Cognitive Behavioral Therapy in Elderly Populations
Cook  implemented a cognitive behavioral pain management program for elderly nursing home residents with chronic pain. Specifically, a randomized group design was used comparing a CBT intervention to an attention/support control intervention. Both interventions were relayed in weekly group sessions lasting 10 weeks. Participants were 13 men and 9 women from two nursing homes, ranging in age from 61 to 98 (M = 72). While the programs were perceived as equally credible both before and after intervention, results found that individuals who received CBT reported less pain and pain-related disability. More treatment effects were maintained at 4-month follow-up, despite an overall increase in reported pain. There were no significant treatment effects for depression or medication ratings. Similarly, Reid et al.  administered CBT to largely female (86%) senior housing residents with chronic pain aged 65 or older (M = 77.4). CBT was administered in 10 weekly individual sessions, with participants phoned 5 days on average after each session to gauge comprehension, usefulness of materials, and adherence to assigned homework exercises. Comprehension of CBT exceeded 97% and mean rating for usefulness of treatment sessions ranged from 7.5 to 9.4. The mean number of days per week that individuals completed homework exercises ranged from 1.8 to 4.0. At 2 weeks, posttreatment versus pretreatment assessment  found significant reductions in pain intensity and pain-related disability scores. While treatment effects were noted to decrease over time, they did not return to pretreatment levels as of 24 weeks.
Morone et al.  looked at a mindfulness-based intervention for chronic pain in the elderly. Community-dwelling adults aged 65 and older (M = 74.9) were randomized to either an 8-week mindfulness-based meditation group or a wait-list control group. Overall, 37 participants were randomized within a 6-month period. The intervention was well tolerated as the study evidenced a high overall completion rate with 68% (13/19) of the intervention group and 78% (14/18) of the wait-list control group after they crossed over to the intervention finishing all aspects of the intervention. Of the 13 participants in the intervention group who completed the 8-week program, 12 were available for 3-month follow-up with no significant difference found in 8-week versus 2-month scores. Participants were found to have meditated an average of 4.3 days/week for an estimated 31.6 minutes/day. Compared to the wait-list control group, the mindfulness group showed significant improvement in pain acceptance, activities engagement, and physical function. Further speaking to long-standing benefit and ease of maintenance, most participants continued to meditate of their own accord at 3-month follow-up.
Arena et al.  evaluated a self-regulatory intervention in the elderly. Here the effects of a 12-session frontal electromyographic biofeedback training regimen on headache in eight elderly tension headache sufferers (N = 8, M = 65 years) was studied. Biofeedback sessions were modified slightly in an effort to increase comprehension, learning, and recall of purpose and instructions. Results of a 3-month posttreatment assessment showed significant decreases in overall headache activity in 50% or greater of the subjects with moderate improvement (35–45%) in three of the four remaining subjects. Significant pre–post difference was also found for number of headache free days, peak headache activity, and medication index.
A prior study by Arena et al.  applied relaxation therapy to tension headache in the elderly (N = 10, M = 69 years). Subjects received an 8-week progressive muscle-relaxation therapy regimen. Similar to Arena et al. , results posttreatment showed significant reductions in overall headache activity (50% or greater) in seven subjects. Significant pre–post differences were also seen for number of headache-free days, peak headache activity, and medication index.
Meta-Analytic Studies Reviewing Cognitive and Behavioral Interventions
Lunde et al.  conducted a meta-analysis reviewing cognitive and behavioral interventions for chronic pain in the elderly. Overall, Lunde et al.  found that cognitive and behavioral interventions (including CBT, mindfulness, and self-regulatory interventions) produce improvement in self-reported pain experience in the elderly with an overall mean effect size for treatment versus pretreatment or control condition of 0.47 at posttreatment and 0.56 at follow-up. Effect sizes were consistent with the Morley et al.  meta-analysis of cognitive and behavioral treatment of chronic pain in adults. Further, meta-analyses for psychological treatments of depression and anxiety in the elderly [52, 61] find overall effect sizes of 0.55 and 0.78. Therefore, in applying nonpharmacological interventions in older persons, overall effect sizes for pain are consistent with effects obtained in younger cohorts and for more traditional psychological disorders.
PSYCHOLOGICAL TREATMENT OF PAIN IN COGNITIVELY IMPAIRED POPULATIONS
As noted above, pain is highly comorbid with dementia and consistently undertreated. Research indicates that both pharmacological and nonpharmacological interventions are underutilized in cognitively impaired populations. In the context of a progressive process whereby ability to effectively communicate needs deteriorates, treatment opportunity is lost when behavioral symptoms are unnoticed, dismissed, or not understood and acknowledged as arising from physiological or psychosocial unmet needs .
Individuals with dementia report pain less often, less spontaneously, and at a lower intensity than their cognitively intact counterparts with ability to verbally communicate pain or discomfort generally decreasing as dementia severity increases [7, 42, 56]. For those who are cognitively impaired, behaviors such as vocalizations (e.g., sighing, moaning, calling out, verbal abuse, repetitive vocalizations, noisy breathing), facial expressions (e.g., grimacing or frowning), restless or strained body expressions (e.g., fidgeting, increased pacing, or rocking), or agitation and resistance to care may represent the most prominent or the only feature of pain. Unfortunately, it is often the case that such symptoms are disregarded or interpreted as characteristic of dementia rather than recognized as a symptom of pain . Recognition of pain when information is less readily available to caregivers calls for more thorough evaluation, this then potentiating opportunities for intervention, both in the treatment of pain and behavioral symptoms. However, in the absence of comprehensive assessment, individuals with dementia are unfortunately more likely to receive psychotropic medications rather than appropriate pain treatment [7, 42, 56].
In light of the possibility of pain serving as a causative factor in the presentation of BPSD, it is important that the control of pain in those suffering from dementia is clear. Research demonstrating the efficacy of pain treatment in reducing behavioral symptoms in dementia has largely focused on pharmacological intervention [10, 33, 34]. As an example, Chibnall et al.  investigated the effect of regularly scheduled analgesic medication on behavior, emotional well-being, and use of psychotropic medication. Here, 25 nursing home residents with moderate to severe dementia received 4 weeks of treatment with acetaminophen and 4 weeks of a placebo. Results showed that those receiving acetaminophen spent more time in social interaction, engaged with media, talking to themselves, engaged in work-like activity, and experiencing unattended distress. They also spent less time in their rooms, less time removed from the unit, and less time in personal care activities. However, despite increased activity and socialization, no effect was seen for agitation, emotional well-being, or need for psychotropic medication.
In contrast to primary pharmacological intervention, there is increasing research available showing the efficacy of psychosocial and behavioral interventions in the reduction of neuropsychiatric symptoms manifesting in behavioral symptoms [5, 8, 44–46]. Ayalon et al.  conducted an investigation of the available research on the effect of nonpharmacologic interventions for neuropsychiatric symptoms among persons with dementia. Three RCTs and six single case designs (SCDs) met criteria and were included in the review. Intervention type included unmet needs interventions (one SCD), behavioral interventions (four SCDs), caregiving interventions (three RCTs), and bright-light therapy (one SCD). The results of the review suggested variable efficacy dependent on the type of intervention used. Approaches that utilized an unmet needs intervention that focused on assessing the motivation driving neuropsychiatric symptoms and designing an intervention to reduce those symptoms or their intensity found moderate reductions in problem behaviors. Comparatively, behavioral interventions treating neuropsychiatric symptoms via contingency management (e.g., reward removal) were found to result in a relative reduction of 50–100% in neuropsychiatric symptoms. Examination of the above approaches when taught to caregivers yielded mixed results with one RCT finding a reduction in four neuropsychiatric symptoms subscales including ideation disturbance, irritability, verbal agitation, and physical aggression at 6-month follow-up, a second RCT finding significant improvement in frequency and severity of intervention-associated target behaviors, and a third finding no effect. Finally, examination of available studies of bright-light therapy where exposure to direct bright light is provided to engender a calming effect on an agitated patient found short-lived improvements in agitated behavior. Overall, findings from this review suggest that nonpharmacological interventions targeting neuropsychiatric symptoms in persons with dementia may be effective.
Despite these inroads in understanding nonpharmacological intervention, particularly those based on the biopsychosocial model, there appears a disconnect in treatment approach and lack of understanding between the relationship between pain and behavioral symptoms, how they co-occur, and which interventions are effective at lessening both pain and behavioral symptoms . The bridging of this disconnect then, as described by Piper et al. , may lie in the appropriate modification of psychological treatments to meet the needs of a cognitively impaired population. Although limited research is available to support such modifications with regard to pain, a relatively larger body of literature exists supporting their modification in the treatment of neuropsychiatric symptoms including depression and anxiety.
As an example of this approach Kraus et al.  discuss a modified version of CBT for anxiety in dementia (CBT-AD), this protocol developed, piloted, and modified over two years with seven mildly demented participants . Modifications were made to the content, structure, and learning strategies of CBT in an effort to adapt to the cognitive limitations of the patients, for example, adapting simplified checklists such that they mainly required recognition skills. Psychoeducation and skills including diaphragmatic breathing, coping self-statements, exposure, and behavioral activation were simplified. For example, breathing was introduced with simplified procedures including breathing deeply and slowly. This contrasted with a traditional multistep diaphragmatic breathing technique. Collaterals (e.g., caregivers such as spouses and children) were trained as coaches (e.g., encouraging participants to breathe slowly and praising effort). Practice was done daily and encouraged in the context of distress. In later sessions participants were encouraged to practice when they had difficulty sleeping at night. This was reinforced with prompting and practicing appropriate responses. Furthermore, sessions were shortened to accommodate for attention and fatigue and were limited to the teaching of one or two skills. In these sessions patients were encouraged to repeat information to facilitate learning and to actively participate in the creating of reminder cues serving retrieval. Finally, therapists relied on spaced retrieval, an evidence-based method for improving learning and retrieval, relying on procedural memory. In their protocol, patients and their collaterals participated in nine sessions over 10 weeks, with sessions lasting 30–45 minutes. The first sessions focused on monitoring and deep breathing with further sessions focusing on the use of coping self-statements (fourth) and behavioral activation (seventh). During each session, participants were provided with simplified written instructions and a concrete homework plan to be completed. In addition, brief phone calls with the patient occurred once a week between sessions to test for comprehension and difficulties using the skills learned in the previous session. These phone calls also encouraged practice. Posttreatment, during a subsequent 3-month period weekly and then bi-weekly phone calls were made to continue to facilitate the use of skills. The results of this early pilot suggested that a modified CBT protocol may be advantageous in a dementia population as clinically meaningful reductions in anxiety were found. Patient one’s posttreatment assessment indicated a decrease in anxiety according to the Rating Anxiety in Dementia Scale (RAID; pre-RAID = 19, post-RAID = 10). Collateral informant also