In this report we present 5 cases of individuals with obsessive-compulsive disorder (OCD) whose symptoms developed late in life. Epidemiological studies indicate the mean age at onset of idiopathic OCD is between 20 and 25 years and that only 15% of cases present after the age of 35.1 Onset of OCD after age 50 is relatively unusual, and information regarding these cases is limited to case reports.2,3 Given its relative rarity, cases of late-onset OCD may be related to structural cerebral damage and may therefore provide further insight into the neurobiology of this disorder.
Theories regarding the cause of OCD have changed remarkably over the years, the earliest writings having been focused on repetitive, blasphemous thoughts believed to be the result of demonic possession. Although Jean Etienne Dominique Esquirol, in an 1838 report describing the case of "Mademoiselle F.," was the first psychiatrist to describe the phenomenon,4 Pierre Janet is often credited with providing the first rational psychological explanation.5 In his 1903 paper "Les obsessions et la psychasthenie," Janet emphasized the role of psychic fatigue as being responsible for a lack of control over one's thoughts.6 Within the decade, Freud's case of obsessional neurosis, the "Rat Man," led to the concept that OCD resulted from unconscious conflicts.7 Alongside this psychodynamic conceptualization, behaviorists began to apply learning theory to the symptoms of OCD, viewing the obsessions and compulsions as conditioned responses that lower anxiety.8
The modern concept of OCD as a brain-based illness developed from three main branches of evidence: 1) cases of "secondary" OCD accompanying or following systemic or neurologic illness; 2) functional neuroimaging studies that highlighted differences in regional blood flow in cases of idiopathic OCD when compared with normal control subjects; and 3) the efficacy of some neurosurgical interventions in cases of refractory OCD. Several disorders have been linked with OCD, including von Economo's encephalitis,9—11 Sydenham's chorea,12—14 and Tourette's syndrome.15,16 Focal cerebral lesions, most commonly involving the frontal lobes and/or the basal ganglia, have also been connected with OCD.2,3,17—20 These cases, though uncommon, have generated interest in the possible neuropathology of "idiopathic" OCD.
Further evidence for the importance of the frontal lobes and striatum in the pathogenesis of OCD has come from functional neuroimaging studies.21—23 These studies reveal increased metabolic activity in the frontal cortex and striatum of patients with OCD. Activity in these brain territories increases with symptom provocation22 and normalizes with appropriate treatment.23 This type of neuroimaging research has highlighted the role of corticostriatal circuits in the development of OCD.24—27 Interruption of (presumably) overactive frontostriatal circuits is hypothesized as the basis for the efficacy of neurosurgical techniques in these patients.28—31
We feel the following 5 cases are of interest because of the relative rarity of late-onset OCD. In fact, they represent the only cases of late-onset OCD identified by the authors.
Case 1. Ms. A., a 70-year-old married white woman, first developed obsessive-compulsive symptoms at the age of 53, which prompted her to seek psychiatric treatment. Her symptoms included obsessions that "something terrible" (e.g., a fire or flood in her home) would happen while she was sleeping. These fears led to checking rituals, usually, but not solely, before bedtime. She felt compelled to check the light switch, turning it on and off several times until she would "put it out right." She repeatedly checked the stove, the sinks and drains, and her pillow, which she feared might be touching the bedside lamp. These fears and rituals disrupted her life and frequently prevented her from falling asleep for hours. Despite trials of fluoxetine, alprazolam, and triazolam, she had not gained relief. She presented for psychopharmacologic consultation.
Her Yale-Brown Obsessive-Compulsive Scale (Y-BOCS)32 score was 24 (scale 0—40), and her score on the Maudsley Obsessive-Compulsive Inventory (MOCI)33 was 9 (scale 0—30). Although she had a history of a prior depressive episode, she was not currently depressed. Her Beck Depression Inventory (BDI)34 score was 16 (scale 0—63). Her medical history was notable only for cirrhosis of the liver secondary to chronic hepatitis C. She did not use alcohol or tobacco. Her family history was significant for major depression in one sister and schizophrenia in one of her three sons. Non—contrast-enhanced cranial CT demonstrated diffuse, mild cortical atrophy. A focal area of low attenuation was also noted in the head of the left caudate, consistent with a lacunar infarct of uncertain age.
The patient continued to have OCD symptoms and was able to tolerate only low-dose (0.25 mg) clonazepam. She died at the age of 75 secondary to complications of hepatic failure.
Case 2. Ms. B., a 56-year-old married white woman, first developed obsessive-compulsive symptoms (contamination fears) at the age of 50. At that time, a child with cytomegalovirus (CMV) enrolled in the daycare center where she worked, and Ms. B. became overwhelmed with the fear that she might somehow pass this infection on to her husband, who was chronically ill. Her worries became so severe that she quit her job and obtained work at a convenience store. While there, she shared a cigarette with a co-worker who had recently been tested for human immunodeficiency virus (HIV) status. Ms. B. became morbidly afraid that she might have contracted HIV, a fear she realized was unreasonable. She sought psychiatric help and was started on clomipramine, with resolution of her symptoms.
She did well for approximately 4 years, and the clomipramine was discontinued. At age 54, following an episode of marital infidelity, she again became obsessed with HIV contamination fears. She worried that people might have residual semen on their hands; touching objects that had been handled by others became anxiety-provoking for her. She found it especially difficult to handle money, and she began to wash her hands excessively. Despite restarting clomipramine and clonazepam, her symptoms persisted. Her Y-BOCS score was 22 and her MOCI score was 15.
Ms. B. had hypothyroidism, controlled with oral hormone replacement, and narrow-angle glaucoma. Recurrent syncopal spells with poor coordination and unclear visual changes had led to a neurological evaluation. No specific diagnosis was made. She had a 20-pack-a-year smoking history but did not use alcohol. Her father and one brother had a history of major depression.
A CT scan of the head, done as part of the syncopal workup, revealed abnormal T2 hyperintensities in the bilateral frontal periventricular white matter adjacent to the frontal horns. A hyperintense lesion was also noted in the central pons.
Case 3. Mr. C., a 75-year-old widowed white man, first developed obsessive-compulsive symptoms related to feelings of guilt over marital infidelity at the age of 52. At 75, he was seen in psychiatric consultation for symptoms of anxiety while hospitalized for three-vessel coronary artery bypass graft surgery. Mr. C.'s symptoms included excessive worries regarding infection and contamination. He ruminated about his current girlfriend's risk of contracting HIV from touching the hand of a homosexual door attendant (of unknown HIV status) at his apartment complex. During the interview, Mr. C. inquired several times whether this was a likely possibility. He ruminated about a variety of medical illnesses (including infectious diseases) and often sought reassurance from doctors. Since his symptoms had begun, he had been washing his hands approximately 20 times per day to deal with fears of contamination and infection. Although he was unsure of his prior medication trials, an ongoing trial of fluvoxamine (150 mg daily) had led to minimal benefit.
His medical history was significant for hypertension as well as coronary and bilateral carotid artery disease. He had not smoked in over 45 years and denied alcohol use. His family history was negative for mental illness. Neurologic exam and bedside cognitive testing were unremarkable. A CT scan of the head done 2 years prior to admission revealed bilateral hypodense areas in the caudate nuclei, consistent with lacunar infarcts of unclear age.
Recommendations were made to increase the fluvoxamine dose and to add a small dose of lorazepam. The patient was discharged to outpatient psychiatric care.
Case 4 (previously published3). Mr. D., a 70-year-old married white man, first developed obsessive-compulsive symptoms at the age of 62. He specifically recalled the exact date and time of symptom onset: it had occurred when he recognized an actor in a television commercial but was unable to recall his name. He became quite anxious and "needed to know"; eventually he called several acquaintances and the television network. This "need to know" worsened, and it became common for him to wake his wife or phone friends in the middle of the night to obtain answers to questions that plagued him. He memorized bumper stickers, newspaper headlines, names of songs and their composers, and even comic strip character names. He started to collect old newspapers to create a database to answer questions that might come up. With time, however, Mr. D. was no longer able to read the newspaper because of a strong urge to memorize all of the headlines. He became unable to attend church because he felt compelled to look up all of the references during the sermon. Trials of nearly 35 medications from several different classes provided little to no relief.
Mr. D. had a history of hypertension, diabetes mellitus, and carotid artery disease (total occlusion of the right internal carotid artery with 30% to 40% stenosis on the left). This latter finding was discovered during a workup of recurrent brief paresthesias involving the left arm and face, eventually thought to be due to sensory seizures.
Neurologic exam revealed rare paraphasic errors in spontaneous speech, a mild constructional apraxia, and mild difficulty with Luria sequencing bilaterally. There was generalized hyporeflexia, with absent ankle jerks bilaterally. There was also evidence of bilateral sensory loss in a stocking distribution. A CT scan of the head revealed a 3-by-4-cm wedge-shaped infarct in the posterior right frontal lobe, with extension to the deep subcortical white matter. A [99mTc] HMPAO SPECT scan revealed hypoperfusion in the area of the infarct, but no other notable abnormalities.
Mr. D. agreed only to an increased dose of clonazepam, with little effect on his obsessive symptoms. However, the paresthesias disappeared, raising the possibility that they were sensory seizures arising from the area of damaged frontal cortex.
Case 5. Mr. E., a 75-year-old married white man, first developed obsessive symptoms at the age of 71; he ruminated about his tongue. For 4 years he had been plagued with near-daily worries about his tongue; feeling that it was "not right," or "too tight," and complaining of intolerable tongue pain. Despite reassurances from several dentists and otolaryngologists that no abnormalities were present, his worries persisted. Although Mr. E. described himself as "perfectionistic" and "a worrier," both he and his wife noted these obsessive symptoms to be excessive and unusual even for him. Trials of lorazepam and diazepam were minimally helpful. He had had two successful series of ECT over the preceding year for treatment of depression; neither had any significant effect on his obsessions. He presented for psychiatric consultation after his wife saw a television report on OCD.
Mr. E. was medically healthy and had no history of substance abuse. He had a history of recurrent depression, with onset at age 57, but he did not have depressive symptoms at the time of his consultation. His family history was notable for major depression in his father, alcoholism in one son, and panic disorder complicated by alcoholism in one daughter. A nephew had died by suicide.
He was a thin man who described himself as an avid gardener. His physical and neurological exams were normal, particularly with reference to the tongue. His Mini-Mental State Examination score was 29/30, with a single error in recall. A CT of the head done at age 73 was described as "within normal limits," but the report itself is not available. An EEG done at the same time was also normal. His Y-BOCS score was 32. His score on the BDI was 14.
In addition to recurrent major depression, he was felt to have OCD or body dysmorphic disorder (i.e., monosymptomatic hypochondriasis). Although psychotic depression or a delusional disorder with somatic features was also possible, his insight into the excessiveness of these concerns and the lack of response to ECT argued against these diagnoses. He has been lost to follow-up.
To our knowledge, this series of 5 cases of late-onset OCD represents the largest such report in the literature. Although a recent study35 suggests that onset of OCD after age 50 might not be as rare as once believed, particularly in women, the scarcity of published information on these patients and our own experience seem to argue against this point.
Because OCD is an illness that usually presents in the second or third decade of life, onset after age 50 should alert the physician to possible "organic" causes of OC symptomatology. Interestingly, 4 of the 5 cases showed evidence of specific cerebral abnormalities on CT scan. Three of the 5 patients had previously seen a neurologist.
The majority of these patients had a relatively poor outcome, with limited response to psychotropic medication. Because of the small number of cases, we cannot comment on the generalizability of this finding to all cases of late-onset OCD (or cases of OCD with intracerebral pathology).
Although the cerebral abnormalities may have been incidental findings, their location in the frontal lobes and caudate nuclei is consistent with several prior reports of secondary OCD related to Sydenham's chorea,12—14 Huntington's disease,36 and focal cerebral lesions.2,3,17—20 It is also consistent with the prevailing theory of OCD pathogenesis, which implicates abnormal activity in the corticostriatal circuit.24—27 It is unclear, however, why some lesions in this circuit cause OCD while other lesions (e.g., cingulotomy) alleviate it. This question is the focus of ongoing functional neuroimaging research.