Synonyms

Applied behavior analysis; Behavior modification; Behavior therapy; Behavioral health; Classical conditioning; Learning theory; Operant conditioning

Theoretical Foundations of Behavior Analysis

Behavior analysis involves the systematic application of learning theory to explain why behavior is occurring. Within this model behavior encompasses everything a person does including observable behavior as well as what the person thinks and feels (Ramnero and Torneke 2008). Learning theory posits that an individual’s behavioral repertoire is a product of their genetic and learning histories (Skinner 1938/1991). The significant heterogeneity of psychological functioning within the population of older adults is both predicted and explained by the model’s assumption that each human being is unique by virtue of their idiosyncratic genetic and learning histories. The utility of behavior analysis for explaining behavior in late life and prescribing interventions to promote behavioral health is significant as the majority of evidence-based behavioral health interventions share roots in learning theory.

Within learning theory genetic history encompasses factors that influence the individual’s physiology and is reflected in the individual’s current physiological status (e.g., health status, sensory functioning, physical conditioning) as well as in responses that are natural, biologically driven, and not learned (e.g., salivating at the smell of food, blinking at a bright light). Behavior that is the product of an individual’s learning history encompasses responses that are influenced by environmental experience.

Learning is conceptualized as occurring within two processes: operant conditioning and respondent conditioning. Behavior analysis assumes that to the extent that behavior, including both developmentally “normal” and dysfunctional behaviors, is learned it can be unlearned (Krasner and Ullman 1965). The assumption of behavioral plasticity is the raison d’ etre for the systematic application of learning principles to produce behavior change at any point in the life span (Bijou 1961/1995; Baltes and Barton 1977).

Operant conditioning and functional relations. The unit of analysis within operant conditioning is behavior in context. Context includes both historical and current physiological, cultural, and social conditions. In order to understand why a particular behavior is occurring, a behavior analyst attempts to identify patterns in the conditions and events that surround the behavior, i.e., whether events or stimuli (called antecedents) reliably precede and follow (called consequences) the occurrence of the behavior. The continuous interaction of antecedent, behavior, and consequence over time is called the operant contingency (Skinner 1938/1991). A contingent relationship between antecedents and consequences is said to occur if systematic observation of behavior reveals that an antecedent reliably increases the probability that a behavior will be emitted and the occurrence of the behavior increases the probability that a consequence will be delivered. Antecedents are environmental events or stimuli that are reliably present in the setting in which a behavior occurs. Antecedents can be verbal or nonverbal. Verbal antecedents may be either the verbal behavior of another person or the individual’s own verbal behavior, for instance, “I know how to do this” or “This isn’t safe.” When an antecedent gains the property of signaling the availability of reinforcement for certain behaviors, the behaviors are said to be under stimulus control in that the presence of the antecedent stimulus increases the probability of the behaviors occurring.

Within the operant model, consequences are defined based on their effect on behavior, i.e., whether the consequence results in an increase or decrease in the probability that the behavior will be emitted in the presence of similar antecedents. If it is determined that the probability of a behavior occurring over time has increased, i.e., its frequency increased, the consequence is labeled a “reinforcer”; if it is observed that the probability of the behavior decreased over time, then the consequence is considered a “punisher.” The adjectives “positive” and “negative” distinguish different types of reinforcers: if a behavior results in access to a stimulus (e.g., smiling is reliably followed by a hug from a friend), the reinforcer is considered “positive.” If behavior is followed by a reduction in the quantity of a consequence (e.g., the intensity of painful emotion declines immediately following suicidal ideation; physical pain is reduced following the ingestion of an analgesic), the consequence is considered a “negative” reinforcer. Reinforcement contingencies are identified based on their effect on the quantity of behavior, i.e., whether behavior frequency increases, is stable, or decreases over time.

The identification of the reinforcement contingencies controlling a behavior is a central goal of behavior analysis and directly informs the identification of interventions that would be effective for increasing or decreasing the behavior. Points of intervention to break the contingency may include: (1) altering the antecedent stimuli to prevent or increase the probability of the occurrence of behavior (e.g., removing electronic devices and reading material from a bedroom to promote sleep, placing medicine containers by the coffee pot to increase the probability medication will be taken in the morning as prescribed, camouflaging a door to prevent a person with dementia from trying to exit) or (2) discontinuing the reinforcement of a behavior (known as extinction) (e.g., discontinuing negative reinforcement of escape from emotionally painful private events through in vivo exposure, instructing family members to stop attending to maladaptive “sick” role behaviors), delivering a reinforcer contingent on behavior that is incompatible with a dysfunctional behavior, or both (Ramnero and Torneke 2008; Skinner 1938/1991). Examples of differential reinforcement of behavior include praising a client’s problem solving when emotionally distressed rather than his engaging in escape of avoidance behavior or praising a caregiver of a person with dementia for engaging in empathic verbal responses rather than corrective feedback with their family member.

Within the behavior-analytic model, it is assumed that a behavior may serve more than one function or purpose. For example, complaining about poor health or pain may result in an individual escaping from household tasks (i.e., the behavior is negatively reinforced in that it results in the removal of aversive stimuli) and also being consistently followed by hugs and comforting statements by family members (i.e., the behavior is positively reinforced in that it is consistently followed by social attention). It is important to determine if a behavior serves more than one function in order to design an intervention that will effectively address all functions of the problem behavior.

The function of a behavior is determined through a variety of assessment methods:

  1. (a)

    Experimental functional analysis during which hypothesized controlling variables, specifically the antecedents and consequences, are directly manipulated while the behavior analyst carefully monitors whether there are changes in frequency of the behavior (Skinner 1938/1991; Bijou 1961/1995). Experimental functional analysis is the most direct and accurate method of identifying the function of a behavior.

  2. (b)

    Descriptive analysis or assessment involves less direct, but often more practical methods for generating but not directly testing hypotheses about the function of a behavior. Descriptive assessment can involve a variety of methods, such as direct observation, self-report methods involving interviews, or paper and pencil questionnaires. Descriptive analysis can also include a variety of sources of information from clients or collateral sources, for instance, a client monitoring and recording the environmental and private events that precede and follow the problem behavior, having the client or an informant (e.g., a caregiver) complete a paper and pencil questionnaire designed to assess the contextual variables reliably associated with the behavior, or having an informant directly observe and record the problem behavior and events or stimuli that precede and follow the behavior (Haynes and O’Brien 2000).

Classical Conditioning

Classical or respondent conditioning involves learning by association. Classical conditioning has been applied to explain the development of many physiological and emotional responses including conditioned fear, sexual arousal, trauma-related anxiety, and responses associated with substance abuse. For example, an individual who has experienced a traumatic event may later experience anxiety when they have contact with environmental stimuli, such as the smells, images, or sounds similar to those present during the initial traumatic event.

Classical conditioning provides the theoretical foundation for several evidence-based therapies such as exposure therapy for anxiety. Classical conditioning was first described by the Russian physiologist Ivan Pavlov (1849–1936). In studying the salivary and gastric secretions of dogs, Pavlov would place a bowl of meat powder in front of a dog and measure the amount of secretions naturally produced. During his studies Pavlov observed that over time the dogs would start salivating when he entered the room, regardless of whether meat powder was presented. Pavlov’s serendipitous, yet astute, observation of this phenomenon led to a series of experiments in which he systematically manipulated the presentation of a neutral stimulus, the sound of a bell or tuning fork, prior to presenting meat powder in order to better understand this learning by association. Pavlov observed that by repeatedly pairing the sound, a neutral stimulus, with the presentation of the meat powder, an unconditioned stimulus because it naturally elicited a response, the dogs would start to salivate at the sound of the neutral stimulus, demonstrating that the neutral stimulus had obtained functional properties similar to the food. The bell had become a conditioned stimulus (CS) in that it elicited a learned reaction or conditioned response similar to the natural reaction or unconditioned response of salivating to the food (an unconditioned stimulus).

Behavior analysis of functioning in late life. Behavior-analytic strategies have been applied to promote the health and quality of life of older adults in a variety of ways. Treatment goals have ranged from increasing health-related behaviors such as exercise, sleep, nutrition, and medication adherence to decreasing behaviors that restrict or prevent access to positive and valued outcomes, (for instance, social withdrawal and isolation, substance abuse, and suicidal behaviors). Examples of behavior-analytic conceptualizations of health-interfering and health-promoting behaviors are described below.

Depression. The behavior-analytic model of depression considers the interaction of the person’s repertoire within its historical and current environmental context. The model posits that the risk of depression increases when individuals experience low rates of positive reinforcement, high rates of aversive events (punishment), or both in their lives (Ferster 1973). Further, the model assumes that an individual’s repertoire, which may include high-frequency negative self-statements, low frequency of eye contact or smiling during interactions, or evidence of a low frequency of instrumental problem-solving skills, may further limit their access to positive reinforcement in day-to-day life or increase the probability of experiencing aversive consequences (Ferster 1973; Lewinsohn and Graf 1973). In addition, behaviors commonly exhibited by persons who meet diagnostic criteria for depression, such as excessive sleeping, social withdrawal and isolation, and alcohol consumption, may be negatively reinforced by escape from or avoidance of potentially aversive consequences such as feeling anxious, ashamed, lonely, or rejected (Martell et al. 2001). The low rates of behavior commonly observed when an individual behaves in a manner typically described as “depressed,” in turn, further limit opportunities for the person to contact pleasant experiences that would elevate his mood. More stable behaviors in the person’s repertoire, such as social skill deficits involving low rates of eye contact or smiling and high rates of vocalizing negative and pessimistic statements during social interactions, may increase the likelihood of contact with aversive consequences, which could include negative affect and withdrawal by others. In addition, these social skill deficits could further limit opportunities for contacting pleasant experiences and hence increase the likelihood of continued depressed mood. Behavioral approaches to the treatment of depression typically target both the depressogenic repertoire and contextual factors.

In regard to age-associated considerations, sensory changes, medical conditions that result in chronic pain or fatigue, and medication side effects may reduce or altogether preclude pleasure during activities that had been historically preferred. Further, a large proportion of elderly persons judged to be depressed are prescribed antidepressant medication although over 50% of older adults who meet criteria for major depressive disorder do not respond to first-line treatment with antidepressant medication (Joel et al. 2014). In many cases, pharmacological intervention for depression may be contraindicated as polypharmacy increases the risk of adverse medication effects (American Geriatrics Society 2015). In contrast, evidence-based non-pharmacological treatments for depression including behavioral activation (BA) (Martell et al. 2001) and cognitive behavior therapy (CBT) have been found to be highly effective for the treatment of depression in older adults (Gallagher-Thompson et al. 1990; Ayers et al. 2007) and have no side effects.

BA is designed to improve mood by targeting the individual’s inactivity, avoidance, and withdrawal behaviors. Specifically, the BA treatment process involves three main steps: activity monitoring (monitoring the actions that preclude and follow depressive behaviors), activity scheduling (replacing prior maladaptive behaviors with positive, productive behaviors that increase contact with pleasant consequences), and modifying activities based on client feedback (continually adjusting the treatment plan until the desired outcome is reached) (Martell et al. 2001). During treatment it is important to consider how an individual’s skill repertoire, sensory functioning, and health may influence their experience of activities, for example, individuals with sensory deficits may prefer activities that involve fewer sensory challenges. It is also important to consider an individual’s socialization history when identifying potentially gratifying experiences. For example, an individual with a history of enjoying solitary activity may find physical activities more rewarding than those involving social interaction. Further, while it is commonly suggested that depressed clients increase social contact by attending community or other organized events, socializing with strangers may be less preferred by older adults who have a history of experiencing the emotion regulation benefits of intimate interactions with close friends and family (Carstensen 1992).

Suicide

Globally, the elderly are at higher risk of suicide than any other age group, with elderly men accounting for the largest proportion of suicides. Older adults tend to use lethal means and are less likely to report suicidal ideation prior to attempting suicide (World Health Organization 2014). Further, the current cohort of older adults is more likely to report somatic symptoms rather than emotional distress when experiencing depressed mood or anxiety (Hinton et al. 2006). A behavior-analytic conceptualization of suicidality considers suicidal thoughts and actions to be learned behaviors that function to allow the individual to avoid or escape overwhelmingly distressing and aversive feelings (Chiles and Strosahl 2005; Linehan 1993). Experiencing physical or emotional pain is a normative event at some point in the course of a long life but the ways in which individuals cope with physical pain and distressing emotions vary significantly. For individuals with a limited repertoire for coping with painful private events and weak social attachments, suicidal thoughts and actions may provide an immediate escape from or avoidance of physical pain and distressing feelings of loss, loneliness, or emptiness.

A behavior-analytic conceptualization of suicidality focuses on the context of suicidal behaviors, both private thoughts and emotions and overt actions, including the circumstances that tend to precede the occurrence of the behaviors and the consequences that reliably follow the suicidal behavior. To illustrate, suicidal ideation would be conceptualized as being negatively reinforced if it produces the consequence of temporarily alleviating painful emotion and the frequency of suicidal ideation following painful emotions has increased or is maintained over time. Alternatively, suicidal behavior would be conceptualized as being positively reinforced if it is consistently followed by a consequence such as access to social attention and comfort from friends and family and the frequency of the behavior increases over time.

Dialectical behavior therapy (DBT), developed by Marsha Linehan (1993) as a treatment for chronically suicidal individuals, focuses on replacing harmful behavior patterns (including thought patterns) with skillful alternatives. A behavior-analytic interpretation of the therapeutic process within DBT considers the contingency between suicidality, private events (e.g., distressing thoughts of abandonment), and environmental events (e.g., attention from or rejection by others). This contingency is broken by building a repertoire of behaviors (e.g., emotion regulation skills, distress tolerance skills, interpersonal skills) that are incompatible with suicidality. Although the primary population Linehan studied when developing DBT was young and female, the behavioral principles underlying the treatment are applicable throughout the life span. DBT has been found to be effective with older adults (Lynch et al. 2006).

Neurocognitive Disorders

Neurocognitive disorders such as Alzheimer’s disease and vascular dementia are among the most debilitating conditions affecting older adults. Behavior analysis has been applied to both support the maintenance of adaptive behaviors (e.g., activities of daily living, speech, etc.) and prevent, decrease, or reverse excess disability and promote the behavioral health of elderly persons with neurocognitive disorders (Buchanan et al. 2011; Fisher et al. 2007). Within this population, excess disability is said to occur when a person is more disabled than expected based on the underlying neurodegeneration (Fisher et al. 2007). Within the behavior-analytic model, excess disability in persons with neurocognitive disorders is evident in the premature decline of functional behaviors that will eventually be lost due to neurodegeneration. Designing environments that support functional behaviors is a fundamental goal of behavior-analytic approaches to enhancing the health and quality of life of persons with neurocognitive disorders (Buchanan et al. 2011; Fisher et al. 2007; Hussian 1981).

The behavior-analytic or “contextual” model of neurocognitive disorders assumes that a person who is experiencing progressive cognitive decline will develop strategies to compensate for the impairment (Hussian 1981) and that the context in which they are experiencing the neurological changes can have a profound effect on their and their family members’ behavioral health and quality of life (Schulz and Sherwood 2008). In this regard, the contextual model treats the interactions between affected persons and their family as essential to well-being.

An important implication of the age-associated risk of neurocognitive disorders is that they affect the functioning of individuals with decades-long genetic and learning histories and hence highly complex verbal, emotional, and interpersonal repertoires. Given the current lack of treatments for these disorders, behavior-analytic approaches to the support of persons with neurocognitive disorders tend to focus on three primary goals: (1) preserving functional repertoires, (2) preventing excess disability, and (3) preventing behaviors that lead to negative outcomes for persons and their families (commonly referred to as “noncognitive neuropsychiatric symptoms,” “behavioral disturbances,” or “challenging behaviors”). The behavior of family and professional caregivers is considered to be a critical feature of the context in which persons with neurocognitive experience changes and therefore the responding of caregivers is commonly targeted within behavior-analytic support services. Support of caregivers tends to focus on (1) increasing caregivers’ knowledge of neurocognitive disorders and the effects of neurological changes on behavior in order to promote perspective taking rather than pathologizing behavior, make the behavior of care recipients predictable, and reduce the likelihood that caregivers will respond to the care recipient with negative consequences, such as negative affect and corrective feedback that may inadvertently or intentionally punish behaviors within the already vulnerable repertoire of the care recipient; (2) promoting the ability of caregivers to cope with the emotional and instrumental challenges that commonly emerge when caring for someone with a neurocognitive disorder; and (3) assisting families in preserving the meaningful and rewarding qualities of their relationship with their family member (Fisher et al. 2007; Nichols et al. 2011).

Persons with dementia experience an array of neurological changes that impact their ability to perform activities of daily life such as personal care and more complex tasks such as managing finances and medications and driving. The behavior changes that accompany neurocognitive disorders are often experienced as confusing, “out of character,” or “intentional” and aversive by family members. In addition, declines in verbal abilities increasingly lead to communication problems within relationships that involve behaviors that have been under powerful stimulus control, for example, responses that have reliably followed an antecedent stimulus – such as a family member’s request or effort to initiate a conversation by inquiring about how the family member is doing – and were reliably reinforced for decades. The confluence of communication deficits and perceived intentional and unpredictable behavior changes (i.e., reflecting the breakdown in the stimulus control of behavior) that occur during the course of neurodegenerative diseases typically lead to high rates of conflict between affected persons and their family members (Fisher et al. 2007). From a behavior-analytic perspective, intra-familial conflicts may be due, in part, to the discontinuation of reinforcement contingencies (i.e., “extinction”) which is typically experienced as emotionally painful by family members.

Within the behavior-analytic or contextual model, the behavior changes that accompany neuropathology represent a natural response to increasingly overwhelming environmental demands (Hussian 1981). Declines in verbal abilities, including the ability to verbally label and respond to private events (e.g., pain, discomfort, fear, boredom, etc.), are a leading risk factor for excess disability in persons with neurocognitive disorders. This is due to the fact that the lack of ability to label, report, and respond to private events such as pain or discomfort increases the likelihood of the emergence of behavior changes that are misattributed to neurodegeneration. The default attribution of behavior change to neuropathology is a leading threat to the behavior health and quality of life of persons with neurocognitive disorders. Knowledge of the distinction between normal or expected behavior change or decline versus unusual behavior changes and careful examination of the context in which behavior changes occurred are critical for understanding the behavioral health of an individual who has a neurocognitive disorder and detecting adverse medical or environmental events. Because neurodegenerative disorders inevitably produce significant declines in behavioral repertoires, there is a risk that all observed behavior changes will be attributed to neurodegeneration, including behaviors that are a response to acute, treatable conditions such as pain, infection, or medication side effects. Through education and guided practice, family and professional caregivers can learn to understand and better predict the behavior of persons with neurocognitive disorders and hence respond in a more empathic and supportive manner (Nichols et al. 2011).

Given the current lack of a cure or effective treatment for neurocognitive disorders, geriatric healthcare advocacy groups have identified the development of support services that promote the quality of life of affected persons and their families as a priority (Odenheimer et al. 2014). The variable nature of the symptom presentation and trajectory of these disorders can limit the utility of traditional medical population-based approaches to disease management. In contrast, the idiographic nature of behavior analysis can readily accommodate the heterogeneous symptom presentation and the influence of idiosyncratic personal histories and contextual factors on the functioning of persons with neurocognitive disorders. An increasing body of literature demonstrates that behavior-analytic strategies are effective for both preventing and reducing excess disability and what are commonly described as noncognitive psychiatric symptoms, including resistance to care, wandering, and disruptive vocalizations exhibited by persons with dementia (Fisher et al. 2007; Hussian 1981).

Behavior Analysis of Health-Related Behaviors

Chronic illness and disability disproportionately affect older adults relative to other age groups. Numerous studies have documented the effectiveness of behavior-analytic strategies for promoting behavioral health and adaptive functioning in order to prevent or delay the onset of morbidity in healthy adults and prevent excess disability in persons with chronic illnesses. Domains targeted have included lifestyle factors that are known risk factors for chronic illnesses including exercise, diet, and smoking (LeBlanc et al. 2011; Roane et al. 2015). The following discussion focuses on how the principles of behavior analysis have been applied to foster health-promoting behaviors.

Behavior analysis of health-related behaviors considers the complex interaction of an individual’s repertoire and contextual variables (antecedent and consequent stimuli) that promote or interfere with the occurrence of desired behavior(s). Analysis of the temporal relationship between a behavior (e.g., eating calorie-dense food or sedentary watching of television) and its consequence(s) (e.g., immediate access to pleasurable sensation or escape from physical discomfort vs. delay of benefits) is particularly important for understanding the probability of the occurrence of health-promoting versus health-interfering behaviors. The more powerful effect of immediate reinforcement relative to delayed consequences can be a significant barrier to behavior change. Further, many health-interfering behaviors are maintained by primary reinforcers (i.e., stimuli that do not require conditioning to function as a reinforcer such as food or the reduction of pain or discomfort) while health-promoting behaviors are often maintained by delayed consequences and/or secondary (i.e., conditioned) reinforcers such as praise, a number appearing on a scale, or fitting into smaller-size clothing. The inherent delay in the consequences of many health-promoting behaviors can limit the effectiveness of setting long-term and abstract verbal goals such as “losing weight” or “getting in shape.” Behavior-analytic strategies of health promotion tend to address the differential effects of immediate versus delayed reinforcement by incorporating goal setting that focuses on increasing specific, concrete behaviors that are achievable in a short amount of time (Roane et al. 2015; King 2001). Consideration of age-associated barriers that may interfere with goal attainment is also important. Potential barriers may include chronic pain conditions, access to transportation and resources such as affordable nutritious food and fitness facilities and equipment, and reduced mobility or endurance.

Self-monitoring of specific behaviors is incorporated within many behavior-analytic health promotion programs as it allows the individual to assess the occurrence of desired behaviors, competing health-interfering behaviors, and progress toward goals in real time. Self-monitoring also enables individuals to identify barriers that require manipulation by providing detailed information about antecedents in instances when the individual deviated from their goals or lapsed. Social support in the form of praise for health behavior adherence has also been found to be effective in promoting health-promoting behaviors in older adults as it can function as an immediate, secondary reinforcement of behaviors that compete with maladaptive behaviors that have been historically maintained by powerful primary reinforcers including food or escape from discomfort by stopping exercise (Roane et al. 2015; King 2001; Penedo and Dahn 2005; Killgore et al. 2013).

Stimulus control or antecedent-based strategies involve arranging the individual’s environment in order to increase the probability that a health-promoting behavior will occur while reduce the probability of the occurrence of health-interfering behaviors. Strategies include removing antecedents for undesired behavior while increasing the salience of antecedents associated with the health-promoting behavior. For example, in promoting physical activity, an antecedent-based intervention might involve placing exercise shoes by the front door, laying out exercise clothes on the bed, or installing a stand-up desk in an office. The promotion of nutritious eating might involve removing unhealthy foods from the home and following a structured meal plan with restricted choices. In the case of smoking cessation, an antecedent-based strategy might be avoiding settings or behaviors associated with smoking, for example, bars or having cigarettes in the home, and involves replacing smoking behavior with an alternative, incompatible adaptive behavior, such as chewing gum or holding a cigarette-shaped object (LeBlanc et al. 2011; Roane et al. 2015).

Behavior-analytic relapse prevention training involves normalizing deviation from goals when attempting behavior change, for instance, missing a scheduled walk or consuming calorie-dense food, and orienting the client to the identification of antecedents or triggers for lapses as well as behavioral skill training like problem solving or relaxation exercises to promote adaptive responding if relapses occur (LeBlanc et al. 2011; Roane et al. 2015; King 2001; Penedo and Dahn 2005; Killgore et al. 2013). For example, an individual who is striving to adhere to a nutritious diet might be instructed to consume high-volume but low-calorie foods prior to social gatherings in order to prevent hunger and reduce the reinforcement value of high-calorie foods.

Sedentary lifestyle, poor diet, medication non-adherence, and smoking are well-established risk factors for an array of preventable, chronic illnesses prevalent within the elderly population including cardiovascular disease, stroke, osteoporosis, many cancers, and type 2 diabetes. Relatedly, there is also vast evidence that even small changes in lifestyle factors such as increasing physical activity can decrease the morbidity and mortality due to cardiovascular disease (King 2001; Penedo and Dahn 2005; Killgore et al. 2013; Chapman et al. 2013). In addition, recent research indicates physical activity can buffer age-related cognitive decline. That is, older adults who are physically active are less likely to demonstrate cognitive decline relative to their more sedentary counterparts (Chapman et al. 2013). Thus, the promotion of adaptive functioning in older adults via behavior-analytic strategies has the potential to contribute to both improved quality of life and the prevention of chronic disease and disability.

Cross-References