Pain is common in patients, suffering from Parkinson’s disease (PD). Painful experiences with various features have been described (muscle cramps, painful dystonia, aching, numbness, tingling, burning, vibrating, and lancinating pains) and a pain classification has been proposed in PD . Based on pathophysiological mechanisms of pain, two kinds of PD-related pain could be recognized: nociceptive pain (musculoskeletal pain due to parkinsonian rigidity and skeletal deformity, dystonic pain due to dystonic posture or dyskinesia) and neuropathic pain resulting from lesions of the Nervous System . The prevalence of pain in PD patients is not known. Few studies have reported an estimation of pain frequency from 30% to 85% [17,18,22,26,32]. These studies were mostly limited due to the lower number of patients involved (n=57–388) and the lack of population control. The aim of this study was to describe the frequency of pain in PD patients using analgesic prescription histories and to compare it with that observed in general population and with two samples of painful patients: diabetics (suffering from neuropathic pain) and osteoarthritis patients (suffering from nociceptive pain). For this purpose, we used a pharmacoepidemiological design allowing (1) the identification of PD patients from antiparkinsonian drugs prescription in a French prescription database and (2) the estimation of pain prevalence using analgesic drugs prescription as an indirect indicator.
2. Materials and methods
Data were obtained from the French Heath Insurance System which covered more than 80% of the French population. The population source was defined by all people covered by the Health Insurance in two French areas: Midi-Pyrénées and Provence-Alpes-Cote-d’Azur – Corse (7,536,000 inhabitants). The database contained patients’ dates of birth, gender, geographical origin and status of Chronic Disabling Disease (CDD), which allowed a total coverage of health expenses. The database also contained all refund prescriptions-only drugs prescribed for each subject and health care covered by the insurance system. Drugs were recorded in terms of brand and drug names, as well as dosages and amounts provided monthly. All data were obtained for the year 2005.
Four populations of patients were compared in this study: PD patients, diabetic patients, patients suffering from osteoarthritis and general population. For each population with specific disease, medications (antiparkinsonian, antidiabetic drugs, and osteoarthritis drugs) were used for identification of these patients.
2.1.1. Definition of parkinsonian cases
At the first stage, we selected all patients older than 34years, who received during the last half-year 2004, prescriptions of medications specifically used in Parkinson’s disease (levodopa, dopamine agonists (pergolide, pramipexole, bromocriptine, ropinirole, and lisuride), anticholinergics (biperidene, trihexyphenidyl, and tropatepine), selective monoamine oxidase B inhibitor (selegiline), and O-methyl transferase inhibitor (entacapone)). During this period of preinclusion, we excluded patients who received neuroleptic drugs and had only one prescription of antiparkinsonian drugs. Patients were definitively included as Parkinsonian patients if they had received at least three prescriptions of antiparkinsonian drugs for 9months (6months of preinclusion plus 3months after inclusion).
2.1.2. Definition of three control populations
For diabetic patients, we selected patients who received a minimum of two prescriptions of antidiabetic medication (insulin or oral hypoglycemic agents [OHAs]) during the last 3months of the preinclusion period (last quarter of 2004). The list of OHAs obtained from this selection included biguanides, sulfonylureas, α-glucosidase inhibitors, thiazolidinediones, and repaglinide.
For osteoarthritis patients, we selected patients who received a minimum of two prescriptions of drugs used for osteoarthritis (Diacerhein, oxaceprol and chondroitin sulphate drugs) during the last 3months of the preinclusion period.
As for the general population, we selected a sample of patients who were not identified by the stages described above (patients without taking antiparkinsonian, antidiabetic or antiarthrosis drugs).
Subjects of the three control populations were excluded if they had received an antiparkinsonian drug during the last 3months of the preinclusion period.
2.2. Analgesic drug prescription
Analgesic drug prescriptions were collected for the year 2005. We investigated two different variables: (1) “pain” (acute and/or chronic) was categorized by at least one prescription of any analgesic drug during 2005; (2) “chronic pain” which was defined as pain persisting for longer than 3months  was identified by at least 90 defined daily doses (DDD) during the years. The DDD system used is validated and based on the WHO Collaborating Centre for Drug Statistics Methodology . One DDD is the drug dose that, on average a person uses daily – 90 DDD would thus theoretically be equivalent to 90days of drug use at standard doses. DDD was chosen as the exposure variable to enable better comparisons with drug classes. The amount of DDD was calculated from the number of deliveries, the number of tablets prescribed and the dose of pills.
We collected analgesic drugs according to the Anatomical Therapeutic and Chemical (ATC) classification index : opioid analgesics (N02A), other analgesic or antipyretic drugs (N02B) and non-steroidal anti-inflammatory drugs [NSAIDs] (M01A).
We also selected some antiepileptic drugs (carbamazepine, oxcarbazepine, clonazepam, and gabapentine) and antidepressants (clomipramine and amitriptyline), which are approved in neuropathic pain in France.
2.3. Statistical analysis
2.3.1. Sample size calculation
All identified PD patients were included in the study. For control groups, we included a number of patients with the same age, gender and geographical origin (urban and rural)-matched with PD population characteristics. Actually, we estimated that we had to include at least the same number of patients in each group to show a relative risk of 2 for the use of analgesic drugs in comparison with the general population, for a sample of at least 3200 PD patients, with a risk α of 0.05 and a power of 0.90.
2.3.2. Data analysis
Descriptive statistics were used for the demographic characteristics of the sample. Proportions were compared using the approximate χ2 test, or Fisher’s exact test when necessary. The t-test was applied when groups were compared in terms of continuous variables provided that they were fairly normally distributed. A p-value of less than 0.05 was considered statistically significant. Association between the use of analgesics and groups of patients were explored in a model of logistic regression adjusted on matching factor (age, gender, and geographical origin) and CDD. Adjusted odd ratios were shown with their 95% confidence intervals. A statistical analysis was performed with SAS®, version 9.1.
The extraction from the French Health Insurance System database was made during the second trimester of 2006 in an anonymous way. Processing of data was possible because the Decision No. 89-117 of the Commission Nationale de l’Informatique et des Libertés (CNIL), according to the French law on privacy.
3.1. Patients (Table 1)
We have identified 11,466 PD patients (let a prevalence of 2.33/1000). The mean age was 77years (range: 35–103) and the sex ratio M/F was 0.8. Sixty-eight percent of PD patients had not seen a neurologist during 2005 and were only treated by general practitioners.
11 200 subjects of the general population, 11 459 diabetics and 11 329 patients with osteoarthritis were included. All sociodemographic characteristics were similar in PD patients and control groups. Most of PD and diabetic patients were covered in the context of CDD. By contrast, this status concerned less patients in the general population and patients with osteoarthritis.
3.2. Analgesic drug prescription (acute and/or chronic) (Table 2)
PD patients significantly received more prescriptions of analgesics than the general population (82% versus 77%, p<0.0001) and fewer than patients with osteoarthritis (82% versus 90%, p<0.0001). There was no significant difference in analgesic drug prescription between PD and diabetic patients.
These results were confirmed by the multivariate logistic regression model adjusted on age, gender and chronic disease. After adjustment, prescriptions of analgesics were more frequently observed in PD patients than in general population (ORAdjusted=1.22, CI95%=[1.10,1.36]), less than in osteoarthritis patients (ORAdjusted=0.50, CI95%=[0.44,0.57]), and similar to diabetics (In MP: ORAdjusted=1.05, CI95%=[0.95,1.16]). Prescriptions of specific analgesics (opiates and other analgesics such as acetaminophen) in PD patients were higher than in the general population, similar to diabetics and less important than in patients suffering from osteoarthritis. Osteoarthritis patients were the most important users of NSAIDs with half of them (53%), who received at least one prescription in 2005. PD patients used up the least of NSAIDs. Prescriptions of other analgesics (antidepressant and antiepileptics) were significantly more observed in PD patients than in all groups of control.
3.3. Chronic analgesic drug prescription (Table 3)
Chronic analgesic drug prescription (estimated by the prescription of at least 90DDD/person/year) was significantly higher in PD patients (33%) than in general population (20.0%) and diabetic patients (26%) (p<0.001). The proportion of patients receiving at least 90 DDD/person/year of analgesics was the same in PD and osteoarthritis patients.
These results were confirmed by the estimated exposure to each class of analgesics from the DDD calculation. PD patients were more exposed than the general population to opiates, acetaminophen, antiepileptics and antidepressants chronic use.
Compared to diabetic patients, they were more frequently chronic users of opiates and acetaminophen (even if the differences were less important than with the general population) and antiepileptic and antidepressant drugs.
Compared to osteoarthritis patients, chronic use of acetaminophen and opiates was similar, but the chronic exposure of antiepileptic and antidepressant drugs was higher whereas that of NSAIDs was lower.
This study represents the largest sample assessing pain in PD. Pain estimated by analgesic drug prescription is more prevalent in PD patients than in the general population. PD patients chronically use opiates, acetaminophen and drugs treating neuropathic pain such as antidepressants and antiepileptics.
Previous studies used drug prescription to estimate the prevalence of diabetes in Europe [13,29] and other chronic conditions like tuberculosis and epilepsy [27,30]. In PD, there are many studies which had also used drug prescription data to estimate disease prevalence or incidence at regional or national levels and the obtained rate was comparable with medical record estimates [3,8,35].
Thus, drug prescription can be considered as a reliable tool for an estimation of prevalence of chronic diseases or symptoms such as pain.
One of the major interests of this study was that our PD patients were not a part of a selected sample but constituted the whole prevalent population of Southern France whoever followed them (a general practitioner or a specialist). A recent study of the French national health fund has reported a close prevalence rate of PD (2.41/1000) in France .
A recent work has reported a pain frequency of 85% with a PD-related pain in 62.6% of patients . In our study, the pain prevalence in PD estimated by analgesic drug prescription was 82%. In this study, we were not able to differentiate pain related and unrelated to PD. Few studies have shown that nociceptive threshold was lower in PD in comparison with healthy volunteers which meant an increase of pain perception [5,12,16]. Thus, it is expected that these PD patients have all kinds of pain (related or not related to PD). We focused on prevalence of chronic analgesic drug prescription which could reflect “chronic pain” recognised as the most disabling and affecting psychological health as well as social and economic well-being [20,25]. Previous studies of pain in PD did not provide epidemiologic findings about chronic pain. Our estimated prevalence of chronic analgesic drug prescription in PD was 33%, which was greater than that in the general population (20%). In the literature, the rates of chronic pain in the general population ranged from 2% to 47% [1,6,14,36]. However, a recent French postal survey has also reported a chronic pain prevalence of 20% .
When compared to the prevalence of pain in two other patient groups with much pain (osteoarthritis patients: 32% and diabetics: 26%), our findings indicated that PD patients experienced more pain than diabetic patients and as much as osteoarthritis patients. Our prevalence ratios of chronic pain in the two painful control populations were lower than those reported in other studies ranging from 60% to 91% for osteoarthritis and from 25% to 60% for diabetic patients [7,9,11,23]. These discrepancies would probably result from the different methods of prevalence assessment. Most of the previous studies used self-report questionnaires for estimating prevalence rate of chronic pain and information on chronic medical conditions was not verified. Thus, recall bias could be present with regards to the participants’ reports of pain duration.
This study is the first to provide detailed information on the use of analgesics in PD patients and to compare it with the general population and other populations. Acetaminophen and opioids were the most commonly prescribed classes of drugs in PD. For instance, 16% of PD patients were chronically prescribed acetaminophen and 11% were prescribed opiates (for a duration ≥3months). Other analgesics like antidepressants and antiepileptic drugs were used with substantially higher frequency in PD patients in comparison with the three control populations. This might indicate that the prescribing doctors have suspected that they have neuropathic pain because these drugs are the most frequently used drugs for neuropathic pain . Several anatomical and pharmacological findings have suggested that the occurrence of neuropathic painful symptoms in PD could be in part due to central modification of nociception and that basal ganglia damage with dopaminergic deficit would be expected to eliminate the inhibitory influence on thalamic nociceptive activity [5,10]. One of the limitations of this study is that these other analgesics may be used not only in the treatment of pain but also in other indications like depression or epilepsy. Even if, at present, the most prescribed antidepressants are serotonin-specific reuptake inhibitors [28,34], the tricyclic antidepressants also represented an indicator of depression that was frequently encountered in PD . So, we could not exclude that depression explained the highest consumption of antidepressants in PD.
Some limitations are worth considering. One is the fact that we did not clinically validate the diagnosis of PD. We could have included some patients with parkinsonism who might not have had true PD, for example, patients with atypical parkinsonism (Multiple System Atrophy and Progressive Supranuclear Palsy) .
Another limitation of this approach is the difference between drug reimbursements prescriptions and the reality of drug consumption. Drug prescription does not always mean that patients take the medications. This could lead to an overestimation of analgesics consumption. Conversely, we do not take into account OTC medication of analgesics (especially acetaminophen or ibuprofen) resulting in an underestimation.
In conclusion, PD should now be considered as a painful disease at least, as much as osteoarthritis and diabetes.
This work was supported by France Parkinson Association. We have no conflicts of interest.
 Blyth FM, March LM, Brnabic AJ, Jorm LR, Williamson M, Cousins MJ. Chronic pain in Australia: a prevalence study. Pain
 Bouhassira D, Lanteri-Minet M, Attal N, Laurent B, Touboul C. Prevalence of chronic pain with neuropathic characteristics in the general population. Pain 2007. Available at: doi:10.1016/j.pain.2007.08.013
 Brandt-Christensen M, Kvist K, Nilsson FM, Andersen PK, Kessing LV. Use of antiparkinsonian drugs in Denmark: results from a nationwide pharmacoepidemiological study. Mov Disord
 Brandt-Christensen M, Lopez AG, Nilsson FM, Andersen PK, Kessing LV. Depressive disorders and anti-parkinson drug treatment: a case register study. Acta Psychiatr Scand
 Brefel-Courbon C, Payoux P, Thalamas C, Ory F, Quelven I, Chollet F, et al. Effect of levodopa on pain threshold in Parkinson’s disease: a clinical and positron emission tomography study. Mov Disord
 Catala E, Reig E, Artes M, Aliaga L, Lopez JS, Segu JL. Prevalence of pain in the Spanish population: telephone survey in 5000 homes. Eur J Pain
 Chan AW, MacFarlane IA, Bowsher DR, Wells JC, Bessex C, Griffiths K. Chronic pain in patients with diabetes mellitus: comparison with a non-diabetic population. Pain Clinic
 Chiò A, Magnani C, Schiffer D. Prevalence of Parkinson’s disease in Northwestern Italy: comparison of tracer methodology and clinical ascertainment of cases. Mov Disord
 Chopra A, Saluja M, Patil J, Tandale HS. Pain and disability, perceptions and beliefs of a rural Indian population: A WHO-ILAR COPCORD study. WHO-international league of associations for rheumatology. community oriented program for control of rheumatic diseases. J Rheumatol
 Chudler EH, Dong WK. The role of the basal ganglia in nociception and pain. Pain
 Clark JA, Spiro A 3rd, Fincke G, Miller DR, Kazis LE. Symptom severity of osteoarthritis of the knee: a patient-based measure developed in the Veterans health study. J Gerontol A Biol Sci Med Sci
 Djaldetti R, Shifrin A, Rogowski Z, Sprecher E, Melamed E, Yarnitsky D. Quantitative measurement of pain sensation in patients with Parkinson disease. Neurology
 Duarte-Ramos F, Cabrita J. Using a pharmaco-epidemiological approach to estimate diabetes type 2 prevalence in Portugal. Pharmacoepidemiol Drug Safety
 Elliott AM, Smith BH, Penny KI, Smith WC, Chambers WA. The epidemiology of chronic pain in the community. Lancet
 Ford B. Pain in Parkinson’s disease. Clin Neurosci
 Gerdelat-Mas A, Simonetta-Moreau M, Thalamas C, Ory-Magne F, Slaoui T, Rascol O, Brefel-Courbon C. Levodopa raises objective pain threshold in Parkinson’s disease: a RIII reflex study. J Neurol Neurosurg Psychiatry
 Giuffrida R, Vingerhoets FJ, Bogousslavsky J, Ghika J. Pain in Parkinson’s disease. Rev Neurol
 Goetz CG, Tanner CM, Levy M, Wilson RS, Garron DC. Pain in parkinson’s disease. Mov Disord
 Guidelines for ATC Classification and DDD Assignment. Oslo, Norway: WHO Collaborating Centre for Drug Statistics Methodology. Available at: http://www.whocc.no/atcddd/
[accessed December 2006].
 Gureje O, Von Korff M, Simon GE, Gater R. Persistent pain and well-being: a WHO study in primary care. JAMA
 International Association for the Study of Pain. Classification of chronic pain. Pain 1986;3:S1–226.
 Koller WC. Sensory symptoms in Parkinson’s disease. Neurology
 Krein SL, Heisler M, Piette JD, Makki F, Kerr EA. The effect of chronic pain on diabetes patients’ self-management. Diabetes Care
 Lajugie D, Bertin N, Chantelou ML, Vallier N, Weill A, Fender P, et al. The prevalence of Parkinson’s disease and its cost to the French national health fund in metropolitan France in 2000. Rev Med Ass Maladie
 Latham J, Davis BD. The socioeconomic impact of chronic pain. Disabil Rehabil
 Lee MA, Walker RW, Hildreth TJ, Prentice WM. A survey of pain in idiopathic Parkinson’s disease. J Pain Symptom Manage
 Maggini M, Salmaso S, Alegiani SS, Caffari B, Raschetti R. Epidemiological use of drug prescriptions as markers of disease frequency: an Italian experience. J Clin Epidemiol
 Montagnier D, Barberger-Gateau P, Jacqmin-Gadda H, Dartigues JF, Rainfray M, Peres K, et al. Evolution of prevalence of depressive symptoms and antidepressant use between 1988 and 1999 in a large sample of older French people: results from the Personnes Agees Quid study. J Am Geriatr Soc
 Papoz L. Utilization of drug sales data for the epidemiology of chronic diseases: the example of diabetes. The EURODIAB Subarea C Study Group. Epidemiology
 Rochat P, Hallas J, Gaist D, Friis ML. Antiepileptic drug utilization: a Danish prescription database analysis. Acta Neurol Scand
 Sindrup HJ, Jensen TS. Efficacy of pharmacological treatment of neuropathic pain: an update and effect related to mechanism of drug action. Pain
 Snider SR, Fahn S, Isgreen WP, Cote LJ. Primary sensory symptoms in parkinsonism. Neurology
 Tolosa E, Wenning G, Poewe W. The diagnosis of Parkinson’s disease. Lancet Neurol
 Ubeda A, Cardo E, Selles N, Broseta R, Trillo JL, Fernandez-Llimos F. Antidepressant utilization in primary care in a Spanish region: Impact of generic and reference-based pricing policy (2000–2004). Soc Psychiatry Psychiatr Epidemiol
 Van de Vijver DA, Roos RA, Jansen PA, Porsius AJ, de Boer A. Estimation of incidence and prevalence of Parkinson’s disease in the elderly using pharmacy records. Pharmacoepidemiol Drug Safety
 Verhaak PF, Kerssens JJ, Dekker J, Sorbi MJ, Bensing JM. Prevalence of chronic benign pain disorder among adults: a review of the literature. Pain