One of the major benchmarks in a child's life is the achievement of urinary continence. Once attained, continence normally persists throughout life. This ability to control storage and emptying of urine requires complex coordination between the lower urinary tract and neural structures and pathways. Researchers, however, have long been aware of the behavioral aspects of urinary continence. Henriksen (1962) noted that it was a "psychophysiologic act permitting the controlled escape of urine from the bladder" (p. 892). At the societal level, continence also has been viewed as a behavior necessary for individual and species survival. For example, during the act of micturating an individual can be vulnerable to physical attack, thus survival is enhanced if the bladder empties under volitional control when the environment is deemed safe, rather than emptying randomly (Blok & Holstege, 1999). Random micturition also raises hygienic issues. Exposure to excreta is considered disgusting across cultures and it is hypothesized that disgust is an aversion that leads to behaviors designed to protect people from disease (Curtis & Biran, 2001). Continence then provides a mechanism to reduce exposure to another's and one's own urine and to potential pathogens. Thus, the behaviors affecting the storage and elimination of urine involve physiologic and neurological mechanisms and intrapersonal and group attitudes and beliefs, and these behaviors are shaped and judged by social norms.

The volitional nature of continence remains a central concept to its definition and the definition of incontinence. The current International Continence Society definition states that urinary incontinence (UI) is the "complaint of any involuntary leakage of urine" (Abrams et al., 2002, p. 168). Recently, considerable knowledge has been gained regarding the complex neurological and biomechanical aspects of normal urinary function and pathophysiologic mechanisms of UI. As more is learned about the biological aspects of UI, concurrent advances in the behavioral aspects of UI must continue. The purpose of this presentation is to discuss historical and recent UI research guided by health behavior change theories and to identify the gaps in current knowledge. Special emphasis is placed on recommendations for future research.

Health Behavior Change Theories and UI Research

Health behavior change theories have been used in UI research for over 40 years. Two distinct populations have been the foci of this research effort: (a) institutionalized psychiatric and cognitively impaired adults; and (b) noninstitutionalized adults, especially women, with overactive bladder and stress urinary incontinence (SUI). The term health behavior change theories encompasses a wide scope of theories, all of which share the assumptions that health is mediated by behavior and that health behaviors have the potential for change. Health behavior change theories can be divided into two broad groups: (a) operant conditioning theory; and (b) a variety of cognitive theories such as the Health Belief Model and Social Cognitive Theory (Table 1). Operant conditioning theory has evolved a wide variety of operant conditioning techniques, such as habit training with material or verbal reinforcement. Derived from the cognitive theories are a variety of cognitive techniques, including relaxation, distraction, and education. A behavioral intervention is any intervention intended to elicit change in behavior. Any of the health behavior change theories, or a combination of them, may be employed by the researcher as the basis of the intervention. Only operant conditioning principles have been used in behavioral interventions for UI with institutionalized psychiatric and cognitively impaired older adults, while both operant conditioning theory and cognitive theories have been used with noninstitutionalized adults.

TABLE 1. Health Behavior Theories That Guide Urinary Incontinence Research

TABLE 1. Health Behavior Theories That Guide Urinary Incontinence Research

Behavioral Interventions With Institutionalized Psychiatric and Cognitively Impaired Older Adults

Carpenter and Simon (1960) tested operant conditioning techniques on institutionalized behaviorally regressed psychiatric patients. Operant conditioning involves the manipulation of antecedents, consequences, or both, which are also known as multicomponent interventions (Burgio & Burgio, 1986), to elicit a desired behavior (Baldwin & Baldwin, 1986). Habit training, habit training with social approval, and habit training with material reward or deprivation were tested to reduce or eliminate UI frequency medically deemed not due to organic causes. Patients were nonrandomly selected for one of three treatment groups or for the control group. Patients in the material reward group experienced greater reduction and return to the continent state than did those assigned to the other groups, including the control group. Carpenter (1963) used data from this study to explore the effect of changes in personnel on the reduction in frequency of incontinence. Carpenter found that as personnel were reassigned across the different intervention groups, changes in the frequency of incontinence did not occur, thus providing evidence that the changes in incontinence frequency were due to the intervention rather than the individuals delivering the intervention. Of note, Carpenter concluded that habit training without social approval or material reward offered little effect on the reduction of incontinence.

Another nursing study involving the use of operant conditioning was conducted with cognitively impaired older adults (Grosicki, 1968). The use of social reinforcement, material reinforcement, and a combination of both types of reinforcement were hypothesized to reduce UI, fecal incontinence, or both. No significant reductions, however, in incontinence occurred in the treatment group. Grosicki noted that several barriers to conducting this study existed including small sample size, staff shortage, and staff "hopelessness regarding outcomes for the aged" (p. 310) that affected the implementation of the study.

Azrin and Foxx (1971) viewed normal toileting as a "complex operant and social learning process" (p. 89). These researchers concluded after a successful trial with long-term male patients having a median IQ of 14 (range 7-45) that manipulation of both stimuli (i.e., antecedents) and consequences (i.e., reinforcement) was effective in reducing incontinence. The complex protocol involved having participants drink extra fluids to bring on the need to void, shaping toileting behavior, giving positive reinforcement, and incorporating staff reinforcement procedures. They noted that continence was more than a physiologic response to the sensation of a full bladder. Manipulating social factors by providing immediate (direct) reinforcement for correct toileting behavior and introducing a negative reaction (punishment) to incontinent episodes played a major role in the reduction of incontinent behavior.

Atthowe (1972), in an attempt to reduce nocturnal enuresis in long-term chronic psychiatric patients, devised a two-phase intervention. In phase one an "aversive environment" was created; during phase two, a token (reward) system was established. Incontinence dramatically decreased in phase one and remained low during phase two and at a 22-month followup. Atthowe noted that staff motivation and support of the intervention were integral components of its success.

Operant conditioning dominated incontinence research for over 20 years. Stimuli used to induce the desired behavior included providing extra fluids to create the need to void, giving assistance to the toilet, and prompting to use the toilet. Contingencies (consequences) included receiving verbal praise or tokens, getting physical contact such as hugs, and wearing one's own clothes. Because frail adults relied on caregivers for toilet access, researchers also focused on changing caregiver attitudes and behaviors.

A complex behavioral intervention based on operant conditioning principles manipulated antecedents to and consequences for UI in cognitively impaired elderly nursing home residents. This intervention, prompted voiding, also included a staff performance component usually in the form of supervisory feedback. The goal was to decrease incontinent episodes and to increase self-initiated toileting (Burgio, Engel, McCormick, Hawkins, & Scheve, 1988;Schnelle et al., 1983). Schnelle, Newman, and Fogarty (1990) found nursing home residents' dryness increased, but this behavioral change did not act as reinforcement for staff members to continue providing cues to the older adult about using the toilet and offering assistance (Palmer, 2004).

Despite evidence that older adults became drier with the prompted voiding intervention, clinicians and researchers believed that the intervention was too labor-intensive for the staff levels in nursing homes. As clinicians adapted the protocol for the clinical environment they removed one component, patient-initiated toileting requests. Attempts to increase self-initiated requests for toileting are rarely found in current clinical prompted-voiding programs. Also, emphasis on supervisory feedback has been reduced and efforts to better select appropriate candidates (such as those who respond to a 3-day prompted voiding trial) have increased (Palmer, 2004). Thus the principles of operant conditioning, the theoretical foundation for incontinence interventions with cognitively impaired older adults, have been modified in the clinical setting yet little research has been conducted on these changes.

Behavioral Interventions With Noninstitutionalized Adults

Adults, especially women, with SUI and overactive bladder have been a focus of behavioral intervention research. Much of this research has been guided by theories that involve changing knowledge level, attitudes, and beliefs as well as physiologic responses to stimuli. Sampselle (2003) in a state-of-the-art article noted that bladder training and pelvic floor muscle exercise (PFME), two techniques employing operant conditioning principles, have been the interventions most studied. Other techniques used with bladder training include cognitive learning strategies derived from Social Cognitive Theory (Table 1) such as relaxation and distraction to delay voiding, self-monitoring, and patient education. This intervention is designed to alter the individual's cognitive and physical response to an antecedent stimulus (sensation of the need to void) until a specified period of time has elapsed. The bladder is "trained" to relax and fill to achieve a normative (i.e., 4-hour) bladder emptying schedule. This intervention combines elements of both operant conditioning and Social Cognitive Theory. Little research has been conducted using other components of Social Cognitive Theory, such as self-efficacy or outcomes expectancy, to improve physical outcomes such as incontinent episode frequency.

Pelvic floor muscle exercise is an effective intervention for SUI and overactive bladder (Sampselle, 2003) that incorporates both cognitive learning concepts and exercise physiology principles. The goal in treating SUI is to increase muscle bulk and tone and to improve volitional control over pelvic muscle relaxation during times of increased intra-abdominal pressure. The training phase of PFME often includes a computerized biofeedback component. Biofeedback is considered a multicomponent intervention that assumes physiologic responses are behaviors (Burgio & Burgio, 1986). The patient can observe the magnitude and duration of muscle contraction on a computer monitor, thus receiving instant reinforcement for that behavior. If the incorrect muscle or muscle group is contracted, instant corrective feedback (i.e., instruction) from the practitioner can be provided to alter or modify the behavior. Also, biofeedback has been used to treat overactive bladder, by filling the bladder and having the patient engage in bladder and abdominal muscle relaxation and pelvic muscle contraction. The patient receives immediate feedback about the physiologic response to his or her attempts to voluntarily control bladder filling and inhibit emptying (Burgio & Burgio, 1986).

Another example of the use of cognitive learning strategies in incontinence research is the use of a quick intentional contraction of the pelvic floor muscle, also called the knack, for women with SUI (Miller, Ashton-Miller, & DeLancey, 1998). Although these authors did not identify an explicit health behavior change theory or measure behavioral outcomes such as self-efficacy they did note that cognitive ability and the ability to make a connection between a stimulus (i.e., cough) and outcome (i.e., incontinence) were necessary to effectively perform the knack.

In interventions based on operant conditioning, consequences are central to shaping overt behaviors. In interventions based on Social Cognitive Theory, cognitive components are included to change both overt and covert behavior (i.e., attitudes, beliefs, and knowledge). Behavioral interventions based on operant conditioning are appropriate to shape overt volitional behavior in individuals with significant cognitive and learning impairments. Interventions based on Social Cognitive Theory may be effective for individuals who can change covert behaviors by learning vicariously through role-modeling or instruction.

Adherence or Compliance to Behavioral Interventions

Many behavioral interventions based on health behavior change theories require extensive periods of time to determine their effect. For example, women may need to perform PFME daily for several weeks before an effect on incontinent episodes is evident. Adherence to the regimen, defined as following the advice about performing specific behaviors, is an issue for both practitioner and patient. Henrikson (1962) found that the more severe the symptoms of incontinence the more compliant the patient is to PFME.

Researchers in The Netherlands described a program to develop adherence to a behavioral intervention for incontinence (Alewijnse, Mesters, Metsemakers, & van den Borne, 2002), described predictors of long-term adherence (Alewijnse, Mesters, Metsemakers, & van den Borne, 2003), and tested the effectiveness of an exercise program supplemented with an educational component (Alewijnse, Metsemakers, Mesters, & van den Borne, 2003). These researchers used Social Cognitive Theory, elements of the Health Belief Model (e.g., self-efficacy), and the Transtheoretical Model (i.e., stages of change;Table 1) to develop the program. To promote adherence they used reminders and provided structured feedback to women who were using a self-management process. One of the gaps in knowledge about adherence or maintenance of a behavioral change program is what the experience of the behavioral change means to the individual. The role of the individual's assessment of initiation of the behavior change and the decision to maintain that change is poorly understood (Rothman, 2000).

Primary Prevention Research

Primary prevention of incontinence, which is the maintenance of continence (Palmer, 1994, 2002), and theory development in this area were neglected until recently. The Health Belief Model and Subjective Expected Utility Theory were used to predict adoption of PFME by postpartum women (Dolman & Chase, 1996). The authors concluded that for women to be compliant, PFME must become a habit rather than requiring a conscious decision. How this shift from conscious decision-making to a habitual behavior is to occur is unclear.

Diokno and colleagues (2004) reported the findings of a randomized clinical trial with postmenopausal women using a behavioral modification program to prevent incontinence and to increase pelvic floor muscle strength and voiding control. The intervention involved group instruction from an urologist and nurse specialist and daily use of audiotapes that carried instructions for performing PFME or bladder training. The underlying theory, although not explicitly stated, was Social Cognitive Theory, as increasing participants' knowledge and skills to perform PFME were considered essential components to the intervention. Further refinement and testing of incontinence primary prevention theories are essential for the advancement of efforts to reduce the incidence of incontinence in vulnerable groups such as postpartum and elderly women.

Recommendations for Future Research

Behavioral interventions are being used daily by incontinent individuals in a variety of settings. Little long-term follow-up research on the efficacy of these interventions has been reported. Modifications to theory-based protocols such as prompted voiding have been made but little research has been conducted to determine if these changes result in better patient outcomes or staff performance. It may be that the expectation of self-initiated requests for toileting from individuals with advanced dementia is not feasible. This area should be explored especially because researchers have little insight into the factors that act as barriers to maintaining a behavioral change.

Many of the theories that underpin incontinence behavioral interventions do not take environmental and social factors into account (Table 1). For example, little intervention research is being conducted in the long-term care setting that explores the effect of social factors and changing patient and staff expectations about incontinence care. Compelling evidence exists that nursing home residents hold "reduced expectations" (p. 1760) for receiving toileting assistance; that is, because they do not receive frequent toileting they do not expect to receive it (Schnelle et al., 2003). The norm in long-term care is for incontinent residents to wear absorbent products (Watson, Brink, Zimmer, & Mayer, 2003). Needed are theory development and testing to explain how staff and patient expectations that incontinence cannot be successfully treated affect patient outcomes. For example, qualitative studies that explore staff and patient expectations and perceived barriers could lead to theory development concerning the role of the social environment on individual and group behavior related to continence.

Criticisms of behavioral interventions in the long-term care setting include their labor-intensive nature (Palmer & Johnson, 2003). Therefore, behavioral research on advances in technology that may reduce the intensity of labor, such as advances in absorbent products and portable urine containment devices (e.g., female urinals), is needed. New models of care guided by theory that shift emphasis from containment of urine to active treatment and prevention of incontinence, perhaps by reinforcing continent behaviors, is clearly needed in long-term care.

Nurse researchers benefit from incorporating health behavioral theories into their research. Because of the prevalence of incontinence, nurse researchers need to apply their skills to the development of theory-based interventions directed at populations rather than individuals. For instance, there are many reports that women do not discuss incontinence with their healthcare providers (Palmer & Fitzgerald, 2003). Incorporation of social marketing strategies that target specific groups, such as middle-aged women, may help increase the number of women seeking help with incontinence or willing to participate in research. Strategies may be designed to promote benefits for adopting behaviors and reduce barriers, such as feelings of stigma and shame, to getting information and healthcare.

Incorporation of stages of change discussed in the Transtheoretical Model (Prochaska & DiClemente, 1983;Table 1) into population-based interventions may help some women change from their current behavior of containing urine with absorbent products (precontemplation stage) to an active self-management intervention (action stage). Understanding a person's willingness or even awareness of the need for behavioral change is essential for the design of relevant and acceptable interventions. Because factors other than intrapersonal ones affect a person's ability to adhere to treatment protocols (Alewijnse, Mesters, Metsemakers, Adriaans, & van den Borne, 2001;Kincade, Johnson, Ashford-Works, Clarke, & Busby-Whitehead, 1999), theories that take into account the effects of external factors are needed.

New paradigms about the pathophysiology of incontinence are emerging, especially in neuro-urology. As the role of neurotransmitter levels on continence is better understood (Thor, 2003), health behavior change theories will require modification to accommodate this new knowledge. Refined measurement tools that can determine the effect of a behavioral intervention on relevant anatomical structures and physiology, including neurotransmitter levels will be needed. Incorporation of biobehavioral concepts into theory will help to advance the design of nursing behavioral interventions designed to prevent or treat UI.

Great advances in behavioral research related to UI have taken place. New insights into environmental factors, social influences, and physiological function and greater understanding about the differentiation between initiation and maintenance of behavior will drive the development of complex health behavior theories. The development and testing of these new theories will guide the next generation of incontinence researchers and ultimately lead to reducing the incidence and prevalence of incontinence.

References