You are viewing the site in preview mode

Skip to main content
Download PDF

Prevalence of anxiety symptoms in a Ugandan population sample and psychometric properties of the Generalized Anxiety Disorder-7 scale (GAD-7) in Luganda and Runyoro

BMC Psychiatry volume 25, Article number: 502 (2025) Cite this article

Abstract

Background

Locally validated measures are required for robust clinical and epidemiological assessments of anxiety symptoms and disorders. Few studies on the African continent have examined the psychometric properties of the Generalized Anxiety Disorder-7 scale (GAD-7). We aimed to investigate certain psychometric properties of GAD-7 translated into Runyoro and Luganda, two Ugandan languages, and to measure the prevalence of anxiety symptoms in a Ugandan population sample.

Methods

Data were collected in 2021–22 through the African Medical and Behavioural Sciences Organization (AMBSO) Population Health Surveillance (APHS), a population cohort study in Uganda. GAD-7, Patient Health Questionnaire-9 (PHQ-9), and questions on socio-demographic factors were administered during individual face-to-face interviews. Construct validity, internal consistency, and concurrent validity of the Runyoro- and Luganda-translated GAD-7 were examined using confirmatory factor analysis (CFA), two internal consistency coefficients (Revelle’s omega total and Cronbach’s alpha), and Pearson’s correlation coefficient, respectively. The prevalence of anxiety symptoms and likely anxiety disorders was also estimated.

Results

A total of 4107 individuals aged 13–80 participated, with 2206 and 1901 speaking Runyoro and Luganda, respectively. The Runyoro-translated GAD-7 retained its one-factor structure (RMSEA = 0.097, CFI = 0.977, TLI = 0.966), had good internal consistency (omega total = 0.85), and correlated moderately with the PHQ-9 (r = 0.67, p < 0.01). The Luganda-translated GAD-7 also fit a one-factor structure (RMSEA = 0.097, CFI = 0.989, TLI = 0.983), exhibited excellent internal consistency (omega total = 0.90), and strong concurrent validity with PHQ-9 (r = 0.71, p < 0.01). Overall, participants reported low levels of anxiety symptoms. Using GAD-7 ≥ 10 binary cut-offs, the total prevalence of likely anxiety disorders was 1.5% (0.5% among males, 2.2% among females).

Conclusions

GAD-7 was translated into Runyoro and Luganda, two Ugandan languages, and both translations showed good psychometric properties. The prevalence of likely anxiety disorders was low in this diverse large Ugandan population sample, the largest population study on anxiety in Uganda to date. The low prevalence could partly be due to individuals with more severe anxiety symptoms not participating or GAD-7 not including local idioms of anxiety. Further clinical validation is required.

Peer Review reports

Background

Mental disorders constitute a substantial public health burden globally. Anxiety disorders are estimated to be the most prevalent mental disorders, but estimates differ widely between cultures and countries [1]. Examples of anxiety disorders include generalized anxiety disorder (GAD), different phobias and panic disorder, previously also post-traumatic stress disorder (PTSD) and obsessive–compulsive disorder (OCD). Effective therapeutic and pharmacological treatments for anxiety disorders exist, but access is insufficient in many countries due to inadequate funding for prevention, detection, or treatment of mental disorders in general [2, 3].

Validated measures are needed to accurately assess the burden of disease attributable to anxiety disorders and inform allocation of resources aimed at addressing this public health issue. The Generalized Anxiety Disorder-7 scale (GAD-7) and Patient Health Questionnaire-9 (PHQ-9) were jointly developed in the early 2000s to screen US-American primary care patients for GAD [4] and major depressive disorder [5] because of the high prevalence and overlap of these disorders. GAD-7 has since then been widely adopted for research purposes [6] and is freely available in multiple languages [7]. All cultures have local idioms of distress, i.e. common expressions to describe different types of suffering, and local idioms are important to consider when translating constructs such as anxiety [8].

Any screening tool to be used outside of its original setting requires both psychometric and clinical validation to ensure local utility and reliable estimates [9, 10]. Psychometric validation evaluates a scale’s structure and quantitative properties (e.g., construct validity, internal consistency, and concurrent validity) while clinical validation compares a screening tool with a gold standard, which for mental disorders usually consists of a structured interview or assessment by trained professionals. Clinical validation occurs less frequently, in part because it requires trained clinical providers to administer the diagnostic assessment. Although GAD-7 has been used previously in Uganda [11, 12], is has not been validated psychometrically or clinically. Previous assessments from other low and middle-income countries have found GAD-7 to be psychometrically robust across diverse settings, such as in Malaysia, Peru, and Kenya [13,14,15].

As evidenced in a recent meta-analysis of the PHQ-9 [16], epidemiological studies using screening tools to measure the burden of mental disorders frequently overestimate their true prevalence. However, screening tools are easier to administer, faster and less expensive than structured interviews and, when carefully interpreted, can provide valuable information for mapping and comparing mental illness globally.

In a 2020 systematic review, GAD-7 and Kessler-10 were found to be the most validated anxiety screening tools in low and middle-income countries, with wide ranges of sensitivities and specificities [17]. In a 2016 global meta-analysis, GAD-7 had an 83% pooled sensitivity and 84% pooled specificity for identifying GAD using the cutoff ≥ 8 points. Accordingly, GAD-7 can yield high numbers of false positive and false negative responses in settings with low prevalences (like in population surveys), and the actual prevalence of anxiety disorders is usually substantially lower than screening results [18].

A 2021 scoping review of GAD in Africa reported broad heterogeneity in point prevalence estimates (0.7–9.6%, with one outlier at 36.5%) in general population samples and cautioned against the use of screening tools (such as GAD-7) in assessing prevalence [19]. A 2021 systematic review of studies among Ugandan adults found a higher pooled prevalence of anxiety disorders (20%) than regional estimates. However, it is worth noting that these estimates included GAD as well as PTSD, a condition disproportionately burdening Ugandans residing in historically war-affected regions [20].

In this study, we aimed to investigate certain psychometric properties of the Runyoro- and Luganda-translated GAD-7, and to measure the prevalence of anxiety symptoms in a Ugandan population sample using the psychometrically validated measures.

Methods

Setting

This study used data from the Africa Medical and Behavioural Sciences Organization (AMBSO) Population Health Surveillance (APHS), an ongoing population-based cohort study in Uganda. Methods of the APHS have been described in detail previously [21]. APHS is conducted annually across six communities (one urban, semi-urban and rural per district) in Hoima and Wakiso, two districts in mid-Western and central Uganda, respectively. Varying community types were selected to represent different characteristics of the Ugandan society, including population density, availability of public infrastructure and income-generating activities. The main language in Hoima is Runyoro. The predominant language in Wakiso is Luganda, which is also the most widely spoken language in the country.

Data collection

Data were collected during the third round of APHS from May 2021 to July 2022. Each round of data collection consists of census activities to enumerate the population and collect household sociodemographic information, followed by individual face-to-face survey administration and collection of biological samples at a community hub. All censused individuals aged 13 years and older were invited to participate in the survey. Surveys lasted approximately one hour, and participants were compensated for their time and any transport costs. Prior to participation, written informed consent (or assent for minors) was obtained. APHS was approved by the local Clarke International University-Research Ethics Committee (CIUREC/0059) and registered with the Uganda National Council for Science and Technology (UNCST/SS4468).

Survey translation

The survey was developed in English, then translated into Runyoro and Luganda by trained social workers fluent in each language and designated by the Common European Framework (CEFR) as C1 or C2 proficient English speakers. Translations were checked by senior researchers to be accurate and locally meaningful and then pilot tested with staff who were not involved in the translation process. Table 1 presents the Runyoro and Luganda versions of GAD-7.

Table 1 GAD-7 in the original English version and translated into Runyoro and Luganda
Full size table

Measurements

Anxiety symptoms in the last two weeks were measured using the 7-item GAD-7 with 4-point Likert-type responses: 0 = not at all, 1 = several days, 2 = more than half of the days, 3 = nearly every day (Table 1). The GAD-7 scale follows the Diagnostic and Statistical Manual of Mental Disorders (DSM)-IV-TR criteria for GAD. A summed score was created from the seven scale items, with higher scores indicating more severe anxiety symptoms (range 0–21). Additionally, continuous GAD-7 scores were categorized using the original 4-category cut-offs (0–4 = no or minimal anxiety, 5–9 = mild anxiety, 10–14 = moderate anxiety, 15–21 = severe anxiety), and dichotomized (0–9 = anxiety disorder unlikely, 10–21 = likely anxiety disorder) for comparability with other published literature [4].

Sociodemographic factors included age, sex, highest education level, household socioeconomic status, and community type. Age was self-reported in completed years. Sex (male or female) was determined by interviewer observation. Highest education level was categorized as none, primary (P1-P7), secondary (S1-S6), and tertiary education. Household socioeconomic status was assessed with an asset index using standard methodology from the Demographic and Health Surveillance Program [22], with quartiles calculated at the household (not individual) level. Urbanity status of communities (i.e., urban, semi-urban, or rural) was defined at cohort creation according to inter alia density of households, the existence of shops, places for social gatherings, employment opportunities, and means of transportation.

Depressive symptoms were measured using the 9-item Patient Health Questionnaire (PHQ-9) [5], which was previously translated and assessed psychometrically by the study team in the same cohort [23]. PHQ-9 measured depressive symptoms in the past two weeks using 4-point Linkert-type responses (0 = not at all, 1 = several days, 2 = more than half the days, 3 = nearly every day). A summed score was created from the nine item responses, with higher scores indicating greater depressive symptomology (range: 0 – 27).

Statistical analysis

Of the individuals who participated in the third round of APHS (N = 4108), the present analysis was restricted to those with complete GAD-7 data (N = 4107). Descriptive statistics were estimated to characterize the study population overall as well as by language (i.e., Runyoro or Luganda). Means (with standard deviations [SD]) and medians (with interquartile range [IQR]) were used for continuous variables and frequencies and percentages were used for categorical variables. Psychometric properties assessed in each language included the translated scale’s construct validity, internal consistency, and concurrent validity.

Construct validity was evaluated using confirmatory factor analysis (CFA) to determine whether the structure of the scale in the current sample corresponded to its designed one-factor structure, whereby all seven scale items share one common underlying characteristic. Factor loadings were calculated using the cfa function with a diagonally weighted least squares estimator in the lavaan package in R [24] and estimated for each scale item the proportion of the variance that was explained by the underlying factor. The more variance accounted for, the stronger the agreement with the one-factor structure [25]. Model fit was evaluated using three model fit indices: the Root Mean Square Error of Approximation (RMSEA), Comparative Fit Index (CFI), and Tucker-Lewis Index (TLI). RMSEA values below around 0.1 indicate acceptable model fit [26], CFI and TLI values above around 0.9 indicate good model fit.

Internal consistency is a type of reliability that measures the homogeneity of the items that comprise a scale. More specifically, reliability is high when the proportion of variance attributable to random error is small compared to a scale score’s total variance. Internal consistency is not a feature of the scale itself but varies based on study population and setting. Longer scales and more heterogeneous samples have a larger total variance and, subsequently, have higher internal consistency [27, 28]. Two reliability coefficients (i.e., Revelle’s omega total and Cronbach’s alpha) were used to assess the internal consistency of the translated GAD-7 scales using the omega function in the psych package in R [29]. Revelle’s omega total is a more robust measure of reliability due to its fewer and more realistic underlying assumptions [28], while Cronbach’s alpha coefficient was estimated for comparison with existing literature.

Concurrent validity assesses how well a measure correlates with similar or related measures at the same point of measurement [30]. A Pearson’s correlation coefficient between anxiety symptoms (GAD-7 score) and depressive symptoms (PHQ-9 score) was estimated using the pwcorr statement in STATA. Correlations were classified as weak (r < 0.3), moderate (r = 0.4–0.6), or strong (r > 0.7) [31].

The prevalences of anxiety symptoms and likely anxiety disorders were estimated, with the latter stratified by sex for comparability with other published literature. Due to small sample sizes, no further subgroup analyses by sex were performed.

Analyses were performed in STATA SE Version 16.1 and R Studio Version 3.6.1.

Results

Overall, participants had a median age of 29 years (IQR 21–39 years), 2444 (59.5%) were female, 2056 (50.1%) had at least some secondary education, and community types were evenly distributed within districts as designed (Table 2). Age, sex, and education level distributions were approximately comparable between Runyoro-speaking participants in Hoima and Luganda-speaking participants in Wakiso. Compared to Luganda-speaking participants in Wakiso, a greater proportion of Runyoro-speaking participants in Hoima had higher socioeconomic status (70.1% in the upper two quartiles compared to 42.6%, respectively) and fewer lived in a semi-urban community (24.9% compared to 31.1%, respectively).

Table 2 Descriptive statistics of the study sample overall and by language, N (%)
Full size table

The frequency of endorsing specific scale items were similar between Runyoro- and Luganda-speaking participants (Table 3); the most (i.e., item 1) and least (i.e., items 5 and 7) endorsed items were the same in both samples.

Table 3 GAD-7 item descriptives and factor loadings in Runyoro and Luganda
Full size table

Assessing construct validity, the factor loadings in each scale were moderate to strong, ranging from 0.58 to 0.92 in Runyoro and from 0.68 to 0.96 in Luganda (Table 3). All factor loadings were higher in the Luganda-speaking sample compared to the Runyoro-speaking one. Model fit indices supported a one-factor structure of GAD-7 in both languages (Table 4), with an acceptable RMSEA value of 0.097 for both the Runyoro- and Luganda-translated GAD-7, and all CFI and TLI values above 0.96, indicating a good fit. Results of invariance testing by sex are provided as a Supplementary file.

Table 4 Psychometric properties of GAD-7 translated into Runyoro and Luganda
Full size table

The internal consistency coefficients indicated good (Revelle’s omega total = 0.85) and excellent (Revelle’s omega total = 0.90) internal consistency for the Runyoro- and Luganda-speaking sample, respectively (Table 4).

The correlation between GAD-7 and PHQ-9 was moderate in the Runyoro-speaking sample (r = 0.67, p < 0.01) and strong in the Luganda-speaking one (r = 0.71, p < 0.01), indicating that the translated anxiety scales had good concurrent validity with depression symptomology (Table 4).

GAD-7 median summed scores were low among both Runyoro- and Luganda-speaking participants (1 and 0, respectively; Table 5); 34.7% (765/2206) of Runyoro-speaking participants in Hoima and 50.1% (953/1901) of Luganda-speaking participants in Wakiso had a GAD-7 score of 0. When considering GAD-7 categorically, 91% of the Runyoro- and Luganda-speaking samples reported no or minimal anxiety symptoms. Using the dichotomous cut-off, 1.1% of the Runyoro-speaking sample screened positive for likely having an anxiety disorder (Table 2), with a higher prevalence among females compared to males (1.8% [23/1276] versus 0.2% [2/930], respectively). Similarly, 1.9% of the Luganda-speaking sample screened positive for likely having an anxiety disorder, with a higher prevalence among females compared to males (2.6% [30/1168] versus 1.0% [7/733], respectively).

Table 5 Descriptives statistics for GAD-7 overall and by language, N (%)
Full size table

Discussion

Psychometrically, the Runyoro-translated GAD-7 had a one-factor structure, good internal consistency, and correlated moderately with depression symptomology as measured by the PHQ-9. The Luganda-translated GAD-7 also had a one-factor structure, excellent internal consistency, and correlated strongly with the PHQ-9. In both population-based samples, the prevalence of anxiety symptoms was low.

Both the Runyoro- and the Luganda-translated GAD-7 scales performed well in terms of the psychometric components that were assessed. Our findings are in line with previous research reporting good psychometric properties of the GAD-7 in different African settings. Studies among Ghanaian high school students [32], people living with HIV in Kenya [15], and primary care patients in Zimbabwe [33] confirmed its unidimensional factorial structure and good internal consistency. The studies in Ghana and Kenya also reported that the GAD-7 correlated with the PHQ-9, in line with our findings. In contrast, an evaluation among Sesotho-speaking South African tuberculosis patients found that a modified two-factor structure, comprising somatic and cognitive-emotional symptoms, fit their data better than a unidimensional model [34]. Invariance testing indicated that a one-factor model had suboptimal fit for Runyoro males, however these also had a very low estimated prevalence of anxiety and therefore we judge a one-factor structure to provide the best fit in both languages based on the more robust non-stratified analyses.

Comparing with population studies of anxiety disorders in other African countries, prevalence estimates in our sample were in the lower range of what has been previously reported [19]. A 2022 systematic review reported a sizeable increase in pooled anxiety prevalence in Africa during the COVID-19 pandemic to 47%, with estimates based on the GAD-7 ranging from 24 to 76% [35]. This review pooled studies using different screening tools and diagnostic instruments, complicating the interpretation of findings. However, our findings are comparable to studies from other East African nations. A 2013 community study in Western Kenya using the Clinical Interview Schedule-Revised found a 1% prevalence of GAD and 4% of any anxiety disorder [36]. In Rwanda, a nationally representative population survey in 2018 using the semi-diagnostic Mini International Neuropsychiatric Interview (MINI) found minimal levels of GAD (< 1%), but instead higher levels of panic disorder (8%), OCD (4%), and PTSD (4%) than estimates for neighbouring countries [37]. No general population studies from Tanzania reporting prevalences of anxiety disorders could be identified.

In the preliminary subgroup analysis by sex, a greater proportion of females had likely anxiety disorders compared to males, which supports worldwide findings regarding anxiety. However, the overall low prevalence contrasts with the estimated 20% prevalence of anxiety disorders in Uganda in the previously mentioned 2021 review [20]. Community studies not included in that review found adult GAD prevalences of 13% using the MINI in rural South-Western parts of the country in 2022 [38], and 11% in villages in Central Uganda using GAD-7 ≥ 10 as cut-off in 2020 [12]. Of note is that the latter study used a different Luganda version of GAD-7, locally adapted by researchers from Makerere University using a more sophisticated translation process. Low sensitivities and specificities of GAD-7 to detect anxiety disorders have been reported even when it was found to be psychometrically robust [39]. It is also important to consider that the prevalence of anxiety disorders in our sample may be even lower than the estimated 1.5%, considering the frequent overestimation of mental disorder prevalence by screening tools and assuming sensitivities and specificities of GAD-7 in the range of published results.

“Anxiety” can be understood as a process of creating an unpleasant feeling of tension, likely existing among all humans. However, the symptoms through which this tension is expressed and categorized into “anxiety disorders” were shaped by locally prevailing norms in European and US-American societies. Given the wide variation in estimates of anxiety symptoms/disorders globally and building on extensive transcultural research [40,41,42,43], one may question the appropriateness of attempting to measure anxiety uniformly worldwide. On the one hand, through the global spread of European and US-American psychiatric traditions and their conceptualizations of psychopathology, “anxiety” (like “depression”) became a de facto standard term and entered cultural spheres far from the term's origins [44]. On the other hand, a long tradition of anthropical research has been working side by side with the medical profession to try to understand the local aspects of the international terminology [45]. Our findings imply that anxiety disorders as conceptualized in Euro-American psychiatry and screened through GAD-7 are uncommon in this sample (i.e., GAD-7 is a robust measure but the real prevalence is low), or that individuals with more severe anxiety symptoms were systematically missed (i.e., GAD-7 is a robust measure and the real prevalence is higher than we found), or that local idioms are required to accurately measure the occurrence of anxiety (i.e., the real prevalence is higher than found and the current formulations of GAD-7 are inadequate). Prior qualitative research by the authors in the same population showed that community members experienced symptoms of mental distress, such as social withdrawal and “thinking too much”, and largely attributed these to contextual factors (such as economic insecurity). Discussion of mental distress on a social level was culturally taboo [46]. The discrepancy between this qualitative work and estimates of anxiety in the present study underscores a potential mismatch between conceptualizations of mental disorders in ICD and DSM and how “anxiety” may manifest and be understood in this setting. To strengthen future research and clinical practice, we recommend validating the Runyoro and Luganda GAD-7 against a diagnostic gold standard or, better yet, to develop more nuanced local measures of “anxiety”.

Substantial efforts have been made to put a stronger emphasis on culturally distinct conceptualizations when measuring human illness and wellness including the Shona Symptom Questionnaire [47], a Mozambican combined screening tool for depression and anxiety [48], and a South Sudanese assessment to capture local idioms of distress and mental illness [49]. In Uganda, a visual depression screening tool has been developed for low literacy settings [50]. These developments are promising examples of contextually adapted measures.

Strengths and limitations

To our knowledge, this study is the first to examine certain psychometric properties of the GAD-7 translated into any Ugandan languages. Other strengths include the general population sample and large sample size, making this one of the largest investigations of anxiety symptoms in East Africa to date. Our samples cover a wide age range, large populations of both males and females, and diverse settings including semi-urban communities, which have received less in focus in mental health research than binary rural–urban categorizations which do not reflect actual urbanization processes.

Potential limitations include a potential selection bias towards individuals with minimal or mild anxiety symptoms. Those with moderate or severe symptoms may have avoided participation. Reasons for this could include logistical barriers, that participation would create additional anxiety, or that affected individuals avoid other community members due to stigma. The Runyoro and Luganda versions of GAD-7 were translated using a simplified process and considered to correspond in meaning by senior bilingual researchers, whose understanding of the items may not be consistent with that of the general population and the translations thus may lack salient local idioms of anxiety. Given only minor differences between our GAD-7 Luganda version and the one used by Makerere researchers, it is unlikely that different translations alone can explain the differing results. Modifications of the one-factor model for the Runyoro version might be considered in future work. We could not evaluate additional aspects of validity and reliability, such as convergent validity and test–retest reliability, because these data were not collected. Furthermore, our statistical approach did not address potential clustering within households as this was deemed of minor relevance. Using a screening tool which so far lacks clinically validated sensitivity and specificity in our settings, we were unable to calculate how different cut-off scores would impact the estimated population prevalence of likely anxiety disorders.

Conclusions

GAD-7 was translated into Runyoro and Luganda, two Ugandan languages, and both translations showed good psychometric properties. The prevalence of likely anxiety disorders was low in this diverse large Ugandan population sample, the largest population study on anxiety in Uganda to date. The low prevalence could partly be due to individuals with more severe anxiety symptoms not participating or GAD-7 not including local idioms of anxiety in the questionnaire. Further clinical validation is required.

Data availability

The data that support the findings of this study are available from Africa Medical and Behavioural Sciences Organization but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of Africa Medical and Behavioural Sciences Organization.

Abbreviations

AMBSO:

Africa Medical and Behavioural Sciences Organization

APHS:

AMBSO Population Health Surveillance

CEFR:

Common European Framework of Reference for Languages

CFA:

Confirmatory factor analysis

CFI:

Comparative Fit Index

DSM:

Diagnostic and Statistical Manual of Mental Disorders

GAD:

Generalized anxiety disorder

GAD-7:

Generalized Anxiety Disorder-7 scale

IQR:

Interquartile range

OCD:

Obsessive-compulsive disorder

PHQ-9:

Patient Health Questionnaire-9

PTSD:

Post-traumatic stress disorder

RMSEA:

Root Mean Square Error of Approximation

SD:

Standard deviations

TLI:

Tucker-Lewis Index

References

  1. GBD 2019 Mental Disorders Collaborators. Global, regional, and national burden of 12 mental disorders in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Psychiatry. 2022;9(2):137–50.

    Article  PubMed Central  Google Scholar 

  2. Patel V, Saxena S, Lund C, Thornicroft G, Baingana F, Bolton P, et al. The Lancet Commission on global mental health and sustainable development. Lancet. 2018;392(10157):1553–98.

    Article  PubMed  Google Scholar 

  3. Mostert CM, Nesic O, Udeh-Momoh C, Khan M, Thesen T, Bosire E, et al. Who should pay the bill for the mental health crisis in Africa? Public Health Pract. 2024;1(7):100458.

    Article  Google Scholar 

  4. Spitzer RL, Kroenke K, Williams JBW, Löwe B. A Brief Measure for Assessing Generalized Anxiety Disorder: The GAD-7. Arch Intern Med. 2006;166(10):1092–7.

    Article  PubMed  Google Scholar 

  5. Kroenke K, Spitzer RL, Williams JBW. The PHQ-9. J Gen Intern Med. 2001;16(9):606–13.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Kroenke K, Spitzer RL, Williams JBW, Löwe B. The Patient Health Questionnaire Somatic, Anxiety, and Depressive Symptom Scales: a systematic review. Gen Hosp Psychiatry. 2010;32(4):345–59.

    Article  PubMed  Google Scholar 

  7. Screener Overview. Patient Health Questionnaire (PHQ) Screeners. https://www.phqscreeners.com/select-screener/. Accessed 10 Oct 2024.

  8. Hinton DE, Lewis-Fernández R. “Idioms of distress” (culturally salient indicators of distress) and anxiety disorders. In: Schneier F, Simpson HB, Lewis-Fernández R, Neria Y, editors. Anxiety Disorders: Theory, Research and Clinical Perspectives. Cambridge: Cambridge University Press; 2010. p. 127–38.

    Chapter  Google Scholar 

  9. Flaherty JA, Gaviria FM, Pathak D, Mitchell T, Wintrob R, Richman JA, et al. Developing instruments for cross-cultural psychiatric research. J Nerv Ment Dis. 1988;176(5):257–63.

    Article  CAS  PubMed  Google Scholar 

  10. Prince MJ. Cross-Cultural Research Methods and Practice. In: Patel V, Minas H, Cohen A, Prince M, editors. Global Mental Health: Principles and Practice. Oxford University Press; 2013. p. 63–81.

  11. Mugisha J, Byansi PK, Kinyanda E, Bbosa RS, Van Damme T, Vancampfort D. Moderate to severe generalized anxiety disorder symptoms are associated with physical inactivity in people with HIV/AIDS: a study from Uganda. Int J STD AIDS. 2021;32(2):170–5.

    Article  PubMed  Google Scholar 

  12. Akena D, Kiguba R, Muwhezi WW, Kwesiga B, Kigozi G, Lukwata H, et al. The prevalence and factors associated with mental disorders in a community setting in central Uganda. PLoS ONE. 2023;18(5):e0285091.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Sidik SM, Arroll B, Goodyear-Smith F. Validation of the GAD-7 (Malay version) among women attending a primary care clinic in Malaysia. J Prim Health Care. 2012;4(1):5–11.

    Article  PubMed  Google Scholar 

  14. Villarreal-Zegarra D, Paredes-Angeles R, Mayo-Puchoc N, Arenas-Minaya E, Huarcaya-Victoria J, Copez-Lonzoy A. Psychometric properties of the GAD-7 (General Anxiety Disorder-7): a cross-sectional study of the Peruvian general population. BMC Psychol. 2024;12(1):183.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Nyongesa MK, Mwangi P, Koot HM, Cuijpers P, Newton CRJC, Abubakar A. The reliability, validity and factorial structure of the Swahili version of the 7-item generalized anxiety disorder scale (GAD-7) among adults living with HIV from Kilifi, Kenya. Ann Gen Psychiatry. 2020;28(19):62.

    Article  Google Scholar 

  16. Levis B, Benedetti A, Ioannidis JPA, Sun Y, Negeri Z, He C, et al. Patient Health Questionnaire-9 scores do not accurately estimate depression prevalence: individual participant data meta-analysis. J Clin Epidemiol. 2020;1(122):115-128.e1.

    Article  Google Scholar 

  17. Mughal AY, Devadas J, Ardman E, Levis B, Go VF, Gaynes BN. A systematic review of validated screening tools for anxiety disorders and PTSD in low to middle income countries. BMC Psychiatry. 2020;20(1):338.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Plummer F, Manea L, Trepel D, McMillan D. Screening for anxiety disorders with the GAD-7 and GAD-2: a systematic review and diagnostic metaanalysis. Gen Hosp Psychiatry. 2016;1(39):24–31.

    Article  Google Scholar 

  19. Greene MC, Yangchen T, Lehner T, Sullivan PF, Pato CN, McIntosh A, et al. The epidemiology of psychiatric disorders in Africa: a scoping review. Lancet Psychiatry. 2021;8(8):717–31.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Opio JN, Munn Z, Aromataris E. Prevalence of Mental Disorders in Uganda: a Systematic Review and Meta-Analysis. Psychiatr Q. 2022;93(1):199–226.

    Article  PubMed  Google Scholar 

  21. Mugamba S, Ziegel L, Bulamba RM, Kyasanku E, Johansson Århem K, Sjöland CF, et al. Cohort Profile: The Africa Medical and Behavioral Sciences Organization (AMBSO) Population Health Surveillance (APHS) in rural, semi-urban and urban Uganda. Int J Epidemiol. 2023;52(2):e116–24.

    Article  PubMed  Google Scholar 

  22. Wealth-Index-Construction. The DHS Program. https://dhsprogram.com/topics/wealth-index/Wealth-Index-Construction.cfm. Accessed 10 Oct 2024.

  23. Miller AP, Espinosa da Silva C, Ziegel L, Mugamba S, Kyasanku E, Bulamba RM, et al. Construct validity and internal consistency of the Patient Health Questionnaire-9 (PHQ-9) depression screening measure translated into two Ugandan languages. Psychiatry Res Commun. 2021;1(2):100002.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Rosseel Y. lavaan: An R Package for Structural Equation Modeling. J Stat Softw. 2012;24(48):1–36.

    Google Scholar 

  25. Floyd FJ, Widaman KF. Factor analysis in the development and refinement of clinical assessment instruments. Psychol Assess. 1995;7(3):286–99.

    Article  Google Scholar 

  26. Kenny DA, Kaniskan B, McCoach DB. The Performance of RMSEA in Models With Small Degrees of Freedom. Sociol Methods Res. 2015;44(3):486–507.

    Article  Google Scholar 

  27. Streiner DL. Starting at the Beginning: An Introduction to Coefficient Alpha and Internal Consistency. J Pers Assess. 2003;80(1):99–103.

    Article  PubMed  Google Scholar 

  28. McNeish D. Thanks coefficient alpha, we’ll take it from here. Psychol Methods. 2018;23(3):412–33.

    Article  PubMed  Google Scholar 

  29. Revelle W. psych: Procedures for Personality and Psychological Research. 2017. https://CRAN.R-project.org/package=psych. Accessed 10 Oct 2024.

  30. Song J, Howe E, Oltmanns JR, Fisher AJ. Examining the Concurrent and Predictive Validity of Single Items in Ecological Momentary Assessments. Assessment. 2023;30(5):1662–71.

    Article  PubMed  Google Scholar 

  31. Akoglu H. User’s guide to correlation coefficients. Turk J Emerg Med. 2018;18(3):91–3.

    Article  PubMed  PubMed Central  Google Scholar 

  32. Adjorlolo S. Generalised anxiety disorder in adolescents in Ghana: Examination of the psychometric properties of the Generalised Anxiety Disorder-7 scale. Afr J Psychol Assess. 2019;1:7.

    Google Scholar 

  33. Chibanda D, Verhey R, Gibson LJ, Munetsi E, Machando D, Rusakaniko S, et al. Validation of screening tools for depression and anxiety disorders in a primary care population with high HIV prevalence in Zimbabwe. J Affect Disord. 2016;1(198):50–5.

    Article  Google Scholar 

  34. Kigozi G. Construct validity and reliability of the generalised anxiety disorder-7 scale in a sample of tuberculosis patients in the Free State Province, South Africa. South Afr J Infect Dis. 2021;36(1):6.

    Google Scholar 

  35. Bello UM, Kannan P, Chutiyami M, Salihu D, Cheong AMY, Miller T, et al. Prevalence of Anxiety and Depression Among the General Population in Africa During the COVID-19 Pandemic: A Systematic Review and Meta-Analysis. Front Public Health. 2022;10:814981.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Jenkins R, Othieno C, Ongeri L, Sifuna P, Ongecha M, Kingora J, et al. Common mental disorder in Nyanza province, Kenya in 2013 and its associated risk factors –an assessment of change since 2004, using a repeat household survey in a demographic surveillance site. BMC Psychiatry. 2015;15(1):309.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Kayiteshonga Y, Sezibera V, Mugabo L, Iyamuremye JD. Prevalence of mental disorders, associated co-morbidities, health care knowledge and service utilization in Rwanda – towards a blueprint for promoting mental health care services in low- and middle-income countries? BMC Public Health. 2022;22(1):1858.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Kalungi A, Kinyanda E, Akena DH, Gelaye B, Ssembajjwe W, Mpango RS, et al. Prevalence and correlates of common mental disorders among participants of the Uganda Genome Resource: Opportunities for psychiatric genetics research. Mol Psychiatry. 2024;14:1–9.

    Google Scholar 

  39. Sawaya H, Atoui M, Hamadeh A, Zeinoun P, Nahas Z. Adaptation and initial validation of the Patient Health Questionnaire – 9 (PHQ-9) and the Generalized Anxiety Disorder – 7 Questionnaire (GAD-7) in an Arabic speaking Lebanese psychiatric outpatient sample. Psychiatry Res. 2016;30(239):245–52.

    Article  Google Scholar 

  40. Hofmann SG, Hinton DE. Cross-Cultural Aspects of Anxiety Disorders. Curr Psychiatry Rep. 2014;16(6):450.

    Article  PubMed  PubMed Central  Google Scholar 

  41. Kaiser BN, Jo WL. Culture-bound syndromes, idioms of distress, and cultural concepts of distress: New directions for an old concept in psychological anthropology. Transcult Psychiatry. 2019;56(4):589–98.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Patel V, Gwanzura F, Simunyu E, Lloyd K, Mann A. The phenomenology and explanatory models of common mental disorder: a study in primary care in Harare. Zimbabwe Psychol Med. 1995;25(6):1191–9.

    Article  CAS  PubMed  Google Scholar 

  43. Wasil AR, Venturo-Conerly KE, Gillespie S, Osborn TL, Weisz JR. In Their Own Words: Using Open-Ended Assessment to Identify Culturally Relevant Concerns among Kenyan Adolescents. Cult Med Psychiatry. 2022;46(2):297–321.

    Article  PubMed  Google Scholar 

  44. Watters E. Crazy like us: The globalization of the American psyche. New York, NY, US: Free Press; vii, 306 p. Crazy like us: The globalization of the American psyche; 2010.

    Google Scholar 

  45. Kaiser BN, Haroz EE, Kohrt BA, Bolton PA, Bass JK, Hinton DE. “Thinking too much”: A systematic review of a common idiom of distress. Soc Sci Med. 2015;1(147):170–83.

    Article  Google Scholar 

  46. Miller AP, Ziegel L, Mugamba S, Kyasanku E, Wagman JA, Nkwanzi-Lubega V, et al. Not Enough Money and Too Many Thoughts: Exploring Perceptions of Mental Health in Two Ugandan Districts Through the Mental Health Literacy Framework. Qual Health Res. 2021;31(5):967–82.

    Article  PubMed  PubMed Central  Google Scholar 

  47. Patel V, Simunyu E, Gwanzura F, Lewis G, Mann A. The Shona Symptom Questionnaire: the development of an indigenous measure of common mental disorders in Harare. Acta Psychiatr Scand. 1997;95(6):469–75.

    Article  CAS  PubMed  Google Scholar 

  48. Belus JM, Muanido A, Cumbe VFJ, Manaca MN, Wagenaar BH. Psychometric Validation of a Combined Assessment for Anxiety and Depression in Primary Care in Mozambique (CAD-MZ). Assessment. 2022;29(8):1890–900.

    Article  PubMed  Google Scholar 

  49. Ng LC, Solomon JS, Ameresekere M, Bass J, Henderson DC, Chakravarty S. Development of the South Sudan Mental Health Assessment Scale. Transcult Psychiatry. 2022;59(3):274–91.

    Article  PubMed  Google Scholar 

  50. Akena D, Joska J, Musisi S, Stein DJ. Sensitivity and Specificity of a Visual Depression Screening Instrument Among HIV-Positive Individuals in Uganda, an Area with Low Literacy. AIDS Behav. 2012;16(8):2399–406.

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

We thank survey participants who generously gave their time to participate, community leaders and volunteers for hosting the APHS, Hoima and Wakiso District Health Departments, ABMSO Community Advisory Board, and the AMBSO staff and leadership for enabling this research. We further thank Zangin Zeebari and Maxim Kan for data analysis support.

Funding

Open access funding provided by Karolinska Institute. Core funding for APHS is provided by Uro Care Ltd. and its affiliated research unit, AMBSO. This study was supported by grants from Fredrik and Ingrid Thuring Foundation (grant number Thuring-2020–00635), Erik and Edith Fernström Foundation (grant number FS-2022:0004), and the Swedish Research Council (grant number VR-2022–04799). The funders had no role in study design, data collection, data analysis, data interpretation, or reporting of findings.

Author information

Authors and Affiliations

  1. Department of Global Public Health, Karolinska Institutet, Solna, Stockholm, 17177, Sweden

    Leo Ziegel, Emmanuel Kyasanku, Anna Mia Ekström & Anna-Clara Hollander

  2. Department of Medicine, University of California San Francisco, San Francisco, CA, USA

    Cristina Espinosa da Silva

  3. Africa Medical and Behavioural Sciences Organization, Nansana, Uganda

    Robert Bulamba, Alex Daama, Grace Kigozi, Godfrey Kigozi, Emmanuel Kyasanku, Stephen Mugamba & Fred Nalugoda

  4. School of Public Health, San Diego State University, San Diego, CA, USA

    Amanda P. Miller

  5. Centre for Psychiatry Research, Department of Clinical Neuroscience, Karolinska Institutet and Stockholm Health Care Services, Region Stockholm, Stockholm, Sweden

    Anders Hammarberg

  6. Department of Infectious Diseases (Venhälsan), South General Hospital, Stockholm, Sweden

    Anna Mia Ekström

Authors
  1. Leo Ziegel
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Cristina Espinosa da Silva
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Robert Bulamba
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Alex Daama
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Grace Kigozi
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Amanda P. Miller
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Godfrey Kigozi
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Emmanuel Kyasanku
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Stephen Mugamba
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Anders Hammarberg
    View author publications

    You can also search for this author inPubMed Google Scholar

  11. Anna Mia Ekström
    View author publications

    You can also search for this author inPubMed Google Scholar

  12. Fred Nalugoda
    View author publications

    You can also search for this author inPubMed Google Scholar

  13. Anna-Clara Hollander
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

LZ, CEDS and APM conceptualized the work and conducted the analysis. RB, AD, GrK, GoK, EK, SM and FN made substantial contributions to data acquisition and interpretation. AH, AME and ACH made substantial contributions to data interpretation. All authors have drafted the work and approved the submitted version.

Corresponding author

Correspondence to Leo Ziegel.

Ethics declarations

Ethics approval and consent to participate

This research was conducted in accordance with the Declaration of Helsinki. APHS was approved by the local Clarke International University-Research Ethics Committee (CIUREC/0059) and registered with the Uganda National Council for Science and Technology (UNCST/SS4468). Informed consent was obtained from all participants prior to participation in the study, for minors assent was obtained and consent from their legal guardians.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ziegel, L., da Silva, C.E., Bulamba, R. et al. Prevalence of anxiety symptoms in a Ugandan population sample and psychometric properties of the Generalized Anxiety Disorder-7 scale (GAD-7) in Luganda and Runyoro. BMC Psychiatry 25, 502 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12888-025-06944-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12888-025-06944-8

Keywords

BMC Psychiatry

ISSN: 1471-244X

Contact us