Sample

We found 56 articles: 25 in the Web of Knowledge Cross Search, 30 in PubMed/MEDLINE and one in SciELO. After the exclusion of redundant cross-references (n = 11), 45 studies were selected. Thirty-one studies were excluded for different reasons, as shown in Table 1. We added one study manually, and our sample comprised the complete texts of 15 studies.

Resilience and psychological factors (n = 15)

Higher levels of resilience were associated with psychological factors, such as coping strategies,^{10 , 12} positive cognitions,¹ resources,¹ optimism,⁶self-efficacy,⁶ internal locus of control,⁶ full engagement in daily activities⁴ and search for challenges.⁴ Higher levels of resilience were also associated with decreased burden,^{1 , 2 , 28} stress,⁶ neuroticism,² perceived control⁴ and distress.^{5 , 24} A study found that caregivers that provided care for a longer duration of time had higher levels of resilience.²⁸ Another study showed that resilience did not significantly explain caregiver's sense of marital satisfaction.¹³

Resilience and biological factors (n = 9)

Higher levels of resilience were associated with less depression^{4 , 7 , 29} and better physical health.^{2 , 3} Other biological factors associated with higher levels of resilience were: older age,¹⁴ African-American ethnicity²⁹ and female sex.²⁸

Resilience and social factors (n = 5)

Studies found a positive association between higher levels of resilience and social factors, such as good social support,^{2 , 15 , 20} satisfaction with social support²⁰ and individual, family, and community resources.¹⁹

Methods

Most studies were cross-sectional,^{1 - 3 , 5 , 6 , 10 , 12 - 14 , 18 , 24} but longitudinal^{4 , 28 , 29} and descriptive^{1 , 3 , 5 , 10 , 14 , 24} designs were also used.

Most samples included caregivers of people with Alzheimer's disease (AD).^{1 , 3 , 4 , 7 , 12 , 14 , 18 , 29} Caregivers of people with AD and vascular dementia (VD) were included in one study,¹⁰ and caregivers of people with AD, VD and mixed dementia (MD) were compared in 13% of the studies.^{2 , 6} Two studies compared AD and similar disorders,^{13 , 28} and two did not specify the type of dementia under study.^{5 , 24}

The most frequent assessment instrument was the complete and brief versions of the Resilience Scale developed by Wagnild & Young.^{3 , 5 , 12 , 13 , 18} Other instruments were the Connor-Davidson Resilience Scale (CD-RISC)^{2 , 7} and the Dispositional Resilience Scale.⁴ Two studies did not have any specific instrument to assess resilience, but used scales that assess associated constructs, such as the Resourcefulness Scale¹ and the Battery of Generalized Expectancies of Control Scales (BEEGC-20).⁶ One study²⁸ administered the Zarit Burden Interview to assess resilience and analyzed results according to reports of burden and frequency of demands of the person with dementia. In one study²⁹ the Center for Epidemiological Studies Depression Scale (CES-D), Life Satisfaction Index (LSI-Z) and the brief and revised version of the Social Support Questionnaire (SSQSR) were used. Finally, two studies used qualitative interviews that were fully transcribed and analyzed.^{10 , 24}

The determinant factors and the methods of studies about resilience are shown inTable 2.

Resilience definition

We conducted a search to retrieve studies that discussed the definitions and determinants of resilience among caregivers of people with dementia. We did not find a consensus about how to define resilience, as several terms are used as synonyms for it: adaptation when faced with adversity;^{1 , 2 , 4 , 5 , 13 , 14 , 18} flexibility¹³; belief in oneself and internal locus of control⁶; efficacious coping^{2 , 18}; psychological well-being and strength¹²; development of a healthy life³; perceived burden²⁸; and amount of and satisfaction with social support.²⁹ Therefore, we concluded that each study defined resilience according to its objectives and assessment methods, rather than based on any clearly established theoretical framework of resilience in dementia. These various definitions may explain the differences found in the results of the studies under assessment here.

The lack of a theoretical framework may give rise to several methodological problems, such as the absence of standardized assessment instruments^{1 , 4 , 6 , 7 , 14 , 28} and the heterogeneity of samples.^{6 , 12 , 13 , 28} However, most studies converged on the definition of resilience as the capacity to adapt successfully when faced with the stress of adversity.^{1 , 2 , 4 - 6 , 13 , 14 , 18} Therefore, resilience seems to have some degree of conceptual overlap with similar constructs used in dementia caregiving, such as mastery or degree to which individuals attribute outcomes to their own efforts or abilities, self-efficacy, internal locus of control, and learned helplessness. Other constructs, possibly shared by resilience, are personality traits, such as neuroticism, which may predispose some caregivers of people with dementia to reporting greater feelings of burden, behavior problems or subjective ratings of health. Therefore, it remains unclear whether resilience is a learning ability or a personality trait.

Resilience should be understood not only as a personal attribute that may lead to success, but also as the dynamic interaction between biological and psychosocial processes. Resilience of caregivers of people with dementia should be defined in association with caregiving tasks, because it implies resistance to stress due to the perception of caregiving demands and burden. This definition suggests that resilience should be assessed in association with reports of burden and frequency of demands from care recipients, which may be associated with neuropsychiatric symptoms and functional dependence.²⁸

Psychological factors

In general, higher control over life helps adaptation and improves understanding of the situation, acceptance of changes and confidence of caregivers on their own strength and on the affective bonds shared with care recipients.² Therefore, resilience may be understood as a force, a strategy and a set of skills used in a specific context by a particular individual to deal with a particular adverse event. The coping strategy adopted by caregivers with greater resilience is task focused.^{10 , 12} This strong association between task-focused coping and resilience suggests that task-focused skills training programs may improve and protect caregiver mental health and prolong the time a loved one may remain at home.¹²

According to Contador et al.,⁶ optimistic people make a subjective estimate of the probability of achieving goals or desired results based on other factors, such as self-efficacy (belief in their own capacity of starting actions, achieving goals and dealing with life) and internal locus of control or contingency (how individuals expect life events or results to be motivated by their actions). Self-efficacy and internal locus of control are psychological factors that support the construct of resilience and are associated with less suffering from the negative effects of stress and burden. Therefore, people with grounded optimism, high level of self-efficacy and internal locus of control may be more resilient to stressful situations, which may reduce vulnerability to burden.⁶

O'Rourke et al.⁴ assessed three constructs that compose resilience. The first was commitment to life, as resilient people tend to engage fully in daily activities. Second, resilient people enjoy challenges and prefer changes rather than stability, and life travails provide opportunities to develop skills and self-knowledge.⁴ Finally, psychological resilience entails perceived control, which results in a sense of personal autonomy and a belief in being able to directly affect the course of life events.⁴

These characteristics, associated with burden, stress and distress, may be protective factors or risk factors for people with dementia. Some authors suggest that higher levels of resilience are inversely associated with low burden,^{1 , 2 , 28} stress⁶ and distress levels.^{5 , 24}

Bekhet¹ concluded that positive cognitions are protective factors for caregivers of people with dementia. Positive cognitions, or positive thinking patterns, enhance mental health and improve the capacity to manage daily activities.¹

Biological factors

Higher resilience scores were associated with improved physical health.² Caregivers with lower resilience scores tended to visit the doctor less frequently and take less medicine. This might be related to a lower degree of self-care. Comparatively, higher degrees of resilience were associated to fewer alterations of health habits² and better physical health.³ Moreover, higher degrees of resilience were associated to low somatization and depression.³

The role of mood in resilience was demonstrated by Lavretsky et al.,⁷ who concluded that antidepressant use in depressed family dementia caregivers with either major or minor depression might improve depression symptoms, boost resistance to stress and improve coping strategies. Moreover, there is a significant association between resilience and aging, which indicates that resilience is acquired as the person gets older.¹⁴ This finding is important because caregivers are the same age as elderly people with dementia. Further studies should investigate the neurophysiologic mechanisms of resilience and their association with psychological functioning and social factors.

Social factors

Social support has been positively associated with resilience,^{2 , 18 , 29} but studies have defined support in different ways: network size and frequency of contact; support received; frequency of help from other people; and general satisfaction with social network.²⁹ Social support associated with perceived help and satisfaction with support seem to be more consistent than actual support or size of support network.²⁹ A study²⁹ that assessed ethnic differences in resilience found that resilience of Afro-American caregivers was partially explained by higher levels of satisfaction with social support.

Wilks & Croom¹⁸ found an inverse association of social support and stress resulting from the combination of: a) negative influence of stress on caregivers' perception about available social resources; b) negative influence of stress on caregivers' capacity of using these resources; and c) higher perception of available social resources associated with lower probability of perceived stress. Therefore, social support is a moderating factor of resilience, as several types of support may relatively relieve the physical and mental burden caused by stress.¹⁸

Therefore, resilience is a dynamic process believed to arise from interactions between biological and psychosocial factors. This process involves physiological adaptation to stress, psychological adjustment, self-efficacy, effective coping strategies and cognitive redefinition of experience. However, studies in the area often focus on only one factor, such as coping.^{1 , 3 , 6 , 7 , 12 - 14 , 18} Clinically, partial assessments of resilience may compromise the development of interventions focused on the improvement of this capacity among caregivers of people with dementia.

Methods and instruments: difficulties in the assessment of resilience

Although the assessment of dynamic constructs, such as resilience, may be methodologically limited, most studies in this review were cross-sectional.^{1 , 2 , 5 , 6 , 10 , 12 , 13 , 18 , 24} Longitudinal studies would be more appropriate because they would follow the development of resilience among caregivers and the cognitive decline of people with dementia, as well as the increase of care demands.

Most samples included caregivers of people with AD,^{1 , 3 , 4 , 7 , 12 , 14 , 18 , 29} but some assessed, without comparing, caregivers of people with AD, VD, MD, mild cognitive impairment (MCI) or unspecified dementia.^{2 , 6 , 10 , 13 , 28} There are important clinical differences between types of dementia and, therefore, between different forms of caring. Heterogeneous samples, found in some studies, precluded the assessment of some factors, such as resilience.

In all studies, most participants were informal female caregivers^{1 - 7 , 10 , 12 - 14 , 18 , 24 , 28 , 29} and spouses or children of elderly people with dementia,^{1 - 6 , 10 , 12 - 14 , 18 , 24} whose mean age was 60 years^{2 - 5 , 7 , 10 , 13 , 14 , 24 , 28 , 29} and whose mean schooling was eight years.^{1 , 2 , 3 , 5 - 7 , 13 , 14} This caregiver profile indicates that elderly people have been taking care of elderly people, and the most important challenge is to prevent and treat the diseases that may affect caregivers. The study of resilience may bring results that define a protective factor for caregivers.

The use of assessment instruments that provide only partial data might have limited the understanding of resilience. The instrument most often used to assess resilience was the complete or brief version of the Resilience Scale of Wagnild & Young.^{3 , 5 , 12 - 14 , 18} This scale measures factors associated with resilience, such as self-esteem, locus of control, coping styles and sense of humor. However, the lack of a consensus about the definition of constructs resulted on the use of different instruments, including those designed to assess depression and burden, which makes it difficult to replicate these studies and raises questions about the reliability of their results.