This study is part of a broader research entitled “Mental health profile of elderly in the city of Campina Grande-PB”, with a quantitative approach and ex-post facto design. Cluster sampling from the city of Campina Grande-PB was used. Forty-four census sectors were drafted (used as clusters), which are delimited by the IBGE and are considered as the smallest territorial unit with identifiable physical limits and appropriate dimensions to undertake research (IBGE 2011).
Sample
The minimum sample size estimated for the research was 381, for a 5 % sampling error and σ = 1.96 (represents the 95 % confidence interval). The establishment of this value was estimated based on the sample size needed to obtain a proportion of 50 % for the occurrence of a certain characteristic of the elderly population (value at which the sampling size obtained in the highest possible for p = 0.50 and q = 0.50). Considering a population of 42,817 elderly people (IBGE 2011), the formula for finite populations was used: n = σ2 . p.q.N/e2 . (N-1) + σ2 . p.q. Therefore, approximately nine elderly were collected per census sector.
In total, 381 male and female elderly accepted to participate in the research free of charge, aged 60 years or older, who reached the cut-off point established by the Mini-Mental State Examination (MMSE), assessed according to the participant’s education level (Brucki et al. 2003). In this assessment, it was considered that the answers of elderly with cognitive problem scores could compromise the reliability of the tools to be applied. Thus, to comply with the psychometric requisites of the scales, attention, concentration and preserved reasoning skills are needed.
This research found a predominance of the female sex (73.4 %), with a mean age of 71.50 years (SD = 8.00), most elderly being concentrated in the age range between 60 and 69 years (44.4 %). The majority is present or lives with a partner (44.1 %). As regards education, 53.3 % declared having finished primary education. Most elderly self-declared Catholic (71.9 %), had their own home (75.9 %) and a monthly personal income of up to one minimum wage (60.8 %).
Elderly with a cognitive deficit were excluded; as well as patients with severe sequelae after Cerebrovascular Accident; patients with Parkinson’s Disease in a severe or unstable stage; patients with severe hearing or sight impairments; and patients unable to participate due to severe illness. These criteria were necessary, considering that elderly people committed by these problems would face difficulties to answer the research questionnaires.
Tools
The elderly answered a protocol consisting of a demographic questionnaire, the Mini-Mental State Examination (MMSE) (Folstein et al. 1975), the Problem Coping Mode Scale (PCMS) (Seidl et al. 2001), Satisfaction with Life Scale (SLS) (Diener et al. 1985) and the Positive and Negative Affect Schedule (Diener 1984).
To describe the demographic characteristics of the sample, demographic questions were applied, specifically: sex, age, education, marital status, religion, housing and income of the elderly.
The Mini-Mental State Examination (MMSE) is one of the most used cognitive screening tools in the world. Its validity has been evidenced in different studies. Folstein et al. (1975) found no statistically significant differences between the test and retest with a 28-day interval and a Pearson correlation coefficient of 0.83. Lourenço and Veras (2006) found sensitivity, specificity, positive and negative predictive values of 80.8, 65.3, 44.7 and 90.7 %, respectively. The same authors also weighted that education should constantly be considered for the adoption of cut-off points. The criteria from the FIBRA study (Frailty of Brazilian Elderly) were used: 17 (illiterate); 22 (education between 1 and 4 years); 24 (education between 5 and 8 years); 26 (9 or more years of education) (Brucki et al. 2003).
The Problem Coping Mode Scale (PCMS) (Seidl et al. 2001) reveals the coping strategies the subjects use when in stressful situations. It consists of 45 items, answered on a Likert scale, divided in four factors, which express how people cope with their problems. In the validation article, the four factors were extracted using the main axis method with orthogonal rotation and explained 25 % of the shared variance: Problem-focused coping strategies (α = 0.84); Emotion-focused coping strategies (α = 0.81); Religious practices/fanciful thinking (α = 0.74) and search for social support (α = 0.70).
The Satisfaction with Life Scale (SLS) was elaborated by Diener et al. (1985) to assess how satisfied people judge they are with their lives. It consists of five items and is answered on a Likert scale ranging from 1 (I completely disagree) to 7 (I completely agree). The scale was adapted and validated for the context of Paraíba by Albuquerque et al. (2010), in a sample of 342 elderly. In its validation, it showed a one-dimensional structure with an explained variance of 59.7 % and Cronbach’s Alpha of 0.84.
The Positive and Negative Affect Schedule was formulated by Diener (1984) and consists of ten adjectives, being five positive and five negative, with alternative answers ranging between 1 (Nothing) and 7 (Extremely). The objective is to assess the intensity of the positive (“Happy”, “Satisfied”, “Amusing”, “Optimistic” and “Joyful”) and negative affects (“Depressed”, “Frustrated”, “Mad”, “Concerned” and “Unhappy”). The scale was validated by Albuquerque et al. (2010) for the context of Paraíba, in a sample of 342 elderly, explaining 40.88 % (positive affects) and 18.18 % (negative affects), respectively; with a Cronbrach’s Alpha of 0.78 in each factor.
Procedures
The study started after the Committee for Ethics in Research at the Universidade Estadual da Paraíba (UEPB) gave its approval, registered under number 0655.0.133.000-11, in compliance with the standards of the National Health Council Resolution No. 196/96.
Ten psychology students from a public education institution in the state of Paraíba, properly trained and identified, visited the homes in the sectors drafted for the research, according to the geographical region mapped, working in pairs. If any elderly was present at the home visited, questions were asked about his willingness to participate in the research. After the acceptance and signing of the Free and Informed Consent Term (FICT), the data collection process started with the application of the tools. The first tool to be applied was the MMSE, followed by the scale of Coping strategies, with the researcher informing the participants that the questions dealt with how they were facing their health problems. The final two questionnaires were the SWB (Positive and Negative Affect Schedule and Satisfaction with life) and the demographic questionnaire.
Data analysis
The collected data were analyzed using the software PASW and AMOS (versions 18). Descriptive statistics were calculated for the demographic variables and for the PCMS, Satisfaction with Life scale and Positive and Negative Affect Schedule. In bivariate analyses, Student’s t-test (sex and occupation with SWB), Pearson’s correlation (between income and SWB) and ANOVA for repeated measures with Bonferroni’s post-hoc test (used among the factors of the SWB). To verify the relation between the coping strategies and the SWB dimensions, multiple linear regression (using the stepwise method) was used to determine the predictive capacity of the coping strategies in each of the three subscales of subjective well-being. The four factors of the PCMS were included as independent variables and the SWB variables as dependent variables.
The software AMOS was used to confirm the factorial structure of the measures and thus top up the reliability of the results. Therefore, the following adjustment indicators were considered (Tabachnick & Fidell 2013):
The index χ
2/df (degrees of freedom). This is considered a subjective quality of adjustment. An index between two and three is recommended, accepting values up to five to indicate the fitness of the theoretical model to describe the data.
The Goodness-of-Fit Index (GFI) and the Adjusted Goodness-of-Fit Index (AGFI) are adjustment indicators that reflect the proportion of variance-covariance in the data explained by the model. The indices range between 0 and 1. Values of 0.90 or higher are recommended.
The Comparative Fit Index (CFI) is an additional comparative index of adjustment to the model, with values closer to 1 expressing better adjustment. Commonly, values close to 0.90 are admitted as a reference of an adjusted model.
The Root-Mean-Square Error of Approximation (RMSEA) and its 90 % confidence interval (90 % CI) are based on the residual values; high values indicate a non-adjusted model. Therefore, values between 0.05 and 0.08 are recommended, admitting up to 0.10.