Cross-cultural adaptation and psychometric validation of the World Health Organization quality of life-old module (WHOQOL-OLD) for Persian-speaking populations

The study sample consisted of 400 older adults (200 Zoroastrian and 200 Muslim individuals) aged 60 years old and above who were registered in the two primary health care center in the city of Yazd, the capital city of Yazd province, center of Iran. The sample size was decided based on the recommended sample size (8 people per each of the instrument’s items) [20] for performing the confirmatory factor analyses. The invited aged people for participation in the study were 480 registred Zoroastrians (260) and Muslim (220) dwellers of them 400 consented to cooperate in the study (particiaption rate (83.33%).

The inclusion criteria were being permenant resident of the city of Yazd, having age equal to 60 or above, being diagnosed with at least one to three simulataneous chronic diseases, ability to speak and understand Persian language, not having considerable hearing loss to make interviews absurd and also not having a major cognitive problems which was assured based on the respondents’ Abbreviated Mental Test scores (AMT) [21]. The study exclusion criteria composed of not willing to continue participation after giving informed consent, progression of a chronic disease that make participation in the study medically inconsistent or development of a severe disability that make the respondent epistemically dependent on others.

Standard translation and back translation procedures were followed after formal permission from the WHO Quality of Life (WHOQOL) assessment group to prepare a preliminary draft of the WHOQOL-OLD-P [22, 23]. The original WHOQOL-OLD was first translated into the Persian language by the two professional translators, one of them was familiar with the study subject. At the next stage the translated draft was back translated into the English language and this back-translated version was compared side by side to the original WHOQOL-OLD in terms of clarity, common language use, conceptual equalization and quality of the overall translation. Any inconsistencies between the original and the translated version was resolved at this stage and the final Persian version was administered to a group of representative older people (n = 10) unfamiliar with the topic, to ask for their comments about vague items or phrases, format of the questions and response items. All the feedbacks were considered to ensure rapport and having an exactly equivalent translated version to the original WHOQOL-OLD.

A panel of experts consisting 15 gerontologists, geriatricians, health education and promotion specialists, psychologists and nursing specialists were contacted for their comments about the simplicity, grammatical soundness, clarity and understandability of the WHOQOL-OLD-P. The Lawshe’s Content validity index (CVI) and content validity ratio (CVR) were estimated based on the panelists feedbacks and values of CVI ≥ 0.8 [24] and CVR ≥ 0.49 [25] were deemed acceptable. To ensure face and qualitative content validity of the WHOQOL-OLD-P it was also pilot tested on the 28 older adults for their comments about understandability of the questions and response items however; no important issue was mentioned.

The Cronbach’s alpha coefficient of internal consistency and the Intra-Class Correlation (ICC) coefficient of reliability were calculated to assess performance of the WHOQOL-OLD-P and reproducibility of the data collection process.

Two trained and qualified interviewers were hired to collect data in face-to-face interviews from the Zoroastrian and their socio-economically matched aged Muslim counterparts who had been registered in the two purposefully selected urban health care centers.

The skewness and kurtosis of the collected data were tested for normal distribution, mean and standard deviation of the WHOQOL-OLD-P’s scores were estimated consequently [26]. The ceiling and floor effects were also checked for detection of imbalance in the percentage of the study participants (greater than 20%) with low and high levels of WHOQOL-OLD-P’s scores and to assess ability of the instrument in discriminating the respondents across the high and low categories of quality of life [10]. The internal consistency reliability coefficient of Cronbach’s Alpha and the intra-class correlation coefficient (ICC) to test the WHOQOL-OLD-P stability over time were estimated and values above 0.7 were deemed satisfactory [27, 28].

To verify compatibility of the study findings with the factor structure reported for the original WHOQOL-OLD module, confirmatory factor analysis (CFA) was performed and the CFA fit indices including Chi-squared/df < 5.00, Standardized Root Mean Square Residual (SRMR) < 0.08, Comparative Fit Index (CFI), and Tucker Lewis index (TLI) > 0.90 were deemed to be acceptable [29]. The IBM SPSS version 24 (IBM Corporation, Software Group, Armonk, NY, USA; 2016) [30], and STATA version 14 (Stata Corporation, College Station, TX: LP; 2015) [31] software were used for data analysis purposes.

Mean age of the study participants was 70.63 ± 8 and their mean WHOQOL-OLD-P’s scores was 77.72 ± 10.41. Other characteristics of the respondents were presented in Table 1.

The study participants’ overall and subscales’ scores of the WHOQOL-OLD-P along with the test-retest reliability and internal consistency coefficients were tabulated in Table 2. The outputs of ceiling and floor effects’ analysis for overall and subscales of the instrument were also presented in this table. As indicated the estimated values are in the range of acceptable range.

The WHOQOL-OLD-P was sent to a panel of experts for their feedbacks about the instrument items’ conceptual, cultural and linguistic appositeness. Based on the panelists’ given scores to the listed items’ attributes on a Likert type scale the CVI and CVR were estimated and the values were in the range of acceptable variability (0.88 and 0.86 respectively).

The conducted CFA in a sample of 400 Muslim and Zoroastrian older adults yielded an acceptable model fit with the six factor solution that was similar to the endorsed model fit for the original WHOQOL-OLD (χ²/df = 1.92, P < 0.001), RMSEA = 0.04, CFI = 0.94, TLI = 0.93, SRMR = 0.06).

The study sample in this research consisted of elder people with at least one to three diagnosed simultaneous chronic diseases who had been registered in the two primary health care centers in the city of Yazd, the capital city of Yazd province, central part of Iran. The purposively selected health care centers and therefore, non-probabilistic sample that was selected for inclusion might be limited representativeness of the study participants and restraint extrapolation and generalizability of the findings. Comparison of the within-subjects perceived QOL measures and association of the factors such as religion, age, gender, level of education and overall health status with the estimated QOL values were not speculated in this study. Therefore; interim data collection and analysis in future corresponding studies could add further detail and explanation about precipitating factors of QOL.

Iran is a multi-cultural, multi-religious and multi-ethnic country and the study findings might not be generalizable to elder people residing in the country’s other provinces and geographical areas due to socio-cultural diversities.

There is also possibility of the reporting bias based on the respondents’ willingness and ability to provide accurate responses especially concerning the length of the WHOQOL-OLD-P, although the concern was attempted to be mitigated by emphasis on the anonymity of the participants and allocation of proper time for interview.

Elder people who suffer from multiple chronic diseases are more likely to develop depression and therefore; tend to report lower quality of life [38]. Such an impediment must be taken into account in interpretation of the study data.

Multidimensional nature of QOL and the effects posed by certain underlying factors warrant further research to examine how these circumstances might relate to self-reported quality of life among older people. QOL assessment by use of measures that actually highlight specific domains e.g. physical, emotional and social functioning or over emphasis on health-related aspect of daily life experiences may also cause bias especially in measurement of elder population’s QOL [39]. Aged people themselves, might have a priority need or interest that could potentially be reflected in their responses to the items of QOL assessment instruments. The judgmental approach (which entails comparison between the respondents’ state in different time frames e.g. present to past) which is salient generally in health related quality of life assessment tools could also further contribute to ambiguity in reliable determination of QOL. A number of other baseline socio-demographic factors e.g. gender, being divorced, unemployed, migrant and bereaved might also have same or even higher impact on several dimensions of perceived QOL in aged populations [39]. All these facets warrants interpretation of the QOL data with caution.

Introduction of a valid and reliable instrument for measurement of quality of life among Iranian and other Persian-speaking elder adults was the primary objective of this study. Taking into account all above mentioned limitations, the WHOQOL-OLD-P application could be recommended as an explicit instrument for measurement of perceived QOL as a proxy measure of aged people’s overall health in the Iranian and other Persian-speaking elder populations across the world. The WHOQOL-OLD-P application could enhance reliability of QOL assessment by a wide range of healthcare providers (HCPs) (e.g. nurses, psychologists, geriatricians) in primary health care centers, secondary and tertiary care providing settings (e.g. nursing homes, day care centers and hospitals). The instrument can be utilized in routine care provision activities or in assessment of influences a specific health condition or morbidity and also subsequent restorative therapeutic intervention might have on perceived overall QOL among the aged people. Other social care providing organizations could also use the WHOQOL-OLD-P for estimation of impacts their policies, services or targeted interventions might have on elder people within the Persian-speaking countries.