratified into TD-PD and PIGD-PD and based on age at onset (≤50 vs >50) of PD.
Results.
There were 487 PD probands (168 with TD-PD, 239 with PIGD-PD, and 80 who could not be classified as either) and 409 control probands. Action tremor (defined as a rating ≥1 [mild or greater tremor] on one of the 12 items on the ET examination) was present in 95.7% of TD-PD probands and 93.8% of PIGD-PD probands (χ2 = 0.39; p = 0.53), and the total tremor score in TD-PD probands was 9.8 ± 5.8 vs 6.4 ± 4.4 in PIGD-PD probands (t = 4.80; p = 0.001). Five thousand five hundred sixty-three first-degree relatives did not meet criteria for PD, including 1,026 relatives of probands with TD-PD, 1,526 relatives of probands with PIGD-PD, 453 relatives of PD probands who could not be classified as either, and 2,558 relatives of control probands.
The PD probands were younger than the control probands (table 1). First-degree relatives of PD probands were younger than first-degree relatives of control probands, were less likely to be white, and were more likely to be living (table 2). First-degree relatives of TD-PD probands were younger, had more years of education, and were more likely to be living than were first-degree relatives of control subjects (see table 2).
Table 1. Demographic and clinical characteristics of PD and control probands
Table 2. Demographic and clinical characteristics of the first-degree relatives of PD probands and first-degree relatives of control probands
A larger proportion of first-degree relatives of TD-PD probands had action tremor compared with first-degree relatives of control subjects (table 3). In an unadjusted Cox model, the risk of action tremor (defined using the liberal definition) was higher in first-degree relatives of PD probands than in first-degree relatives of control subjects (RR = 1.58; 95% CI = 1.20 to 2.09; p = 0.001). In a Cox model, adjusting for gender and vital status of the relative, the risk of action tremor was higher in first-degree relatives of PD probands than in first-degree relatives of control subjects (RR = 1.55; 95% CI = 1.18 to 2.05; p = 0.002). In a Cox model, adjusting for gender, education, race (white vs nonwhite), and vital status of the relative, the risk of action tremor was higher in first-degree relatives of PD probands than in first-degree relatives of control subjects (table 4). In unadjusted Cox models and in Cox models adjusting for gender, education, race, and vital status of the relative, the risk of action tremor was higher in first-degree relatives of TD-PD probands compared with first-degree relatives of control subjects but similar in first-degree relatives of PIGD-PD probands compared with first-degree relatives of control subjects (see table 4). In unadjusted and adjusted Cox models, the risk of action tremor did not differ between first-degree relatives of young-onset vs old-onset PD probands (see table 4).
Table 3. Proportion of first-degree relatives of PD probands and first-degree relatives of control probands with action tremor
Table 4. Risk of action tremor in first-degree relatives
Ninety-six relatives (71 patient relatives and 25 control relatives) were excluded from the main analyses because they were diagnosed with PD. However, a patient may have action tremor and PD. In a subanalysis, we included these relatives. In a Cox model, adjusting for gender, education, race, and vital status of the relative, the risk of action tremor (defined using the liberal definition) was higher in first-degree relatives of TD-PD probands compared with first-degree relatives of control subjects (RR = 2.25; 95% CI = 1.65 to 3.05; p ≤ 0.001) but was similar in first-degree relatives of PIGD-PD probands compared with first-degree relatives of control subjects (RR = 1.30; 95% CI = 0.94 to 1.80; p = 0.11).
We performed additional analyses to examine the possibility of recall bias involving over-reporting of tremor by probands with TD-PD vs control probands. For this purpose, we stratified by whether the data were derived from the proband or from self-report of the first-degree relative. Recall bias may result in higher estimates of RR based on the probands’ reports than based on self-reports. The adjusted RR of action tremor (defined using the liberal definition) in first-degree relatives of TD-PD probands compared with first-degree relatives of control subjects was 1.80 (95% CI = 1.14 to 2.84; p = 0.01) based on the proband report and 2.24 (95% CI = 1.22 to 4.12; p = 0.009) based on self-report. Hence, the magnitude of increased risk did not differ by source of information.
We also examined whether probands, and especially control probands, under-reported tremor in their relatives. For this purpose, we examined the validity of PD proband and control proband reports by comparing them with a gold standard (i.e., an in-person tremor examination, which was performed blinded to any knowledge of reported tremor). One hundred forty-four first-degree relatives (4.0% of 3,617 living relatives) underwent an in-person tremor examination, and 14 met published diagnostic criteria for ET (a moderate or greater amplitude action tremor during three or more tasks, or a head tremor).14 Ten were relatives of PD probands. Nine of these 10 (sensitivity of PD proband’s report = 90%) were reported by the proband to have action tremor (liberal criteria). Four relatives of control subjects met diagnostic criteria for ET based on the tremor examination. All four (sensitivity of control proband’s report = 100%) were reported by the proband to have action tremor (liberal criteria). Overall, reported tremor was a sensitive method of diagnosing ET, with 13 of 14 (92.8%) who were diagnosed with ET based on examination having been reported by the proband to have action tremor.
Discussion.
We studied the familial aggregation of action tremor in first-degree relatives of probands with PD compared with first-degree relatives of control probands. Our analyses adjusted for the ages of the relatives and for other potential confounding variables. The risk of action tremor was increased in relatives of PD probands, particularly when they had the tremor-dominant form of the disease.
Our observation of higher risk of action tremor in the relatives of PD probands may reflect a genuine increase in the risk of action tremor in the family members of PD probands. Whether the action tremor represents ET or a limited manifestation of PD is not clear. In favor of ET is that we found that the risk of physician-diagnosed ET was higher in relatives of PD probands than in relatives of control probands, suggesting that some of these relatives may have had ET. Also, in other analyses,24 the risk of PD was similar in the relatives of probands with TD-PD and probands with PIGD-PD, which suggests that the increased risk in action tremor that we observed only in the relatives of probands with TD-PD was not the result of an increased risk for PD but might have been caused by an increased risk for ET. In favor of PD is that isolated action tremor that resembles ET has been reported to occur as the only phenotypic manifestation of an underlying PD susceptibility genotype in familial forms of PD,23 so that it is not totally clear whether the reported tremor was ET or just resembled ET.
There are several other potential explanations for the increased risk of action tremor in relatives of PD probands. The proportion with action tremor could have been similar in the two groups, but either 1) control probands and their relatives under-reported the presence of action tremor in their family members or 2) PD probands and their relatives over-reported action tremor. We performed several additional analyses to examine the possibility that our results could be caused by reporting bias. First, we examined data collected from the proband’s report and compared these to data from self-reports. PD probands, all of whom had neurologic symptoms and signs, might have been more likely to report action tremor in their relatives than control subjects. The RR of tremor in relatives of TD-PD probands vs relatives of control probands was similar, whether based on proband report or self-report (1.80 vs 2.24), suggesting that over-reporting by PD probands did not explain the increased risk. Second, based on the relatives who were examined, we found that sensitivity of reporting action tremor was similar for PD and control probands, although we recognize that these numbers were small.
The risk of action tremor was increased only in relatives of probands with TD-PD and not in relatives of probands with PIGD-PD, suggesting that the type of PD in the proband might influence the risk of action tremor in relatives. Alternatively, a shared familial susceptibility for tremor may influence the risk of tremor in relatives and the type of parkinsonism manifested by the proband.
One limitation of this study is that most of the relatives were not examined. To deal with this limitation, we analyzed data, when available, from other informants. In addition, a subsample of relatives was examined, and we were able to demonstrate that the sensitivity of the information provided by patient and control probands was high (92.8%). Second, it is possible that some of the probands with PD, particularly those with PIGD-PD, had other forms of parkinsonism (e.g., multisystem atrophy). However, this form of diagnostic misclassification may have resulted in lower (more conservative) estimates of RR when comparing relatives of PD probands with control probands. Strengths of this study include the large number of relatives whose disease status was ascertained and adjustment for age and other potential confounders in Cox models, which is a strategy that has not been used in previous studies of the familial aggregation of action tremor in PD.
Acknowledgments
Supported by NS36630, RR00645, AG10962, NS39422, NS29993 (National Institutes of Heath) and the Parkinson’s Disease Foundation.
Acknowledgment
The authors thank Richard Mayeux, MD, MSc, Ralph Sacco, MD, MS, Susan Bressman, MD, and Paul Greene, MD, for their assistance in the recruitment of patients for participation.
- Received March 19, 2003.
- Accepted June 18, 2003.
References
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
