Convergent validity between SF-36 and WHOQOL-BREF in older adults.

OBJECTIVE: To compare the reliability and convergent validity of instruments assessing quality of life in Brazilian older adults.
METHODS: Cross-sectional study of 278 literate, community-dwelling older adults attending a municipal university for the elderly in Sao Carlos, SP, Southeastern Brazil between 2006 and 2008. The Brazilian versions of the SF-36 and WHOQOL-BREF instruments to assess quality of life were compared. Cronbach's alpha coefficient was used to estimate reliability and Pearson's correlation for comparison between the two scales.
RESULTS: Most of participants were women (87.8%) with a mean age of 63.83±7.22 years. Both scales showed an acceptable internal consistency - WHOQOL-BREF Cronbach's alpha was 0.832 and SF-36 was 0.868. There was a weak (r ≤ 0.6) correlation between the related fields in the two questionnaires.
CONCLUSIONS: The SF-36 and WHOQOL-BREF are reliable instruments for clinical and research uses in Brazilian older women. To select one, researchers should consider which aspects of quality of life they aim to capture because of weak convergent validity signs. This study's results indicate that WHOQOL-BREF may be more relevant to evaluate changes in the quality of life of older women because it prioritizes responses to the aging process and avoids focusing on impairment.

INTRODUCTION

The 36-item form constructed to survey health status (SF-36) and the World Health Organization quality of life assessment instrument (WHOQOL-BREF) are widely used to evaluate quality of life in old age. A pilot study involving a longitudinal follow-up of middle-aged and older people showed differences in scores of these two instruments.[3] However, only one study included participants over 60 years old and examined psychometric properties of these two instruments.[12] Therefore, questions about their reliability and validity in healthy older populations remain unanswered. Indeed, some studies have shown conflicting results between these instruments in healthy adults and in patients with coronary disease.[7,9] Other studies reported similarities between the two instruments for groups of patients with specific clinical conditions.[6,8] Few studies have compared these two instruments in Brazilians. Indeed, a preliminary study of a young population showed weak convergent validity for Brazilian individuals.[11] Given this scenario of scarce knowledge, it is important to understand the psychometric properties of quality of life instruments to evaluate and to follow-up Brazilian older adults.

Therefore the aim of this study was to compare the reliability and convergent validity of life quality instruments in the older population.

METHODS

This was a cross-sectional study with 278 volunteer participants of at least 60 years of age from University of the Third Age of Sao Carlos, SP, Southeastern Brazil, in 2006-2008. The sociodemographic characteristics of the university population is similar to those of this study participants, with an average age of less than 65 years and both education and social class of great variability.[a]

All participants were literate and clinically assessed as being in good health by a physician.

The instruments used were the Brazilian versions of WHOQOL-BREF[5] and SF-36.[4] Both questionnaires were answered by the participants in a single meeting, first the WHOQOL-BREF and next the SF-36. When asked to help, the researcher was limited to re-reading the questions slowly. Whenever a participant had difficulty in reading or understanding the questions, the assessment was conducted by interview.

WHOQOL-BREF consists of: first, 26 questions of which 24 are divided into four domains: physical health, psychological, social relationships and environment. Second, overall quality of life score formed by question one: “How would you rate your quality of life?” and question two: “How satisfied are you with your health?”. The score for each domain varies from zero to 20, zero being considered the worst and 20 the best quality of life.[5] For questions one and two the maximum score is 25 points.

SF-36[4] is a multicultural questionnaire comprised of 36 items measuring eight subscales/domains, including physical functioning, role limitations due to physical health problems, bodily pain, general health perceptions, vitality, social functioning, role limitations due to emotional problems, and mental health. Subscale scores range from 0 to 100, with higher scores representing better health.

The validation methods of these two instruments were complex and broad and this paper aims only to understand a convergent validity between them.

Cronbach’s alpha coefficient was used to estimate reliability and Pearson’s correlation for comparison between the two scales. Coefficients r from 0.21 to 0.60 were considered weak correlations magnitudes, according to Rodrigues.[b]

The study was approved by the Ethics in Human Research Committee of the Federal University of Sao Carlos (Process 116/2006).

RESULTS

The 276 participants completed both questionnaires. Women represented 87.8% of the enrolled participants. The mean age was 63.83 (SD = 7.22); the maximum age was 91 years and the minimum 60 years. Score distributions of the WHOQOL-BREF and the SF-36 are summarized in Table 1. Of all the domains, social functioning achieved the highest value for SF-36 and psychological for WHOQOL-BREF. The lowest was vitality for SF-36 and physical health for WHOQOL-BREF. In most domains, the median was higher than the mean – negatively skewed – revealing distributions with more participants scoring among the more favorable states.

Table 1

Score distributions of the instruments WHOQOL-BREF and SF-36. Municipality of Sao Carlos, Southeastern Brazil, 2006-2008.

Variable	Mean	SD	Median	Minimum	Maximum	Lower quartile	Upper quartile
SF-36
PF	77.68	18.5	83.50	13.50	98.50	68.50	93.50
RP	68.79	17.9	80.00	30.00	80.00	55.00	80.00
BP	66.60	23.1	70.80	8.80	98.80	49.80	82.80
GH	71.06	19.3	75.75	15.75	98.75	55.75	85.75
VT	65.27	18.1	63.80	13.80	98.80	53.80	78.80
SF	79.52	20.7	86.25	11.25	98.75	61.25	98.75
RE	65.33	24.3	80.00	13.33	80.00	46.67	80.00
MH	68.92	18.0	70.80	18.80	98.80	58.80	82.80
WHOQOL-BREF
Physical	12.95	2.2	13.14	6.29	17.14	11.43	14.86
Psychological	14.94	2.4	15.33	4.00	19.33	13.33	16.67
Social	13.68	2.3	13.33	4.00	20.00	12.00	14.67
Environmental	14.21	2.4	14.50	6.50	20.00	12.50	16.00
Overall	15.26	2.8	16.00	4.00	20.00	14.00	16.00

PF: physical functioning; RP: role limitations due to physical health problems; BP: bodily pain; GH: general health perceptions; VT: vitality; SF: social functioning; RE: role limitations due to emotional problems; MH: mental health

WHOQOL-BREF Cronbach’s alpha was 0.832 and SF-36 was 0.868, representing acceptable reliability.

Pearson’s correlation results are displayed in Table 2. There were weak correlations magnitudes between the two questionnaires in the following fields: i) WHOQOL-BREF social relationships domain and SF-36 social functioning, ii) WHOQOL-BREF physical health domain and SF-36 role limitations due to physical health problems, iii) WHOQOL-BREF physical domain and SF-36 physical functioning, and iv) WHOQOL-BREF psychological domain and SF-36 role limitations due to emotional problems. Also, the WHOQOL-BREF Question 2 (on general health) showed a weak correlation with SF-36 general health perceptions and role limitations due to physical health problems. WHOQOL-BREF overall score also showed a weak correlation with all SF-36 parameters.

Table 2

r values according to Pearson’s correlation between WHOQOL-BREF and SF-36. Municipality of Sao Carlos, Southeastern Brazil, 2006-2008.

	WHOQOL-BREF
Variable	Q1	Q2	Physical	Psycol	Social	Environ	Overall
SF-36
PF	0.21[a]	0.33[a]	0.58[a,b]	0.35[a]	-0.10	0.32[a]	0.32[a,b]
RP	0.27[a]	0.31[a,b]	0.48[a,b]	0.22[a]	-0.12[a]	0.27[a]	0.34[a,b]
BP	0.23[a]	0.34[a]	0.51[a]	0.22[a]	-0.19[a]	0.26[a]	0.34[a,b]
GH	0.31[a]	0.41[a,b]	0.55[a]	0.29[a]	-0.15[a]	0.29[a]	0.42[a,b]
VT	0.40[a]	0.48[a]	0.64[a]	0.45[a]	-0.03	0.36[a]	0.51[a,b]
SF	0.29[a]	0.31[a]	0.44[a]	0.43[a]	0.17[a,b]	0.31[a]	0.35[a,b]
RE	0.16[a]	0.23[a]	0.33[a]	0.23[a,b]	0.01	0.22[a]	0.23[a,b]
MH	0.46[a]	0.41[a]	0.55[a]	0.54a	0.12	0.39[a]	0.50[a,b]

Physical: physical health; Q1: question 1; Q2: question 2; Psycol: psychological; Social: social relationships; Environ: environmental; PF: physical functioning; RP: role limitations due to physical health problems; BP: bodily pain; GH: general health perceptions; VT: vitality; SF: social functioning; RE: role limitations due to emotional problems; MH: mental health

p ≤ 0,05 according to Pearson’s correlation.

discussed comparisons.

DISCUSSION

The study participants were mostly women, probably because most of the Universities of the Third Age have more participants of this gender.[1,2,c] Therefore, the findings of this study will be discussed taking this female population perspective into consideration.

Both scales showed acceptable reliability in this study. Despite the subjectivity of health-related quality of life and self-reported quality of life, this good reliability means that WHOQOL-BREF and SF-36 can be used to evaluate and propose strategies and policies aimed at aging. Other studies performed with different populations also observed similar results concerning reliability.[6-9,11,12] However, it is important to be cautious in generalizing the results when applying these measures to evaluate one particular patient, since the statistical analyses of these studies are based on large samples and do not consider individual differences.

Our results suggest that the questionnaires’ related domains do not provide similar measures in evaluating Brazilian older women. This result is consistent with the findings of Huang et al,[7] who observed a weak correlation or an absence of correlation between these areas. Higher correlation results were found by Nedjat et al[10] for a population of young Iranians. The divergences between these results may be due to differences in the participants’ age or the discrepancy between the Brazilian and the Iranian versions of the instruments. This study’s results also differ from studies on patients with specific clinical conditions.[6,8] The literature indicates that these instruments have a weak convergent validity for healthy populations,[7,11] and high magnitude correlation for studies with patient groups.[6,8]

The observed divergences may be a consequence of differences in the goals adopted by developers of the instruments. SF-36 measures aspects that are linked to health and functional performance of patients, whereas WHOQOL instruments attempt to measure a broad range of factors concerning the organism, task and environment.[4,5] Nevertheless, it is likely that SF-36 is a more objective measure, because its questions are about capability and disability, whereas WHOQOL-BREF focuses on individual opinions about quality of life.

According to Huang et al,[7] SF-36 allows a better discrimination between health-related known-groups whereas WHOQOL-BREF seems to be a better choice for an overall assessment of quality of life in a population of young individuals. Because this study results observed a weak convergent validity as before,[7] the differences between the characteristics of the instruments appointed can also hold true for an adult population.

Although age-related changes can interfere in quality of life, the presented results indicate that older individuals’ opinion about themselves is more important than health-related quality of life. That is because a person can adapt and develop strategies to handle morbidity or functional decline in aging process, resulting in a happier late life. Therefore, it is considered that WHOQOL-BREF provides a relevant output on changes in older women’s quality of life after an intervention program, because it prioritizes responses to the aging process, without focusing on their impairment.

We studied a convenience sample containing a predominance of women and therefore no comparison with health-related known-groups was performed. Thus, caution is needed when generalizing the results beyond this study group. Further studies are necessary to fully understand the convergent validity for senescent men and patient groups.

This study data suggest that SF-36 and WHOQOL-BREF are reliable for clinical and research uses. Nevertheless, to select one of them, researchers should consider which aspects of quality of life they aim to capture because of the observed weak convergent validity signs. The WHOQOL-BREF can be considered more suitable for this study population because it valued older individuals’ personal opinions.