Received: 16 November 2020  |  Revised: 28 March 2021  |  Accepted: 4 May 2021 DOI: 10.1002/brb3.2193 D A T A P A P E R Validation of the multidimensional WHOQOL- OLD in Ghana: A study among population- based healthy adults in three ethnically different districts Adote Anum1  | Samuel Adjorlolo2,3 | Charity S. Akotia4 | Ama de- Graft Aikins5,6 1Department of Psychology, University of Ghana, Legon- Accra, Ghana Abstract 2Department of Mental Health, School of Objectives: Study of well- being of older adults, a rapidly growing demographic group Nursing and Midwifery, College of Health in sub-S aharan Africa, depends on well- validated tools like the WHOQOL- OLD. This Sciences, University of Ghana, Legon-A ccra, Ghana scale has been tested on different populations with reasonable validity results but 3Research and Grant Institute of Ghana, has limited application in Africa. The specific goal of this paper was to examine the Accra, Ghana factor structure of the WHOQOL- OLD translated into three Ghanaian languages: Ga, 4Department of Psychology, University of Ghana, Legon-A ccra, Ghana Akan, and Kasem. We also tested group invariance for sex and for type of community 5Institute of Advanced Studies, University (distinguished by ethnicity/language). College London, London, UK Methods: We interviewed 353 older adults aged 60 years and above, selected from 6Regional Institute for Population Studies, University of Ghana, Legon- Accra, Ghana three ethnically and linguistically different communities. Using a cross- sectional design, we used purpose and convenience methods to select participants in three Correspondence Adote Anum, Department of Psychology, geographically and ethnically distinct communities. Each community was made up of University of Ghana P. O. Box LG 84, Legon- selected rural, peri-u rban, and urban communities in Ghana. The questionnaire was Accra, Ghana. Email: aanum@ug.edu.gh translated into three languages and administered to each respondent. Results: The results showed moderate to high internal consistency coefficient and Funding information This project was made possible with factorial validity for the scale. Using confirmatory factor analysis, we found that the financial support from the University of results supported a multidimensional structure of the WHOQOL- OLD and that it did Ghana Research Fund. Grant Number: UGRF/10/ILG- 079/2016-2 017 not differ for males and females, neither did it differ for different ethnic/linguistic groups. Conclusions: We conclude that the translated versions of the measure are adequate tools for evaluation of quality of life of older adults among the respective ethnic groups studied in Ghana. These results will also enable comparison of quality of life between older adults in Ghana and in other cultures. K E Y W O R D S aged, Ghana, quality of life, WHOQOL- OLD This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2021 The Authors. Brain and Behavior published by Wiley Periodicals LLC Brain and Behavior. 2021;11:e02193. wileyonlinelibrary.com/journal/brb3  |  1 of 9 https://doi.org/10.1002/brb3.2193 2 of 9  |     ANUM et Al. measure that has cross- cultural relevance. Considering the applica- Keypoints bility of these two instruments for older adults, the WHOQOL re- search group developed a WHOQOL-O LD module for older adults, 1. The WHOQOL_OLD has factorial validity in a cross- containing six facets (Power et al., 2005). This measure has been cultural context. translated into several language versions with reasonable psycho- 2. Translation into multiple languages does not affect the metric properties (Eser et al., 2010; Fleck et al., 2006). factor structure. The World Health Organization quality of life instrument for older 3. There are no differences in sex on the issue of quality of adults largely assesses multifaceted quality of life and psychological life. well- being. The measure has been used to accurately distinguish be- 4. It appears death and dying is an uncomfortable subject tween depressed patients and patients in remission (Hussenoeder among older adults. et al., 2020; Skevington et al., (2020)) or healthy patients (Bonicatto et al., 2001). Although the measure provides adequate validity coef- ficients, there are differences that result from cultural specificities. It 1  | INTRODUC TION was therefore recommended that it is important to validate the tool within each cultural context (Fleck et al., 2006; Power et al., 2005). The rapid pace of economic development in low- and middle- income We designed the study to examine the validity of the instrument countries (LMICs) has resulted in demographic shifts from younger in a multiethnic and multilingual population. We therefore tested populations to a more aged population. This is fueled largely by im- the factorial validity of the WHOQOL- OLD among a cross section proved health care and increase in life expectancy in many LMICs of healthy older adults in Ghana. Specifically, (1) we tested the as- (Ahmad, 2016; Prina et al., 2020). Improved health of the population sumption that the underlying dimensions of the measure would be in LMICs is accompanied by improved quality of life which also means confirmed, and we also examined whether (2) the WHOQOL-O LD that more people are living into advanced old age (Ahmad, 2016; is invariant for males and females and (3) invariant for three ethno- Gyasi & Phillips, 2020). The demographic shift in life expectancy linguistic groups. has caused changes in the disease burden profiles of LMICs, with It was our expectation that the use of the WHOQOL-O LD among chronic noncommunicable diseases (NCDs) becoming a more com- Ghanaians will yield a concise evaluation of older people's rating of mon and growing public health challenge (Aikins & Agyeman, 2017). their quality of life and furthermore provide caregivers, healthcare Consistent with this change, governments’ concerns are moving providers, and potentially policy makers with a more holistic idea of toward developing comprehensive policies on provision of inter- what older people need in order to have quality of life. ventions that meet the health needs of older adults. An effective comprehensive policy is developed on the basis of accurate need- based scientific research. There is increasing effort to provide this 2  | METHODOLOGY need-b ased research globally, especially on physical health needs of the aged. 2.1 | Research setting and sampling In LMICs, especially in sub- Saharan African countries like Ghana, the research is gradually shifting from physical health needs research The sample for this study was selected using a multi-s tage process to quality of life and mental health research, but this change is slow that began with selection of districts and then households. Three and therefore limited information is available on psychological health districts were purposively selected: Accra (Ga West district) in the needs of the older adults (Aikins & Apt, 2016). This gap has slowed south, Sunyani (Sunyani East and West districts) in the middle belt, the ability to generate evidence- based policies and interventions to and Navrongo in the Upper East Region. Three factors guided the meet the psychological health needs of the increasing adult popula- selection of the districts. First, the districts are geographically and tion. The major reason for this is that behavioral and mental health ethnically distinct, and they have easily accessible rural, peri-u rban, research in LMICs is dogged partly by limited and skewed allocation and urban communities. Second, their locations in the South, Middle, of funding resources (Anum et al., 2020). One important element for and the Northern- most parts of the country ensured that the sample research in quality of life and mental health is the use of contextually for the study was close to a nationally representative sample. Third, validated tools that are accurate and allow for cross-c ultural compa- all three districts had their unique languages which then allowed us rability. Mental health intervention policies for older adults will re- to examine the factorial validity of the quality of life scale in ethno- quire extensive investment into accurate assessment and diagnosis linguistically different groups. The selected ethnic groups are three of psychological well- being and/or psychiatric morbidity. of the six major ethnic groups in Ghana. Each selected district has a In response to developing cross-c ulturally validated measure of dominant language although other languages may be spoken within well- being, the World Health Organization Quality of Life Group has the districts. developed the quality of life measures, the WHOQOL-1 00, and its The criteria for inclusion were that the participants had to be short form to address the issue of measurement of quality of life 60 years and above and did not have any signs of ill- health that could (WHOQOL Group, 1993, 1998). Another purpose is to develop a impede their ability to participate in the interview. In each selected ANUM et Al.      |  3 of 9 district, a community facilitator who lives in and knows the com- in older population cohorts (Power et al., 2005). As indicated, there munity very well was hired to help stratify the communities in the are six facets or domains, which are Sensory abilities, Autonomy, districts in order to select participants from all different sections of Past, present, and future, Social interaction, Death and dying, and the selected district. In each selected locality, any household with an Intimacy. Each facet is measured by four items. The original version, individual who was 60 years and older was contacted. The distribu- designed to assess quality of life cross- culturally in health and health tion of the sample is presented in Table 1. care, the WHOQOL, has 100 items. We also asked questions about health status and other demographic characteristics required for the study. 2.2 | Measures The World Health Organization Quality of Life- Old (WHOQOL- 2.3 | Translation procedure OLD) is a 24- item, 6- facet instrument with cross-c ultural reliability (Power et al., 2005; Van Biljon et al., 2015). This was developed by The standard translation and back translation methods were used. the World Health Organization Quality of Life group, a collaborative First, the original WHOQOL- OLD was translated into the three re- effort among numerous researchers from various countries which spective local languages: Ga in Accra, Akan in Sunyani, and Kasem led to the development of a measure focusing on the quality of life in Navrongo, following WHO translation guidelines for assessment of instruments (Üstun et al., (2005)). Second, the translated versions were back translated into the English language by other language TA B L E 1   Descriptive statistics for key variables in the study participants experts who were not familiar with the original English version. The third step involved the evaluation of the back- translated Measures (N) Percent Mean SD Min Max versions, comparing them to the original WHOQOL- OLD by the Study site first author who is literate in Akan and Ga languages. In the case of Ga West (111) 31.40 Kasem, the back- translated version was evaluated with one of the Navrongo (120) 34.00 research assistants. During this phase, the first author corrected any Sunyani (122) 34.60 discrepancies, focusing on contextual and linguistic meaning. Items Age 71.65 9.22 60 85 that lacked clarity were referred to the translators. The final phase Sex involved a discussion of contextual and linguistic equivalence of the items during training. The first author who did the training led Female (229) 67.20 the discussion on the items, and when there was no consensus, the Male (112) 32.80 item(s) was referred to the translators. Marital status Married (125) 35.40 Unmarried (228) 64.60 2.4 | Data collection procedure Employment status Employed (96) 27.30 Eight research assistants were trained for the study; five had de- Unemployed (257) 72.50 grees in psychology, one had a degree in sociology, one had a degree Income status in social work, and one had a degree in education. The training of Regular income 29.70 research assistants was in two phases. The first phase involved train- (105) ing on the original English version. In the second phase, the research Nonregular 70.30 assistants were trained on the translated versions of the question- income (248) naire. During this phase, the research assistants and the first author WHOQOL- OLD Domains had discussions about contextual and linguistic accuracy and items Sensory abilities 13.25 4.36 4 20 on which there was no consensus were referred to the translators. (352) We pretested the questionnaire in a sample of 25 older adults Autonomy (353) 15.62 3.74 4 20 in a peri- urban town in the northern part of Accra. This is a typical Past, present, 15.14 3.61 4 20 Ghanaian community which shared similar characteristics with the future (351) communities for the main study. There were minimal modifications Social interaction 12.99 4.69 4 20 to item translations following feedback from the participants. For (351) example, one item on the Death and Dying dimension— “Fear pain Death and 5.03 3.16 3 15 before death”— was deleted after multiple translations could not re- dyinga  (347) sult in consensus on meaning of the item. Intimacy (351) 15.75 3.78 4 20 All participants completed the WHOQOL- OLD scale and a de- aBased on three items mographic questionnaire that included questions about age, marital 4 of 9  |     ANUM et Al. TA B L E 2   Item means, standard deviations, and internal consistency coefficients of the WHOQOL- OLD items Cronbach's Item Number Domain Description Mean SD Skewness Kurtosis Alpha 1 Sensory abilities Impairment to senses affect daily life 3.33 1.19 −0.498 −0.736 0.937 20 Rate sensory functioning 3.35 1.18 −0.403 −0.839 2 Loss of sensory abilities affect 3.35 1.18 −0.625 −0.602 participation in activities 10 Problems with sensory functioning 3.27 1.16 −0.464 −0.723 affect ability to interact 3 Autonomy Freedom to make own decisions 4.16 0.967 −1.536 2.437 0.885 4 Feel in control of your future 3.69 1.044 −0.743 0.184 11 Able to do things you'd like to do 3.63 1.225 −0.844 −0.253 5 People around you are respectful of 4.14 1.086 −1.396 1.305 your freedom 19 Past present future Happy with things to look forward to 3.68 1.036 −0.780 0.184 0.844 12 Satisfied with opportunities to 3.52 1.077 −0.718 −0.067 continue achieving 13 Received the recognition you deserve 3.97 1.381 −1.070 0.352 in life 15 Satisfied with what you've achieved 3.96 1.304 −1.165 0.720 in life 16 Social interaction Satisfied with the way you use your 3.49 1.086 −0.524 −0.348 0.908 time 17 Satisfied with level of activity 3.52 0.899 −0.470 −0.001 14 Have enough to do each day 3.61 1.032 −0.695 −0.082 18 Satisfied with opportunities to 3.69 0.981 −0.672 0.002 participate in the community 6 Death & dying Concerned about the way you will die 2.05 1.382 0.871 −0.797 0.920 7 Afraid of not being able to control 1.50 1.084 2.049 2.797 Death 8 Scared of dying 1.48 1.417 2.260 3.588 9a  Fear pain before death 21 Intimacy Feel a sense of companionship in life 3.72 0.995 −0.799 0.351 0.930 22 Experience love in your life 3.95 0.904 −1.024 1.217 23 Opportunities to love 4.06 0.900 −1.134 1.431 24 Opportunities to be loved 4.02 0.916 −0.997 0.881 aThis item was deleted from the main study because of poor internal consistency indicator during pretesting of the questionnaire. status, living arrangements, occupational, and income statuses. We 2.6 | Statistical analyses administered the questionnaire in the dominant language for each district. Interviews were conducted by trained interviewers (research The questionnaires were coded into SPSS Version 24. The data were assistants) with the use of the questionnaire as described in measures. managed and analyzed using this software. The Cronbach's alpha coefficients were estimated to assess internal consistency. A series of confirmatory factor analysis (CFA) was also conducted to test for 2.5 | Ethical issues the theorized model and factorial structure and to test for group invariance. The research received ethical approval from the Ethical Committee for Humanities of the University of Ghana. The study number is ECH 105/17-1 8. The research was done in compliance with ethical 3  | RESULTS requirements in Ghana. All participants signed or thumb-p rinted an informed consent form. Participants received either phone credits The descriptive statistics of background characteristics of the study or cakes of soap valued at five Ghana cedis (approximately $1.00). sample are presented in Table 1. The sample is made up of 353 ANUM et Al.      |  5 of 9 older adults (> 60 years of age) living in rural, peri-u rban, and urban The correlation analysis showed there were moderate correla- communities in three districts (Ga West, Navrongo, and Sunyani, in tions among the WHOQOL subdimensions. The correlation with Ghana). death and dying was the lowest. The results for correlations are pre- Majority of the respondents were females (67.2%), and the av- sented in Table 3. erage age was 71.65 years. Almost two- thirds of the respondents were not married or lived with a regular partner. More than 70% were not employed or on a regular income which is expected of the 3.1 | Confirmatory factor analysis demographic group studied. Internal consistency was measured using the Cronbach alpha co- Confirmatory factor analysis (CFA) was used to estimate the fac- efficient, and they were all found to be reasonably acceptable ranging tor structure of the WHOQOL in the full sample. This was followed from 0.844 to 0.937 for the WHOQOL domains. These are considered by multi- group CFA to determine measurement invariance of the moderate to high coefficients (Cortina, 1993). Death and dying has the WHOQOL based on sex and geographical location. In the first of series lowest mean score. As indicated earlier, one item, “Fear pain before of analyses, separate models were estimated for males and females, fol- death,” was not included in the main study because its inclusion resulted lowed by estimation of unconstrained (baseline) model. In this model, in low psychometric coefficients across all the sites. Two other items on the parameters were freely estimated across the groups, with satisfac- this domain— “Afraid of not being able to control Death” and “Scared tory fit indices indicating the attainment of a configural invariance. In of dying”—w ere within acceptable Skewness and Kurtosis limits of 2.0 subsequent model estimations, some constraints were introduced, with (George & Mallery, 2019). Intimacy has the highest mean score. each successive model containing all the constraints of its predecessor. TA B L E 3   Correlations (Pearson r) among key constructs Constructs 2 3 4 5 6 7 1 Sensory abilities 0.621** 0.504** 0.600** −0.143* 0.354** 0.384** 2 Autonomy 0.635** 0.676** −0.114* 0.550** 0.519** 3 Past, Present, Future 0.616** −0.154** 0.557** 0.634** 4 Social Interaction 0.069 0.527** 0.455** 5 Death & Dying −0.028 0.170** 6 Intimacy 0.422** 7 Overall WHOQOL- OLD score *=0.05.; **=0.001. TA B L E 4  Confirmatory factor analysis and sex invariance of WHO- quality of life questionnaire Model/Fit Indices χ2(df) χ2/df TLI CFI RMSEA AIC BIC Δχ2 ΔCFI Full sample Original 576.98(160)*** 3.61 0.92 0.94 0.09 676.98 870.31 - Respecified 418.73(157)*** 2.67 0.95 0.96 0.06 524.73 729.66 158.25(3)*** - Males Original 320.43(160)*** 2.01 0.90 0.92 0.09 420.73 557.10 Respecified 284.86(158)*** 1.80 0.92 0.93 0.08 388.86 530.68 35.57(2)*** Females Original 539.33(160)*** 3.37 0.90 0.92 1.00 639.33 813.36 Respecified 424.16(157)*** 2.70 0.93 0.94 0.08 530.16 714.63 115.17(1)*** Sex Invariance Unconstrained 700.30(314)*** 2.23 0.93 0.94 0.06 - - - - Constrained 1 715.53(329)*** 2.18 0.93 0.94 0.06 15.23(15), ns 0.00 Constrained 2 730.24(342)*** 2.14 0.93 0.94 0.06 14.71(13), ns 0.00 Constrained 3 813.12(367)*** 2.22 0.93 0.93 0.06 82.88(25)*** −0.01 Note: Unconstrained = parameters freely estimated; Constrained 1 = factor loadings constrained; Constrained 2 = Factor variances and covariances constrained; Constrained 3 = Error variances constrained; ns = not significant. *p < .001.; **p < .001.; ***p < .001. 6 of 9  |     ANUM et Al. TA B L E 5  Confirmatory Factor Analysis and Ethnic(Regional) Invariance of the WHO-Q uality of Life Questionnaire Model/Fit Indices χ2(df) χ2/df TLI CFI RMSEA AIC BIC Δχ2 ΔCFI Full sample Original 576.98(160)*** 3.61 0.92 0.94 0.09 676.98 870.31 - Respecified 418.73(157)*** 2.67 0.95 0.96 0.06 524.73 729.66 158.25(3)*** - Accra Original 320.03(160)*** 2.00 0.90 0.92 1.00 420.03 555.10 Respecified 279.27(157)*** 1.78 0.92 0.94 0.08 385.27 528.87 40.76(3)*** Sunyani Original 304.81(160)*** 1.91 0.91 0.93 0.08 404.81 545.01 Respecified 293.52(157)*** 1.87 0.91 0.93 0.08 399.52 548.13 11.29(3)*** Navrongo Original 499.64(160)*** 3.12 0.84 0.87 0.13 599.64 739.01 Respecified 325.17(155) 2.09 0.92 0.93 0.09 435.17 588.48 174.47(5)*** Regional Invariance Unconstrained 891.37(465)*** 1.92 0.92 0.93 0.05 - - - - Constrained 1 970.28(495)*** 1.96 0.92 0.93 0.05 78.91(30)*** 0.00 Constrained 2 1,072.91(525)*** 2.04 0.91 0.92 0.06 106.63(30)*** 0.00 Constrained 3 1,428.65(575)*** 2.48 0.87 0.87 0.07 355.75(50)*** −0.01 Note: Unconstrained = parameters freely estimated; Constrained 1 = factor loadings constrained; Constrained 2 = Factor variances and covariances constrained; Constrained 3 = Error variances constrained. *p < .001.; **p < .001.; ***p < .001. First, the factor loadings were held constant across the groups to in- model fit to the data (TLI = 0.92; CFI = 0.94; RMSEA = 0.09). Based vestigate metric invariance, followed by covariance and variances, and on the modification indices (MI), the model was respecified by al- lastly error variances were held to be equal for the groups to determine lowing the residuals of the following items to correlate: Intimacy invariance of the covariance and variances, and error variances, respec- #22 (Experience love in your life) and Past, present and future #15 tively. To determine sex invariance, differences in comparative fit index (Satisfied with what you've achieved in life) (MI = 92.92), Intimacy (CFI; ΔCFI) and chi- square (χ2; Δχ2) were used. A nonsignificant Δχ2 #11 (Feel a sense of companionship in life) and Intimacy #21 and ΔCFI ≥ −0.01 between the restrictive and less restrictive or un- (Experience love in your life ) (MI = 25.09), and Sensory abilities constrained models indicate the attainment of sex invariance. Regional #22 (Rate of sensory functioning) and Sensory abilities #23 (Loss of invariance determination followed the procedure above. sensory abilities affect participation in activities) (MI = 15.55). The Model fit was determined using the following common fit indi- results showed that respecified model was an improvement over the cators: χ2, CFI, Tucker– Lewis Index (TLI), and a noncentrality- based initial model (Δχ2 = 158.25, p < .001), providing a good model fit index, the root mean square error of approximation (RMSEA). The (TLI = 0.95; CFI = 0.96; RMSEA = 0.06). CFA and multi-g roup CFA were conducted with maximum likelihood Sex specific analyses also revealed that the models in which the estimation method in Analysis of Moment Structures (AMOS) ver- residuals of the aforementioned items correlate freely showed sig- sion 21 (Arbuckle, 2011). nificant model improvement over the initial models. The results of Preliminary analysis in the CFA revealed that the model did not sex invariance analyses based on the respecified models indicate provide a good fit to the data (TLI = 0.74; CFI = 0.87; RMSEA = 0.10). that configural invariance (TLI = 0.93; CFI = 0.94; RMSEA = 0.06), Inspection of the items constituting the various dimensions of the metric invariance (ΔCFI = 0.00), and invariance of factor variance scale showed that the items for Death and Dying loaded poorly (i.e., (ΔCFI = 0.00) and (ΔCFI = −0.01) have been attained. The factor ≤. 15) and were not significant (p >.05). A decision was reached to structure from the CFA model and their corresponding coefficients exclude the Death and Dying dimension from further analysis. are summarized in the Figure 1. 3.2 | Confirmatory factor analysis and 3.3 | Confirmatory factor analysis and sex invariance location invariance The CFA and sex invariance analysis results are summarized in Table 5 provides a summary of the CFA and ethnicity (regional) invar- Table 4. The initial model for the full sample did not provide a good iance of the WHOQOL questionnaire. Consistent with the findings ANUM et Al.      |  7 of 9 F I G U R E 1  CFA of the WHOQOL- OLD, based on data collected for the full sample. Note, SAB— Sensory Abilities (s11- Impairment to senses affect daily life, s21- Rate sensory functioning, s22- Loss of sensory abilities affect participation in activities, s23- Problems with sensory functioning affect ability to interact), AUT— Autonomy (a1-F reedom to make own decisions, a2- Feel in control of your future, a3-A ble to do things you'd like to do, a4- People around you are respectful of your freedom), PPF— Past, Present, Future (pp11- Happy with things to look forward to, pp21- Received the recognition you deserve in life, pp22- Satisfied with opportunities to continue achieving, pp23- Satisfied with what you've achieved in life), SOP— Social Participation (sp11- Satisfied with the way you use your time, sp21- Satisfied with level of activity, sp22- Have enough to do each day, sp23- Satisfied with opportunities to participate in the community), INT— Intimacy (in11- Feel a sense of companionship in life, in21- Experience love in your life, in22- Opportunities to love, in23- Opportunities to be loved) in Table 4, the full sample model was respecified to achieve good 4  | DISCUSSION model fit by allowing the error variances of the following items to correlate: Past, present, and future items #22 (Satisfied with oppor- The primary aim of this paper is to establish the factorial validity tunities to continue achieving) and #23 (Received the recognition of the 24- item WHOQOL for older adults on three selected sam- you deserve in life) (MI = 92.92), Intimacy item #21 (Feel a sense of ples from Ghana. Our overarching research question was, therefore, companionship in life) and Intimacy item #22 (Experience love in your whether or not the WHOQOL- OLD can be used as a reliable and life) (MI = 25.09), and sensory abilities items #22 (rate of sensory valid instrument for measuring quality of life among individuals living functioning) and #23 (Loss of sensory abilities affect participation in in three geographically and linguistically different districts in Ghana. activities) (MI = 15.55). As can be seen, the resulting models were This measure has been translated into over 15 languages and within good fit to the data (TLI = 0.95; CFI = 0.96; RMSEA = 0.06). The those contexts provided a good basis for the measurement of quality initial models for participants from Accra and Sunyani were respeci- of life among older adults (Eser et al., 2010; Fleck et al., 2006). fied based on the modifications above, leading to an improvement In this study, we compared results of three culturally different in model fit. For Accra, the fit of the respecified model is as follows: groups using data from three translated versions of the instru- (TLI = 0.92; CFI = 0.94; RMSEA = 0.08). In addition to the above, ment. We found reasonable internal consistency coefficients for the model for participants from Navrongo was respecified by allow- the WHO- QOL OLD subdimensions. The reliability coefficients as- ing the residuals of items for Social interaction items #22 (Satisfied sociated with the six subdimensions point to a reliable instrument, with level of activity) and #23 (Satisfied with opportunity to par- showing comparable coefficients in a sample of Afrikaans- speaking ticipate in community); and Autonomy items #2 (Feel in control of population in South Africa (Van Biljon et al., 2015). your future) and #4 (Able to do things you'd like) to correlate freely. Cross-c ultural assessment issues have been advanced over The resulting model improved over the initial model and provided a the years given the influence of culture on the assessment of good model fit to the data (TLI = 0.92; CFI = 0.93; RMSEA = 0.09). psychosocial constructs. In response, we tested whether the Measurement invariance analyses also revealed that configural in- theoretical factor structure of the WHOQOL-O LD and the in- variance (TLI = 0.92; CFI = 0.93; RMSEA = 0.05), metric invariance variance of same across sex and geographical location would (ΔCFI = 0.00), and invariance of factor variance (ΔCFI = 0.00) and be similar to those reported in previous studies, notably Power error variances (ΔCFI = −0.01) have been attained. and Quinn (2006). In general, the findings of the study have 8 of 9  |     ANUM et Al. revealed that quality of life among adults in Ghana can be for this study was slightly less than optimal. Estimating sample represented by the dimensions stipulated by the WHOQOL- size for structural equation modeling depends on a number of OLD. More importantly, the multidimensional structure of the factors such as number of indicators, number of factors, mag- WHOQOL-O LD is invariant across sex and three different eth- nitude of factor loadings, and magnitude of factor correlations nic and linguistic groups. This finding lends support to previous (Wolf et al., 2013) most of which were not available at the time of studies that have used the WHOQOL-O LD in different coun- determining sample size. Notwithstanding the invariance of the tries, including in South Africa (Van Biljon et al., 2015), Norway factor structure across different geographical locations in Ghana, (Halvorsrud et al., 2008), Brazil (Chachamovich et al., 2008), Iran the generalizability of the study findings to persons in other re- (Rezaeipandari et al., 2020), South Korea (Kim et al., 2020), and gions not sampled for the study may be limited. Ghana is multi- Singapore (Suárez et al., 2018). The confirmation of the structure ethnic and multilingual with at least six broad language groups of the modified version of the WHOQOL- OLD in the Ghanaian (Dakubu, 2015). We studied only three of these languages, and samples has largely extended the multidimensional concept of therefore, we do not make assumptions about generalizing the quality of life beyond Western and non- Western samples on findings to the rest of the country. The translation of the assess- which previous studies were validated. Although previous stud- ment measures to the dominant languages spoken in the selected ies reported a six factor WHOQOL-O LD consistent with the geographical location followed the standard translation ap- original scale (e.g., in Iran and South Korea), our findings support proaches recommended in previous research. In spite of this, we a five- factor structure. We assume that this may be in part to are aware there is the possibility of translation and administration translation issues or to the discussion of a subject matter that errors that could influence participants’ response. is difficult for older adults. These may be tentative assumptions It should be noted that the structure of the WHOQOL- OLD was that need further exploration. It is important, however, to men- confirmed in the Ghanaian sample following model modifications tion that in one study, it is reported that responses about items based on the modification indices. This partly raises concerns about on death and dying were ambiguous, with individuals showing the stability of the structure of the WHOQOL- OLD- R in a similar fear and resignation about death (Melo et al., 2018). sample. The findings of this and previous studies suggest there are Notwithstanding the sociocultural and geopolitical factors (be- problems with the WHOQOL- OLD dimension “Death and Dying” tween and within countries) that exert influence on behaviors, the providing inadequate model fit (Chachamovich et al., 2008) and hav- study has largely re- echoed the notion that the experience and en- ing lower correlations with other dimensions of the measure (Power dorsement of quality of life as a salient psychosocial construct can be et al., 2005; Van Biljon et al., 2015). similar, in accordance with the concept of universality of certain be- havioral repertoires (Adjorlolo et al., 2018). Invariance across sex and geographical locations implies that performance on the WHOQOL- 6  | CONCLUSION OLD may not be biased (i.e., under or overestimated) by sex and geo- graphical locations. Therefore, any mean- level difference based on This study adds to the growing number of studies that have shown the sex of the participants and/or their geographical location on qual- that the WHOQOL_OLD has cross-c ultural validity. However, the ity of life could not be attributed to the biases of the WHOQOL- OLD finding that the death and dying dimension did not measure the for one group (Anum et al., 2019). In relation to geographical location, construct adequately provides impetus for additional studies on the this finding is particularly attractive in that the WHOQOL- OLD could underlying factor structure of the WHOQOL- OLD in the Ghanaian be administered to samples from various regions in Ghana with es- population or similar populations. Given that death and dying is a sentially unique and distinct cultural practices. major theme for the aged, future research in Ghana or on a similar A cross-c ultural measure proven to have sound psychometric population should consider reconceptualizing the items to align with properties will be useful in studying quality of life and mental health Ghanaian cultural perspectives on death and dying. For example, Van needs and to provide the opportunity for comparative analysis in der Geest, in a series of studies among rural populations, has found Ghanaian contexts and across other LMICs. that Ghanaians make reference to and preference for “good death,” dying peacefully and naturally in one's old age (Van der Geest, 2004). 5  | LIMITATIONS CONFLIC T OF INTERE S T The authors report no conflict of interest. There are a number of limitations that need to be considered when interpreting the results from this study. Social desirabil- CONSENT FOR PUBLIC ATION ity and other biases may influence participants’ response to the All authors have given consent for publication. study measures. This possibility is heightened because self- report measures primarily do not provide mechanisms to ascertain the DATA AVAIL ABILIT Y S TATEMENT accuracy of participants’ response, compared with task- based as- The data that support the findings of this study are available from sessment modalities. The sample size for each district selected the corresponding author, AA, upon reasonable request. ANUM et Al.      |  9 of 9 ORCID 75– 83. Published online. https://doi.org/10.1007/s10433 - 020- Adote Anum https://orcid.org/0000-0001-8680-9203 00573 -8 Kim, H. Y., Nho, J. H., Kim, J. Y., & Kim, S. R. (2020). Validity and re- liability of the Korean version of the world health organization R E FE R E N C E S quality of life instrument-o lder adults module. Geriatric Nursing, Adjorlolo, S., Asamoah, E., & Adu- Poku, S. (2018). Predicting delin- 42(2), 548– 554. Published Online, https://doi.org/10.1016/j.gerin quency by self-r eported impulsivity in adolescents in Ghana. urse.2020.10.006 Criminal Behaviour and Mental Health, 28(3), 270– 281. https://doi. Melo, R. L. P. D., Silva Júnior, E. G. D., Souto, R. Q., Leão, Í. S., & Eulálio, M. org/10.1002/cbm.2064 D. C. (2018). Psychometric properties of the complete version of the Ahmad, K. A. (2016). Global, regional, and national disability- adjusted World Health Organization Quality of Life Assessment (WHOQOL- life- years (DALYs) for 315 diseases and injuries and healthy life ex- OLD): Reduced response scale. Psicologia: Reflexão e Crítica, 31, 785– pectancy (HALE), 1990– 2015: A systematic analysis for the Global 791. https://doi.org/10.1186/s41155 - 018-0 084- 1 Burden of Disease Study 2015. The Lancet, 388(10053), 1603– 1658. Power, M., Quinn, K., & Schmidt, S. (2005). Development of the https://doi.org/10.1016/S0140 - 6736(16)31460 - X WHOQOL- old module. Quality of Life Research, 14, 2197– 2214. Aikins, A. D. G. & Agyeman, C. (2017). Chronic non-c ommunicable dis- https://doi.org/10.1007/s1113 6- 005-7 380-9 eases in low- and middle- income countries: Concepts and strategies Prina, A. M., Wu, Y. T., Kralj, C., Acosta, D., Acosta, I., & Guerra, M. for prevention, control and advocacy. CAB Reviews: Perspectives in (2020). Dependence-a nd disability- free life expectancy across Agriculture, Veterinary Science, Nutrition and Natural Resources, 12(027), eight low- and middle- income countries: A 10/66 study. Journal of 1–8 . https://doi.org/10.1079/PAVSN NR201 712027 Aging and Health, 32(5– 6), 401–4 09. https://doi.org/10.1177/08982 Aikins, A. D. G., & Apt, N. (2016). Ageing in Ghana: Setting priorities for 643198 25767 research, practice and policy. Ghana Stud J., 19(1), 35– 45. https://doi. Rezaeipandari, H., Morowatisharifabad, M. A., Mohammadpoorasl, A., org/10.1353/ghs.2016.0002 & Shaghaghi, A. (2020). Cross- cultural adaptation and psychometric Anum, A., Adjorlolo, S., & Kugbey, N. (2019). Depressive symptom- validation of the World Health Organization quality of life-o ld mod- atology in adolescents in Ghana: Examination of psychomet- ule (WHOQOL-O LD) for Persian- speaking populations. Health and ric properties of the Patient Health Questionnaire- 9. Journal Quality of Life Outcomes, 18(1), 1– 7. https://doi.org/10.1186/s1295 of Affective Disorders, 256, 213–2 18. https://doi.org/10.1016/j. 5- 020- 01316- 0 jad.2019.06.007 Skevington, S. M., Rowland, C., Panagioti, M., Bower, P., & Krägeloh, C. Anum, A., Washington- Nortey, M., & Dzokoto, V. (2020). Strategic (2020). Enhancing the multi-d imensional assessment of quality of planning in LAMIC mental health research: A Ghana case study. life: Introducing the WHOQOL-C ombi. Quality of Life Research, 30(3), International Journal of Mental Health, 49(2), 128– 156. https://doi. 1– 13. https://doi.org/10.1007/s11136 - 020-0 2661 - 9 org/10.1080/002074 11.2020.1719621 Suárez, L., Tay, B., & Abdullah, F. (2018). Psychometric properties of Arbuckle, J. L. (2011). IBM SPSS Amos 20 User’s Guide. Amos Development the World Health Organization WHOQOL- BREF quality of life as- Corporation. SPSS Inc. sessment in Singapore. Quality of Life Research, 27(11), 2945– 2952. Bonicatto, S. C., Dew, M. A., Zaratiegui, R., Lorenzo, L., & Pecina, P. https://doi.org/10.1007/s11136 - 018-1 947-8 (2001). Adult outpatients with depression: Worse quality of life Üstun, T. B., Chatterji, S., Mechbal, A., & Murray, C. (2005). Quality assur- than in other chronic medical diseases in Argentina. Social Science ance in surveys: Standards, guidelines and procedures. Household and Medicine, 52(6), 911–9 19. https://doi.org/10.1016/S0277 sample surveys in developing and transition countries. In: United -9 536(00)00192- 1 Nations Statistical Division & United Nations Department of Economic Chachamovich, E., Fleck, M. P., Trentini, C., & Power, M. (2008). Brazilian and Social Affairs (pp. 199– 230). (Eds.). United Nations. WHOQOL- OLD Module version: A Rasch analysis of a new in- Van Biljon, L., Nel, P., & Roos, V. (2015). A partial validation of the strument. Revista de Saude Publica, 42(2), 308– 316. https://doi. WHOQOL- OLD in a sample of older people in South Africa. Glob org/10.1590/S0034 - 891020 0800 0200017 Health Action, 8(1), 28209. https://doi.org/10.3402/gha.v8.28209 Cortina, J. M. (1993). What is coefficient alpha? An examination of theory Van der Geest, S. (2004). Dying peacefully: Considering good death and and applications. Journal of Applied Psychology, 78(1), 98. https://doi. bad death in Kwahu- Tafo, Ghana. Social Science & Medicine, 58(5), org/10.1037/0021-9 010.78.1.98 899– 911. https://doi.org/10.1016/j.socsc imed.2003.10.041 Dakubu, M. E. K. (2015). The Languages of Ghana. Routledge. WHOQoL Group (1993). Study protocol for the World Health Eser, S., Saatli, G., Eser, E., Baydur, H., & Fidaner, C. (2010). The reliability Organization project to develop a Quality of Life assessment instru- and validity of the Turkish version of the World Health Organization ment (WHOQOL). Quality of Life Research, 2, 153– 159. https://doi. quality of life instrument- older adults module (WHOQOL- Old). org/10.1007/BF004 35734 Turkish Psikiyatr Dergisi, 21(1), 1. WHOQol Group (1998). Development of the World Health Organization Fleck, M. P., Chachamovich, E., & Trentini, C. (2006). Development and WHOQOL-B REF quality of life assessment. Psychological medicine, validation of the Portuguese version of the WHOQOL- OLD module., 28(3), 551– 558. 40. Revista de Saude Publica, 40, 785– 791. Wolf, E. J., Harrington, K. M., Clark, S. L., & Miller, M. W. (2013). Sample size George, D., & Mallery, P. (2019). IBM SPSS Statistics 26 Step by Step: A requirements for structural equation models: An evaluation of power, Simple Guide and Reference. Routledge. bias, and solution propriety. Educational and Psychological Measurement, Gyasi, R. M., & Phillips, D. R. (2020). Demography, socioeconomic status 73(6), 913–9 34. https://doi.org/10.1177/001316 4413 495237 and health services utilisation among older Ghanaians: Implications for health policy. Ageing International, 45, 50– 71. https://doi. org/10.1007/s12126 -0 18- 9343- 9 How to cite this article: Anum A, Adjorlolo S, Akotia CS, Halvorsrud, L., Kalfoss, M., & Diseth, Å. (2008). Reliability and va- de- Graft Aikins A. Validation of the multidimensional lidity of the Norwegian WHOQOL- OLD module. Scandinavian Journal of Caring Sciences, 22(2), 292–3 05. https://doi. WHOQOL- OLD in Ghana: A study among population- based org/10.1111/j.1471- 6712.2007.00523.x healthy adults in three ethnically different districts. Brain Hussenoeder, F. S., Jentzsch, D., Matschinger, H., Hinz, A., Kilian, R., Behav. 2021;11:e02193. https://doi.org/10.1002/brb3.2193 Riedel- Heller, S. G., & Conrad, I. (2020). Depression and quality of life in old age: A closer look. European Journal of Ageing, 18(1),