Prevalence and correlates of burnout among collegiate cycle students in Sri Lanka: a school-based cross-sectional study

Prevalence and correlates of burnout among collegiate cycle students in Sri Lanka: a school-based... Background: Even though the concept of burnout has been widely explored across the globe, the evidence base on burnout among high school students in the South Asian context is scanty. Against the backdrop of ever-increasing educational demands and expectations, the present study was designed to determine the prevalence and correlates of burnout among collegiate cycle students in Sri Lanka. Methods: A school-based cross-sectional study was conducted among 872 grade thirteen students in 15 govern- ment schools in an educational zone, Kegalle district, Sri Lanka selected by a stratified cluster sampling technique. The validated Sinhala version of the 15-item Maslach Burnout Inventory-Student Survey (MBI-SS) was used to assess burnout. The adjusted prevalence of burnout was computed based on the clinically validated cut-off values using the “exhaustion + 1” criterion. Multivariable logistic regression was carried out using backward elimination method to quantify the association between burnout and selected correlates identified at bivariate analysis at p value less than 0.05. Results: The response rate was 91.3% (n = 796). The adjusted prevalence of burnout among grade thirteen students was 28.8% (95% CI = 25.0–32.7%). Multivariable analysis elicited a multitude of statistically significant associations with burnout when controlled for other factors included in the model (p < 0.05). Perceived satisfaction related to the school environment (classroom and library facilities), school curriculum (scope, relevance, and difficulty of the subject content), study enthusiasm (preferred subject stream), study support (support from parents to teachers), and future expectations (personal and parental expectations) emerged as statistically significant negative associations with burn- out, whereas having to encounter disturbances while studying and being subjected to bullying at school emerged as statistically significant positive associations with burnout. Conclusions: The burnout prevalence among grade thirteen students in the selected educational zone, Sri Lanka is high. Most of the significant correlates of burnout are directly related to the academic endeavours. It is recommended to strengthen the counseling services at the school level to rectify the problems related to burnout among collegiate cycle students in Sri Lanka. Keywords: Student burnout, MBI-SS, Prevalence, Correlates, Collegiate cycle, Sri Lanka *Correspondence: nuwick74@yahoo.com Department of Community Medicine, Faculty of Medicine and Allied Sciences, Rajarata University of Sri Lanka, Saliyapura 50008, Sri Lanka Full list of author information is available at the end of the article © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/ publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 2 of 11 burnout among Indian undergraduate and postgraduate Background students substantially varies from 10.2 to 52.0% [19, 20]. The concept of student burnout has been in the limelight While there is a paucity of literature assessing burnout since the introduction of the Maslach Burnout Inventory- among high school students in the South Asian context, Student Survey (MBI-SS) in 2002. In accordance with the thus far there is no published literature related to student original definition of burnout, student burnout is defined burnout in Sri Lanka. as, “a three-dimensional syndrome that is characterised In Sri Lanka, the collegiate cycle in the education sys- by feelings of exhaustion due to the demands of study- tem (consists of grade 12 and grade 13) leads to the Gen- ing, a cynical attitude of withdrawal and detachment, eral Certificate of Examination Advanced Level, which and reduced professional efficacy regarding academic is the national level selection examination for state uni- requirements” [1]. In the global literature on student versity admissions. National statistics indicate that the burnout research, the MBI-SS has been cited as the most examination has become extremely competitive for Sri widely used research instrument to assess burnout in Lankan collegiate cycle students, who are in the age range different student populations [2, 3]. Furthermore, the of 17–19 years [21]. Evidence suggests that approximately validity and the reliability including the three-factor con- 40% of adolescents found it stressful to cope with the aca- ceptualisation of MBI-SS have been confirmed in a num - demic pressures exerted on them by parents and teachers ber of different student populations in different countries [22] and almost one in five adolescents in schools have [1, 3–6]. clinically relevant mental health problems with a sub- Amongst the wealth of research concerning student stantial proportion having symptoms classified as defi - burnout, a vast majority has focused on assessing burn- nite or severe, while educational performance is reported out among university undergraduates [1, 3, 7, 8]. How- as the most impacted area of life [23]. Furthermore, the ever, against the backdrop of ever-increasing educational prevalence of mental health problems such as depres- demands and expectations, the amount of research con- sion and anxiety among the collegiate cycle students in ducted among high school students is scanty. Moreover, Sri Lanka is high with examinations being the most com- within the scanty evidence base, the reported prevalence monly cited cause [24]. Though academic endeavours of burnout reflects a substantial variation depending on are usually considered as the ostensible reason for the the definitions and criteria used, intrinsic factors of the resultant mental health problems, other numerous non- samples studied, and the  cut-off values applied for the academic factors could have been the significant contrib - assessment. The prevalence of burnout among Finnish utive reasons. In the light of high prevalence of mental junior high school students was reported as 10.9% [9], health problems in students, it is of utmost importance to while other studies have reported even higher prevalence explore the concept of student burnout, which is directly estimates among high school students [10, 11]. assessing psychological well-being in relation to aca- Studies conducted to explore the associations of burn- demic endeavours. out among different student populations across the globe Given the research vacuum pertaining to student burn- have revealed a multitude of significant associations. out in Sri Lanka, the present study was designed to assess Regarding high school and college students, character- the prevalence and correlates of burnout among colle- istics such as lower self-efficacy [12, 13], maladaptive giate cycle students in a selected educational zone in Sri perfectionism [14], negative temperament [15], negative Lanka. self-image [16], and depressive symptoms [17] are found to have positive associations with burnout. Educational Methods environment related factors such as positive school class Study design and setting atmosphere [18], support from school [18], and academic This school-based cross-sectional study was conducted achievement [12, 15, 17] are negatively associated, while in a selected educational zone in the Kegalle district, high course/work load is positively associated with burn- Sabaragamuwa Province, Sri Lanka from January 2015 out [12, 15]. to April 2015. Altogether there are 144 schools in the Even though the global literature suggests that there is selected educational zone with a total student population a multitude of student burnout correlates, the evidence in of approximately 51,000. the South Asian context is sparse. This existing research gap in the South Asian context needs to be addressed, Participants as differences in the socio-cultural backgrounds, educa - The study population consisted of all grade thirteen stu - tional systems, and the level of education might limit the dents studying in Sinhala medium government schools generalisability of previous study findings. in the educational zone at the time of the study exclud- The scanty evidence base pertaining to student burnout ing students who were unable to read or write in Sinhala in the South Asian context reveals that the prevalence of Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 3 of 11 language. There were 1690 grade thirteen students in The clinically validated cut-off values for the subscale seven category 1AB and 31 category 1C government scores of the MBI-SS were developed by computing schools studying in four main subject streams, viz., Sci- Receiver Operating Characteristic curves using the clini- ence, Arts, Commerce and Technology. Category 1AB cal diagnosis made by a Consultant Psychiatrist as the schools have classes in all subject streams, whereas 1C reference standard in a sample of grade thirteen students schools have classes in Arts and Commerce streams only. in a similar educational setting. The clinical assessment The required sample size was calculated with 95% of burnout by the Consultant Psychiatrist was based on confidence level, 5% absolute precision for an antici - the clinical criteria for the diagnosis of work related neu- pated prevalence of burnout of 30% (as 32% of colle- rasthenia according to the ICD 10 classification [30–35]. giate cycle students were categorised as having burnout Diagnostic accuracy of the MBI-SS test results based on in the validation study of the Sinhala version of the “exhaustion + 1” criterion, which is an accepted crite- MBI-SS) based on standard sample size calculation for- rion used in burnout research [30, 34–37], was assessed mula [25] using OpenEpi software version 3.01. The comparing with the results of the clinical diagnosis. The final sample size was computed as 790, after adjusting clinically validated cut-off values for EX, CY and rPE for a design effect of 2.2 and an anticipated level of non- subscale scores were 12.5, 7.5 and 10.5, respectively. The response of 10%. Stratified cluster sampling technique sensitivity, specificity, positive and negative predictive was used to select the study sample in which a clus- values of the Sinhala version of the 15-item MBI-SS were ter was defined as a grade thirteen class with a median 91.9, 93.2, 86.4 and 96.1%, respectively. Further, the posi- number of 21 students. Hence, the study was conducted tive and negative likelihood ratios were 13.48 and 0.09, in 38 clusters from 15 schools selected according to the respectively [38]. probability-proportional-to-size. The second component of the SAQ, which was intended to gather information related to the correlates Measures of burnout, was developed following an extensive litera- A pre-tested, validated self-administered questionnaire ture search and with the inputs from the experts in the (SAQ), which consisted of two components, was used in fields of psychiatry, psychology, public health, teaching, the present study. student counseling, and medical education. The ques - The first component of the SAQ included the Sin - tionnaire was pre-tested in a sample of grade thirteen hala version of the 15-item MBI-SS in order to assess students in a similar educational setting. The question - the burnout status. The MBI-SS has been used to assess naire consisted of ten sections focused on information burnout status among high school students in several set- related to personal and family characteristics, residence, tings [3, 26–28] and the Sinhala version of the 15-item school environment, curriculum, the pattern of study, MBI-SS was found to be a valid and a reliable instrument support for studies, study enthusiasm, future expecta- to assess burnout among collegiate cycle students in Sri tions, personal behaviours/personal life factors, and Lanka [29]. Out of the total 15 items of the MBI-SS, five, behaviours of others. four and six items are targeted at identifying exhaustion (EX), cynicism (CY) and reduced professional efficacy Procedure (rPE) subscales, respectively. The frequency in which Ethical clearance to conduct the study was obtained the respondents experience feelings related to each sub- from the Ethics Review Committee of the Faculty of scale was assessed using a seven-point, fully anchored Medicine and Allied Sciences, Rajarata University of Sri response format, ranging from 0 (never) to 6 (every day). Lanka (Reference no: ERC/2014/057). Prior to data col- During the validation and cultural adaptation of the lection, administrative clearance was obtained from the Sinhala version of the MBI-SS, a multi-disciplinary panel Zonal Director of Education and the principals of all the of experts in the fields of psychiatry, psychology, public selected schools. The dates for data collection in different health, teaching, student counseling, and medical edu- schools were selected according to the logistic conveni- cation has assessed each item of the questionnaire on ence of the schools in order to minimise the disturbance its relevance, appropriateness, and acceptability in the to the routine academic and other endeavours. Informed local context for assessing burnout among grade thir- written consent was obtained from all the students in teen students. In the confirmatory factor analysis, the each selected classroom. Data collection was took place 15-item three-factor model emerged as an acceptable fit - inside the classrooms and each participant was given a ting model. In addition, the 15-item MBI-SS showed high copy of the printed SAQ to be filled independently. Con - internal consistency (Cronbach’s α > 0.8) and high test– fidentiality of data collected was adhered to strictly and retest reliability (p < 0.001) [29]. the anonymity of the participants was maintained. Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 4 of 11 Data analysis Table 1 Basic characteristics of  the  sample of  grade Data analysis was done by using the SPSS version 17.0. thirteen students (n = 796) After entering, double independent check of entries Characteristic Number Percentage (%) Cumulative was carried out to identify any incompatible entries. percentage (%) The dataset was examined for univariate and multivari - ate outliers using box plots and Mahalanobis distance, Sex respectively. Female 440 55.3 55.3 Scoring of the MBI-SS was carried out according to Male 356 44.7 100.0 the MBI manual instructions [39]. Based on the clini- Religion cally validated cut-off values for the three subscales, a Buddhist 774 97.2 97.2 participant was categorised as having burnout according Hindu 2 0.3 97.5 to the “exhaustion + 1” criterion, in which a participant Christian 16 2.0 99.5 who is having a high score on EX in combination with Islam 1 0.1 99.6 a high score on either of the CY or rPE subscale was Other 3 0.4 100.0 regarded as having burnout [30, 37]. As cluster sampling Monthly family income technique was used to select the study participants and LKR 10,000–20,000 251 31.5 31.5 the number of participants in the different clusters was LKR 20,001–30,000 118 14.8 46.4 not uniform, weighted analysis was conducted to deter- LKR 30,001–40,000 98 12.3 58.7 mine the prevalence of burnout. Each observation in the LKR 40,001–50,000 143 18.0 76.6 sample was assigned a particular weight, which was cal- > LKR 50,001 186 23.4 100.0 culated as the product of inverse selection probabilities Subject stream at each stage of sampling [40, 41]. Incorporating the val- Science 235 29.5 29.5 ues of diagnostic accuracy of validated Sinhala version Arts 276 34.7 64.2 of the MBI-SS [38], the adjusted prevalence of burnout Commerce 229 28.8 93.0 with 95% confidence interval (CI) was calculated in Technology 56 7.0 100.0 grade thirteen students [42]. Total 796 100.0 For the assessment of correlates of burnout, a two-step LKR Sri Lankan Rupees procedure was followed. First, a bivariate analysis was performed to identify potential correlates of burnout. Then, a multivariable analysis using binary logistic regres - The mean age of the grade thirteen students in the sam - sion was conducted to identify the relevant predictors of ple was 18.4  years (SD = 0.32  years). The majority of the burnout and to control for potential confounding among participants were females (n = 440, 55.3%) and 276 stu- the various predictor variables. Categorical data related dents (34.7%) were studying in the Arts subject stream. to predictor variables were amalgamated rationally as dichotomous variables where necessary for the bivariate analysis and the crude odds ratios (OR) were calculated Descriptive statistics of the Sinhala version of the MBI‑SS as the measures of effect with 95% CI. Correlates that subscale scores showed statistical significance at p value less than 0.05 in Table  2 summarizes the mean total scores and the mean the bivariate analysis were included in the multivariable item scores of the three subscales of the MBI-SS Sinhala analysis using backward stepwise elimination method. version. The model produced adjusted odds ratios (AOR) and 95% CI with the significance level for variables of interest. In order to conduct logistic regression, it is recommended to Prevalence of burnout have at least ten observations per an independent variable The prevalence of burnout based on the clinically vali - [43] and the dataset met this criterion. dated cut-off values for each subscale score and the “exhaustion + 1” criterion was 36.8% (95% CI = 33.5– 40.2%). The weighted analysis conducted to compensate for the complex sampling design resulted in a weighted Results prevalence estimate of 31.3% (95% CI = 28.1–34.6%). Characteristics of the sample According to the sensitivity and the specificity of the Sin - Out of the total of 872 grade thirteen students in the 38 hala version of the 15-item MBI-SS, the adjusted preva- identified clusters, 796 students completed the SAQ; lence of burnout among grade thirteen students in the hence, the response rate was 91.3%. Table  1 summarises study was 28.8% (95% CI = 25.0–32.7%). the basic characteristics of the study sample. Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 5 of 11 Table 2 Descriptive statistics of  the  subscale scores Having to encounter disturbances while studying and of  the  MBI-SS Sinhala version among  grade thirteen being subjected to bullying at school emerged as statisti- students (n = 796) cally significant positive associations with burnout. Subscale Mean total score SD Mean item SD score Discussion The present study, which was designed with the objec - EX 11.98 7.16 2.40 1.43 tive of determining the prevalence and correlates of CY 6.80 5.98 1.70 1.49 burnout among grade thirteen students in Sri Lanka, rPE 10.56 6.46 1.76 1.08 addresses an important research vacuum in relation to burnout research in high school students in the South Correlates of burnout in the bivariate analysis Asian context. The study findings reveal that almost one In the bivariate analysis, 35 factors emerged as signifi - in four grade thirteen students is likely to have burnout cant predictors of burnout. These included a number of and burnout is significantly associated with a multitude factors related to the study environment, curriculum, of academic environment related correlates. and behaviours. Table  3 presents the summary of sta- Having a high response rate and computing adjusted tistically significant independent predictor variables of prevalence by compensating for the sampling complexity burnout emerged in the bivariate analysis. and diagnostic uncertainty of the assessment tool pro- vide valid and precise estimates for burnout among grade thirteen students in this study. The prevalence of burnout among the collegiate cycle Multivariable analysis of correlates of burnout students in this study is higher than the reported values All 35 independent predictors identified at bivariate of prevalence of burnout among high school students analysis were included in the multivariable analysis. conducted in different study settings, such as 10.9% None of these predictors had categories with very few among students in Finnish public junior high schools [9], observations, both the dependent and the independ- 14% among Finnish high school students [10], and 12.6% ent variables were dichotomous in nature, and there among middle school and regular secondary school in were no outliers in the data set. Table 4 summarises the Northern China [44]. According to the World Bank sta- results of the multivariable analysis of the correlates of tistics, the total enrolment in tertiary education in Fin- burnout retained in the final model. Out of the 14 fac - land and Sweden are substantially higher than that of Sri tors retained in the final model, 12 factors made unique Lanka; based on which, it can be argued that the com- statistically significant contributions at a p value less petitiveness of the tertiary education enrolment exami- than 0.05. nations and academic endeavour related stress could be Multivariable analysis elicited several statistically sig- higher in the Sri Lankan context as opposed to the study nificant associations with burnout when controlled for settings in other highlighted countries, contributing to other factors included in the model (p < 0.05). Perceived the observed difference of the magnitude of the prob - satisfaction about the facilities available in the classroom, lem of burnout. However, other possible explanations about the library facilities, and about the scope of the such as the crucial differences in the educational context, subject content covered in the curriculum showed statis- burnout assessment tool, assessment cut-off values, and tically significant negative associations with burnout. assessment criteria have to be taken into consideration in Students who thought the content covered in the sub- critically evaluating the research findings for comparative jects are relevant to the curriculum and students who purposes. Provided that there is no universally accepted easily understood the subject content taught in the cur- diagnosis method of burnout, it is important to appreci- riculum had statistically significant lower likelihood of ate that the prevalence estimates reported in the present having burnout in comparison to their counterparts. study is dependent upon the burnout assessment criteria Satisfactory support from parents to satisfactory sup- used in the study. port from teachers were found to have statistically signif- The multivariable analysis revealed that the students icant negative associations with burnout. who encountered disturbances while studying were more The students who selected the subject stream based likely to have burnout in comparison to their counter- on their own decision were less likely to have burnout parts. It can be assumed that the disturbances at study as opposed to those who have selected the current sub- place causing distractions would make the students una- ject stream for other reasons. Students who felt that both ble to study or concentrate on their studies. The result - their own future expectations and parental expectations ant frustration may provoke a feeling of indifference or are encouraging their studies were less likely to have distant attitude towards work. The ultimate result would burnout as opposed to those who did not feel so. Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 6 of 11 Table 3 Statistically significant independent predictor variables of  burnout in  grade thirteen students in  the  bivariate analysis Characteristic Burnout No burnout Total Odds ratio (95% CI) p n (%) n (%) n (%) value Personal characteristics Sex (n = 796) Female 145 (33.0) 295 (67.0) 440 (100.0) 0.7 (0.5–0.9) Male 148 (41.6) 208 (58.4) 356 (100.0) p = 0.012 Residence related factors Satisfaction of study place (n = 705) Satisfactory 139 (32.0) 295 (68.0) 434 (100.0) 0.6 (0.5–0.9) Not satisfactory 116 (42.8) 155 (57.2) 271 (100.0) p = 0.004 Type of disturbances (n = 360) From inside 70 (49.3) 72 (50.7) 142 (100.0) 1.5 (1.0–2.3) From outside/both inside and outside 85 (39.0) 133 (61.0) 218 (100.0) p = 0.054 Disturbances to studies at residence (n = 796) Yes 155 (43.1) 205 (56.9) 360 (100.0) 1.6 (1.2–2.2) No 138 (31.7) 298 (68.3) 436 (100.0) p = 0.001 Duration of travel to school (n = 796) Less than 30 min 131 (33.0) 266 (67.0) 397 (100.0) 0.7 (0.5–0.9) More than 30 min 162 (40.6) 237 (59.4) 399 (100.0) p = 0.026 School environment related factors Facilities of the classroom (n = 796) Satisfactory 81 (17.0) 395 (83.0) 476 (100.0) 0.1 (0.1–0.2) Not satisfactory 212 (66.3) 108 (33.7) 320 (100.0) p < 0.001 School library facilities (n = 796) Satisfactory 194 (33.1) 392 (66.9) 586 (100.0) 0.6 (0.4–0.8) Not satisfactory 99 (47.1) 111 (52.9) 210 (100.0) p < 0.001 School health services (n = 796) Satisfactory 111 (30.4) 254 (69.6) 365 (100.0) 0.6 (0.4–0.8) Not satisfactory 182 (42.2) 249 (57.8) 431 (100.0) p = 0.001 School counseling services (n = 796) Satisfactory 98 (30.9) 219 (69.1) 317 (100.0) 0.7 (0.5–0.9) Not satisfactory 195 (40.7) 284 (59.3) 479 (100.0) p = 0.005 School recreational facilities (n = 796) Satisfactory 85 (28.3) 215 (71.7) 300 (100.0) 0.5 (0.4–0.7) Not satisfactory 208 (41.9) 288 (58.1) 496 (100.0) p < 0.001 Curriculum related factors Scope of subject content (n = 766) Satisfactory 115 (22.5) 395 (77.5) 510 (100.0) 0.1 (0.1–0.2) Not satisfactory 175 (68.4) 81 (31.6) 256 (100.0) p < 0.001 Amount of assignments/workload (n = 777) Satisfactory 103 (28.5) 259 (71.5) 362 (100.0) 0.5 (0.4–0.7) Not satisfactory 180 (43.7) 232 (56.3) 412 (100.0) p < 0.001 Relevance of subject area (n = 781) Satisfactory 141 (30.5) 322 (69.5) 463 (100.0) 0.5 (0.4–0.7) Not satisfactory 144 (45.3) 174 (54.7) 318 (100.0) p < 0.001 Difficulty in understanding (n = 796) Easily understood 79 (18.9) 338 (81.1) 417 (100.0) 0.2 (0.1–0.3) Not easily understood 214 (56.5) 165 (43.5) 379 (100.0) p < 0.001 Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 7 of 11 Table 3 (continued) Characteristic Burnout No burnout Total Odds ratio (95% CI) p n (%) n (%) n (%) value Study pattern related factors Duration of study per day (n = 796) Less than 2 h 119 (42.5) 161 (57.5) 280 (100.0) 1.4 (1.1–2.0) More than 2 h 174 (33.7) 342 (66.3) 516 (100.0) p = 0.014 Methods of study (n = 796) Interactive methods with others 271 (36.0) 482 (64.0) 753 (100.0) 0.5 (0.3–0.9) No interaction with others 22 (51.2) 21 (48.8) 43 (100.0) p = 0.045 Answering past papers (n = 796) Yes 177 (31.8) 379 (68.2) 556 (100.0) 0.5 (0.4–0.7) No 116 (48.3) 124 (51.7) 240 (100.0) p < 0.001 Study support related factors Support from parents (n = 792) Satisfactory 208 (30.4) 476 (69.6) 684 (100.0) 0.1 (0.1–0.2) Not satisfactory 84 (77.8) 24 (22.2) 108 (100.0) p < 0.001 Support from colleagues (n = 759) Satisfactory 160 (33.8) 313 (66.2) 473 (100.0) 0.7 (0.5–1.0) Not satisfactory 117 (40.9) 169 (59.1) 286 (100.0) p = 0.050 Support from teachers (n = 780) Satisfactory 231 (33.3) 462 (66.7) 693 (100.0) 0.3 (0.2–0.5) Not satisfactory 54 (62.1) 33 (37.9) 87 (100.0) p < 0.001 Study enthusiasm related factors Subject stream (n = 740) Science 105 (44.7) 130 (55.3) 235 (100.0) 1.8 (1.3–2.5) Arts/Commerce 154 (30.5) 351 (69.5) 505 (100.0) p < 0.001 Preference of subject stream (n = 796) Own preference 230 (32.5) 478 (67.5) 708 (100.0) 0.2 (0.1–0. 3) Not own preference 63 (71.6) 25 (28.4) 88 (100.0) p < 0.001 Opinion of academic performance (n = 796) Satisfactory 112 (28.6) 279 (71.4) 391 (100.0) 0.5 (0.4–0.7) Not satisfactory 121 (44.7) 224 (55.3) 405 (100.0) p < 0.001 Average marks (n = 577) Less than 50 101 (46.3) 117 (53.7) 218 (100.0) 1.6 (1.1–2.2) More than 50 126 (35.1) 233 (64.9) 359 (100.0) p = 0.007 Future expectation related factors Personal expectations (n = 774) Encouraging 249 (34.8) 467 (65.2) 716 (100.0) 0.3 (0.2–0.6) Not encouraging 36 (62.1) 22 (37.9) 58 (100.0) p < 0.001 Expectations of parents (n = 772) Encouraging 241 (33.8) 473 (66.2) 714 (100.0) 0.2 (0.1–0.3) Not encouraging 44 (75.9) 14 (24.1) 58 (100.0) p < 0.001 Expectations of teachers (n = 761) Encouraging 232 (34.6) 439 (65.4) 671 (100.0) 0.4 (0.3–0.7) Not encouraging 49 (54.4) 41 (45.6) 90 (100.0) p < 0.001 Expectations of colleagues (n = 677) Encouraging 162 (33.5) 322 (66.5) 484 (100.0) 0.7 (0.5–0.9) Not encouraging 83 (43.0) 110 (57.0) 193 (100.0) p = 0.020 Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 8 of 11 Table 3 (continued) Characteristic Burnout No burnout Total Odds ratio (95% CI) p n (%) n (%) n (%) value Personal behaviours/personal life factors Love affair (n = 796) Yes 87 (43.3) 114 (56.7) 201 (100.0) 1.4 (1.1–2.0) No 206 (34.6) 389 (65.4) 595 (100.0) p = 0.028 Eec ff t of hobbies (n = 769) Useful to studies 67 (31.8) 144 (68.2) 211 (100.0) 0.7 (0.5–1.0) Not useful to studies 219 (39.2) 339 (60.8) 558 (100.0) p = 0.055 Eec ff t of religious activities (n = 780) Useful to studies 216 (35.1) 400 (64.9) 616 (100.0) 0.7 (0.4–1.1) Not useful to studies 69 (42.1) 95 (57.9) 164 (100.0) p = 0.098 Factors related to behaviours of other people Household substance abuse (n = 796) Yes 152 (46.2) 177.(53.8) 329 (100.0) 2.0 (1.5–2.7) No 141 (30.2) 326 (69.8) 467 (100.0) p < 0.001 Colleagues substance abuse (n = 796) Yes 130 (42.9) 173 (57.1) 303 (100.0) 1.5 (1.1–2.0) No 163 (33.1) 330 (66.9) 493 (100.0) p = 0.005 Peer pressure for substance abuse (n = 796) Yes 21 (53.8) 18 (46.2) 39 (100.0) 2.1 (1.1–3.9) No 272 (35.9) 485 (64.1) 757 (100.0) p = 0.024 Subjected to bullying at school (n = 796) Yes 160 (51.8) 149 (48.2) 309 (100.0) 2.9 (2.1–3.9) No 133 (27.3) 354 (72.7) 487 (100.0) p < 0.001 Total is not equal to 796 due to responses not included in the analysis and/or missing values Table 4 Correlates of burnout in the sample of 796 grade thirteen students in the multivariable analysis Factor B SE Wald df p value Odds ratio 95% CI Disturbance to studies: yes 0.740 0.366 4.083 1 0.043 2.1 1.1–4.3 Facilities of the classroom: satisfactory − 2.151 0.327 43.156 1 < 0.001 0.1 0.1–0.2 Library facilities: satisfactory − 0.691 0.332 4.338 1 0.037 0.5 0.3–0.9 School health services: satisfactory − 0.578 0.328 3.080 1 0.079 0.6 0.3–1.1 Scope of subject content: satisfactory − 1.464 0. 357 16.818 1 < 0.001 0.2 0.1–0.5 Relevance of subject area: satisfactory − 1.900 0.421 20.333 1 < 0.001 0.2 0.1–0.3 Difficulty in understanding: easily understood − 1.380 0.340 16.444 1 < 0.001 0.3 0.1–0.5 Duration of study: less than 2 h per day 0.574 0.325 3.127 1 0.077 1.8 0.9–3.4 Support from parents: satisfactory − 1.800 0.488 13.612 1 < 0.001 0.2 0.1–0.4 Support from teachers: satisfactory − 1.084 0.449 5.832 1 0.016 0.3 0.1–0.8 Preference of subject stream: own preference − 1.995 0.486 16.825 1 < 0.001 0.1 0.1–0.4 Personal expectations: encouraging − 1.251 0.576 4.716 1 0.030 0.3 0.1–0.9 Expectations of parents: encouraging − 1.002 0.346 8.375 1 0.004 0.4 0.2–0.7 Subjected to bullying at school: yes 0.981 0.302 10.557 1 0.001 2.7 1.5–4.8 Italic values indicate significance of p value (p < 0.05) df degree of freedom; SE standard error; 95% CI 95% confidence interval Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 9 of 11 be a cynical attitude of withdrawal and detachment from the negative association elicited between burnout and academic activities, which is an important aspect of stu- the support from parents to teachers. When academic dent burnout. endeavour related stress increases, the students are in School environment related factors, such as having sat- search for social support to cope with the stress and the isfactory classroom and library facilities was associated mental pressure. with lower likelihood of having burnout and these find - Students, who selected the subject stream that they ings are consistent with the existing evidence [18]. It can are following based on their own decision, were less be assumed that the limited facilities or resources avail- likely to have burnout as opposed to those who opted able at school environment are unable to inspire students to follow the current subject stream for other reasons. to work hard, though in a paradox, the limited resources Similarly, Vasalampi, Salmela-Aro and Nurmi [49] have necessitate students to work hard to achieve better shown that self-concordance was associated with stu- results. The lack of psychological stimulation originated dent burnout. According to Ryan and Connell [50], by the lack of resources may aggravate students’ disinter- even though it is believed that personal goals are self- est towards studies. Furthermore, it may contribute to determined, they are not solely originated based on the further mental exhaustion of students owing to the idea intrinsic values and personal interests. It is believed of having to work hard. These reasons may have lead to that individuals are inclined to adopt their goals for the significant association elicited in the study. external reasons such as social pressure or expectations Curriculum related factors, such as students’ satisfac- of others [51]. These arguments have generated the tion of the scope of the subject content covered in the basis for the self-concordance model for goal selection curriculum and relevance of the subject to the curricu- [51, 52]. Furthermore, according to literature, pursuit lum, emerged as statistically significant predictors with of goals for internalised reasons promotes sustained lower likelihood of having burnout. In a similar vein, efforts leading to a better goal progress [53]. studies suggest a positive association of course work- In relation to future expectations on academic load with burnout [12, 15]. Factors such as the wide endeavours, students who felt that both their own scope covered in the subject content and the difficulty in future expectations and parental expectations are understanding the subject content demand students to encouraging their studies were less likely to have burn- make an extra effort in relation to their academic endeav - out as opposed to those who felt these expectations are ours. Students in failing to understand the deep and dif- not encouraging to their studies. As discussed above, ficult underpinning subject content resort to cram facts as per the self-concordance model for goal selection, a and theories, which hinders the opportunity to acquire certain element of external influence is incorporated in higher-order generic competencies. This results in the the selection of academic goals. In the Sri Lankan con- imbalance between the abilities of the students and the text, parental expectations play an important role in expectations demanded by the academic environment this regard. Hence, the subjective perception of encour- leading to the excess amount of stress. This progressive aging parental expectations together with personal vicious phenomenon culminates in burnout, which is expectations may bring about a better goal progress. recognised as a result of failed attempts to cope with a Thus, the students who are having such subjective per - variety of negative stress situations [45]. ceptions may be less likely to develop burnout. In relation to support received for studies from oth- In the present study, the students who were subjected ers, students who perceived that they receive satis- to bullying at school were more likely to have burnout. factory support from their parents and satisfactory This finding is congruent with existing evidence sug - support from their teachers were found to have statisti- gesting that bullying in schools has a negative impact cally significant negative associations with burnout and on students’ mental health [54, 55]. Being bullied is a these findings are consistent with few other study find - particularly intense and traumatic form of stress, which ings [12, 13, 15]. Social support is regarded as one of the usually involves a loss of control, and poses a threat to a important aspects in explaining the burnout phenom- person’s wellbeing. enon [46]. Social support of an individual includes peo- In sum, the present study revealed several signifi - ple who are caring and valuing the individual and the cant correlates of burnout and the findings are congru - people that can be relied upon [12]. Social support has ent with other study findings. Interestingly, amongst been identified as a resource that enables individuals the significant correlates, almost all were either corre - to cope with stress [47]. Pines and Maslach [48] high- lates that are directly related to the academic endeav- lighted the fact that the social support has been iden- ours or correlates that have a direct effect on academic tified as both a preventive mechanism and a remedy endeavours. against burnout. These collective evidence, elucidate Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 10 of 11 Competing interests Limitations The authors declare that they have no competing interests. There are some limitations related to the study that need to be taken into consideration in interpreting the results. Availability of data and materials The datasets used and analysed during the present study are available from Owing to the fact that the present study was conducted the corresponding author on reasonable request. in a selected educational zone in the Kegalle district, Sri Lanka the generalisation of the study findings to other Consent for publication Not applicable. study settings should be done with caution, consider- ing the educational and cultural differences. Due to the Ethics approval and consent to participate cross-sectional nature of the study, a temporal relation- Ethical clearance to conduct the study was obtained from the Ethics Review Committee of the Faculty of Medicine and Allied Sciences, Rajarata University ship between burnout and correlates could not be estab- of Sri Lanka (Reference no: ERC/2014/057). Informed written consent from all lished based on the present study findings. In the present the participants were obtained prior to data collection. (All the participants study, even though multivariable analysis accounted for were above the age of 16 years). confounding, the effect of unknown confounders could Funding not be accounted for. This work was supported by the University Grants Commission-Sri Lanka, under the Postgraduate Research Grant scheme [Grant number: UGC/DRIC/ PG/2015(I)/RUSL/01]. The funding body did not involve in the design of the study and collection, analysis, and interpretation of data and in writing the Conclusions manuscript. The prevalence of burnout among grade thirteen stu - dents in the selected educational zone, Sri Lanka is high Publisher’s Note (28.8%, 95% CI = 25.0–32.7%). Multivariable analysis elic- Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. ited multiple statistically significant correlates that are directly related to the academic endeavours, including a Received: 2 January 2018 Accepted: 18 May 2018 multitude of factors related to the school environment, school curriculum, study enthusiasm, study support, and future expectations. In light of these findings, it is recommended to encourage the positive and supportive References 1. Schaufeli WB, Martinez IM, Pinto AM, Salanova M, Bakker AB. Burnout and involvement of parents and teachers for students’ aca- engagement in university students: a cross-national study. J Cross Cult demic endeavours and to strengthen the counseling ser- Psychol. 2002;33:464–81. vices at the school level to rectify the problems related 2. Yavuz G, Dogan N. Maslach Burnout Inventory-Student Survey (MBI-SS): a validity study. Procedia Soc Behav Sci. 2014;116:2453–7. to burnout among collegiate cycle students in Sri Lanka. 3. Hu Q, Schaufeli WB. The factorial validity of the Maslach Burnout The present study findings broaden the scanty evidence Inventory-Student Survey in China. Psychol Rep. 2009;105:394–408. base pertaining to the magnitude and the associations of 4. Rostami Z, Abedi MR, Schaufeli WB, Ahmadi SA, Sadeghi AH. The psychometric characteristics of Maslach Burnout Inventory Student student burnout in the South Asian context. Survey: a study students of Isfahan University. Zahedan J Res Med Sci. 2014;16:55–8. 5. Ilic M, Todorovic Z, Jovanovic M, Ilic I. Burnout syndrome among medical Abbreviations students at one university in Serbia: validity and reliability of the Maslach AOR: adjusted odds ratio; CY: cynicism; EX: exhaustion; LKR: Sri Lankan Rupees; Burnout Inventory-Student Survey. Behav Med. 2017;43:323–8. MBI-SS: Maslach Burnout Inventory-Student Survey; OR: odds ratio; rPE: 6. Maroco J, Tecedeiro M. Maslach Burnout Inventory-Student Survey: reduced professional efficacy; SAQ: self-administered questionnaire; SE: stand- Portugal-Brazil cross-cultural adaptation. Psicol Saúde Doenças. ard error; SD: standard deviation; 95% CI: 95% confidence interval. 2009;10:227–35. 7. Galán F, Sanmartín A, Polo J, Giner L. Burnout risk in medical students Authors’ contributions in Spain using the Maslach Burnout Inventory-Student Survey. Int Arch NDW, DSD and GSA were involved in the conception and design of the study. Occup Environ Health. 2011;84:453–9. NDW collected, analysed and interpreted data. DSD and GSA made substantial 8. Oliva Costa E, Santos A, Abreu Santos A, Melo E, Andrade T. Burnout Syn- contribution to data analysis and interpretation. NDW prepared the manu- drome and associated factors among medical students: a cross-sectional script. DSD and GSA made substantial contribution to revise the manuscript. study. Clinics. 2012;67:573–9. All authors read and approved the final manuscript. 9. Virtanen T, Lerkkanen MK, Poikkeus AM, Kuorelahti M. Student behavioral disengagement, school truancy and burnout in Finnish junior high Author details schools: latent profile analysis. Presented at ECER 2014, The Past, the Department of Community Medicine, Faculty of Medicine and Allied Sci- Present and the Future of Educational Research, Porto, 2014. Portugal. ences, Rajarata University of Sri Lanka, Saliyapura 50008, Sri Lanka. Depart- 10. Tuominen-Soini H, Salmela-Aro K. Schoolwork engagement and burnout ment of Community Medicine, Faculty of Medicine, University of Peradeniya, among Finnish high school students and young adults: profiles, progres- Peradeniya 20400, Sri Lanka. Teaching Hospital-Kandy, Kandy 20000, Sri sions, and educational outcomes. Dev Psychol. 2014;50(3):649–62. Lanka. 11. Schraml K, Perski A, Grossi G, Simonsson-Sarnecki M. Stress symptoms among adolescents: the role of subjective psychosocial conditions, Acknowledgements lifestyle, and self-esteem. J Adolesc. 2011;34(5):987–96. Authors would like to acknowledge all the students who participated in the 12. Yang HJ. Factors affecting student burnout and academic achievement in study and the principals and the teaching staff of the selected schools for their multiple enrollment programs in Taiwan’s technical-vocational colleges. support. Int J Educ Dev. 2004;24(3):283–301. Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 11 of 11 13. Yang HJ, Farn CK. An investigation the factors affecting MIS student General Survey (MBI-DV ) in individuals with and without clinical burnout. burnout in technical-vocational college. Comput Human Behav. Stress Health. 2005;21:17–25. 2005;21(6):917–32. 35. Maslach C, Leiter MP. Understanding the burnout experience: 14. Zhang Y, Gan Y, Cham H. Perfectionism, academic burnout and engage- recent research and its implications for psychiatry. World Psychiatry. ment among Chinese college students: a structural equation modeling 2016;15:103–11. analysis. Personal Individ Differ. 2007;43(6):1529–40. 36. Kitaoka-Higashiguchi K, Morikawa Y, Miura K, Sakurai M, Ishizaki M, Kido T, 15. Jacobs SR, Dodd D. Student burnout as a function of personality, social et al. Burnout and risk factors for arteriosclerotic disease: follow-up study. support, and workload. J Coll Stud Dev. 2003;44(3):291–303. J Occup Health. 2009;51:123–31. 16. Slivar B. The syndrome of burnout, self-image, and anxiety with grammar 37. Brenninkmeijer V. How to conduct research on burnout: advantages and school students. Horizons Psychol. 2001;10:21–32. disadvantages of a unidimensional approach in burnout research. Occup 17. Salmela-Aro K, Kiuru N, Leskinen E, Nurmi JE. School burnout inventory Environ Med. 2003;60:16i–20i. (SBI) reliability and validity. Eur J Psychol Assess. 2009;25(1):48–57. 38. Wickramasinghe ND, Dissanayake DS, Abeywardena GS. Clinical validity 18. Salmela-Aro K, Kiuru N, Pietikïnen M, Jokela J. Does school matter? The and diagnostic accuracy of the Maslach Burnout Inventory-Student role of school context in adolescents’ school-related burnout? Eur Psy- Survey. Manuscript submitted for publication. 2017. chol. 2008;13:12–23. 39. Maslach C, Jackson SE, Leiter MP. Maslach burnout inventory manual. 3rd 19. Shetty A, Shetty A, Hegde MN, Narasimhan D, Shetty S. Stress and ed. Palo Alto: Consulting Psychologists Press; 1996. burnout assessment among post graduate dental students. Nitte Univ J 40. Lumley T. Analysis of complex survey samples. J Stat Softw. Health Sci. 2015;5(1):31–6. 2004;9(1):1–19. 20. Bera T, Mandal A, Bhattacharya S, Biswas NM, Ghosh A, Bera S. Burnout 41. Maletta H. Weighting. 2007. http://www.spsst ools.net/Tutor ials/WEIGH among medical students—a study across three medical colleges in TING.pdf. Accessed 17 Mar 2015. Eastern India. Ind Med Gaz. 2013;356–9. http://medin d.nic.in/ice/t13/i9/ 42. Speybroeck N, Devleesschauwer B, Joseph L, Berkvens D. Misclassification icet1 3i9p3 56.pdf errors in prevalence estimation: bayesian handling with care. Int J Public 21. Department of Examinations-Sri Lanka. Statistical handbook 2008–2010. Health. 2013;58(5):791–5. 2011. http://www.doene ts.lk/exam/docs/ebook s/Stati stica l-Handb ook- 43. Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation 2008-2010.pdf. Accessed 21 June 2016. study of the number of events per variable in logistic regression analysis. 22. UNICEF Sri Lanka. National survey on emerging issues among adoles- J Clin Epidemiol. 1996;49(12):1373–9. cents in Sri Lanka. 2004. https ://www.unice f.org/srila nka/Full_Repor t.pdf. 44. Zhang X, Klassen RM, Wang Y. Academic burnout and motivation of Accessed 12 Jan 2016. Chinese secondary students. Int J of Soc Sci Human. 2013;3(2):134–8. 23. Perera H. Mental health of adolescent schoolchildren in Sri Lanka—a 45. Farber BA. Stress and burnout in suburban teachers. J Educ Res. national survey. Sri Lanka J Child Health. 2004;33:78–81. 1984;77(6):325–31. 24. Rodrigo C, Welgama S, Gurusinghe J, Wijeratne T, Jayananda G, Rajapakse 46. Jackson SE, Schwab RL, Schuler RS. Toward an understanding of the S. Symptoms of anxiety and depression in adolescent students; a per- burnout phenomenon. J Appl Psychol. 1986;71(4):630–40. spective from Sri Lanka. Child Adolesc Psychiatry Ment Health. 2010;4:10. 47. Russell DW, Altmaier E, Van Velzen D. Job-related stress, social sup- 25. Lwanga SK, Lemeshow S. Sample size determination in health studies: a port, and burnout among classroom teachers. J Appl Psychol. practical manual. Geneva: World Health Organization; 1991. http://apps. 1987;72(2):269–74. who.int/iris/bitst ream/10665 /40062 /1/92415 44058 _%28p1-p22%29.pdf. 48. Pines A, Maslach C. Characteristics of staff burnout in mental health set - Accessed 14 Jan 2016. tings. Psychiatr Serv. 1978;29(4):233–7. 26. Shin H, Puig A, Lee J, Lee JH, Lee SM. Cultural validation of the 49. Vasalampi K, Salmela-Aro K, Nurmi JE. Adolescents’ self-concordance, Maslach Burnout Inventory for Korean students. Asia Pacific Educ Rev. school engagement, and burnout predict their educational trajectories. 2011;12(4):633–9. Eur Psychol. 2009;14(4):332–41. 27. Aypay A. Elementary School Student Burnout Scale for Grades 6–8: 50. Ryan RM, Connell JP. Perceived locus of causality and internalization: a study of validity and reliability. Kuram ve Uygulamada Egit Bilim. examining reasons for acting in two domains. J Pers Soc Psychol. 2011;11(2):520–7. 1989;57(5):749–61. 28. Lee J, Puig A, Kim YB, Shin H, Lee JH, Lee SM. Academic burnout profiles 51. Sheldon KM, Elliot AJ. Goal striving, need satisfaction, and longitu- in Korean adolescents. Stress Health. 2010;26(5):404–16. dinal well-being: the self-concordance model. J Pers Soc Psychol. 29. Wickramasinghe ND, Dissanayake DS, Abeywardena GS. Factorial struc- 1999;76(3):482–97. ture, reliability and validity evidence of the Maslach Burnout Inventory- 52. Sheldon KM, Elliot AJ. Not all personal goals are personal: comparing Student Survey in Sri Lanka. Manuscript submitted for publication. 2017. autonomous and controlled reasons for goals as predictors of effort and 30. Maslach C, Leiter MP, Schaufeli W. Measuring burnout. In: Cooper CL, attainment. Pers Soc Psychol Bull. 1998;24(5):546–57. Cartwright S, editors. The Oxford handbook of organizational well-being. 53. Sheldon KM, Ryan RM, Deci EL, Kasser T. The independent effects of goal Oxford: Oxford University; 2009. contents and motives on well-being: it’s both what you pursue and why 31. Maslach C, Schaufeli WB, Leiter MP. Job burnout. Annu Rev Psychol. you pursue it. Pers Soc Psychol Bull. 2004;30(4):475–86. 2001;52:397–422. 54. Bonell C, Fletcher A, Fitzgerald-Yau N, Hale D, Allen E, Elbourne D, Legood 32. Schaufeli WB, Bakker AB, Hoogduin K, Schaap C, Kladler A. On the clinical R. Initiating change locally in bullying and aggression through the school validity of the Maslach Burnout Inventory and the burnout measure. environment (INCLUSIVE): a pilot randomised controlled trial. Health Psychol Health. 2001;16:565–82. Technol Assess. 2015;19(53):1–10. 33. Schaufeli WB, Leiter MP, Maslach C, Michael PL, Christina M. Burnout: 55. Sampasa-Kanyinga H, Roumeliotis P, Xu H. Associations between 35 years of research and practice. Career Dev Int. 2009;14:204–20. cyberbullying and school bullying victimization and suicidal idea- 34. Roelofs J, Verbraak M, Keijsers GPJ, de Bruin MBN, Schmidt AJM. Psycho- tion, plans and attempts among Canadian schoolchildren. PLoS ONE. metric properties of a Dutch version of the Maslach Burnout Inventory 2014;9(7):e102145. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Child and Adolescent Psychiatry and Mental Health Springer Journals

Prevalence and correlates of burnout among collegiate cycle students in Sri Lanka: a school-based cross-sectional study

Free
11 pages
Loading next page...
 
/lp/springer_journal/prevalence-and-correlates-of-burnout-among-collegiate-cycle-students-Fo4hLmTGwT
Publisher
BioMed Central
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Psychiatry; Pediatrics; Clinical Psychology; Child and Adolescent Psychiatry; Forensic Psychiatry
eISSN
1753-2000
D.O.I.
10.1186/s13034-018-0238-z
Publisher site
See Article on Publisher Site

Abstract

Background: Even though the concept of burnout has been widely explored across the globe, the evidence base on burnout among high school students in the South Asian context is scanty. Against the backdrop of ever-increasing educational demands and expectations, the present study was designed to determine the prevalence and correlates of burnout among collegiate cycle students in Sri Lanka. Methods: A school-based cross-sectional study was conducted among 872 grade thirteen students in 15 govern- ment schools in an educational zone, Kegalle district, Sri Lanka selected by a stratified cluster sampling technique. The validated Sinhala version of the 15-item Maslach Burnout Inventory-Student Survey (MBI-SS) was used to assess burnout. The adjusted prevalence of burnout was computed based on the clinically validated cut-off values using the “exhaustion + 1” criterion. Multivariable logistic regression was carried out using backward elimination method to quantify the association between burnout and selected correlates identified at bivariate analysis at p value less than 0.05. Results: The response rate was 91.3% (n = 796). The adjusted prevalence of burnout among grade thirteen students was 28.8% (95% CI = 25.0–32.7%). Multivariable analysis elicited a multitude of statistically significant associations with burnout when controlled for other factors included in the model (p < 0.05). Perceived satisfaction related to the school environment (classroom and library facilities), school curriculum (scope, relevance, and difficulty of the subject content), study enthusiasm (preferred subject stream), study support (support from parents to teachers), and future expectations (personal and parental expectations) emerged as statistically significant negative associations with burn- out, whereas having to encounter disturbances while studying and being subjected to bullying at school emerged as statistically significant positive associations with burnout. Conclusions: The burnout prevalence among grade thirteen students in the selected educational zone, Sri Lanka is high. Most of the significant correlates of burnout are directly related to the academic endeavours. It is recommended to strengthen the counseling services at the school level to rectify the problems related to burnout among collegiate cycle students in Sri Lanka. Keywords: Student burnout, MBI-SS, Prevalence, Correlates, Collegiate cycle, Sri Lanka *Correspondence: nuwick74@yahoo.com Department of Community Medicine, Faculty of Medicine and Allied Sciences, Rajarata University of Sri Lanka, Saliyapura 50008, Sri Lanka Full list of author information is available at the end of the article © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/ publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 2 of 11 burnout among Indian undergraduate and postgraduate Background students substantially varies from 10.2 to 52.0% [19, 20]. The concept of student burnout has been in the limelight While there is a paucity of literature assessing burnout since the introduction of the Maslach Burnout Inventory- among high school students in the South Asian context, Student Survey (MBI-SS) in 2002. In accordance with the thus far there is no published literature related to student original definition of burnout, student burnout is defined burnout in Sri Lanka. as, “a three-dimensional syndrome that is characterised In Sri Lanka, the collegiate cycle in the education sys- by feelings of exhaustion due to the demands of study- tem (consists of grade 12 and grade 13) leads to the Gen- ing, a cynical attitude of withdrawal and detachment, eral Certificate of Examination Advanced Level, which and reduced professional efficacy regarding academic is the national level selection examination for state uni- requirements” [1]. In the global literature on student versity admissions. National statistics indicate that the burnout research, the MBI-SS has been cited as the most examination has become extremely competitive for Sri widely used research instrument to assess burnout in Lankan collegiate cycle students, who are in the age range different student populations [2, 3]. Furthermore, the of 17–19 years [21]. Evidence suggests that approximately validity and the reliability including the three-factor con- 40% of adolescents found it stressful to cope with the aca- ceptualisation of MBI-SS have been confirmed in a num - demic pressures exerted on them by parents and teachers ber of different student populations in different countries [22] and almost one in five adolescents in schools have [1, 3–6]. clinically relevant mental health problems with a sub- Amongst the wealth of research concerning student stantial proportion having symptoms classified as defi - burnout, a vast majority has focused on assessing burn- nite or severe, while educational performance is reported out among university undergraduates [1, 3, 7, 8]. How- as the most impacted area of life [23]. Furthermore, the ever, against the backdrop of ever-increasing educational prevalence of mental health problems such as depres- demands and expectations, the amount of research con- sion and anxiety among the collegiate cycle students in ducted among high school students is scanty. Moreover, Sri Lanka is high with examinations being the most com- within the scanty evidence base, the reported prevalence monly cited cause [24]. Though academic endeavours of burnout reflects a substantial variation depending on are usually considered as the ostensible reason for the the definitions and criteria used, intrinsic factors of the resultant mental health problems, other numerous non- samples studied, and the  cut-off values applied for the academic factors could have been the significant contrib - assessment. The prevalence of burnout among Finnish utive reasons. In the light of high prevalence of mental junior high school students was reported as 10.9% [9], health problems in students, it is of utmost importance to while other studies have reported even higher prevalence explore the concept of student burnout, which is directly estimates among high school students [10, 11]. assessing psychological well-being in relation to aca- Studies conducted to explore the associations of burn- demic endeavours. out among different student populations across the globe Given the research vacuum pertaining to student burn- have revealed a multitude of significant associations. out in Sri Lanka, the present study was designed to assess Regarding high school and college students, character- the prevalence and correlates of burnout among colle- istics such as lower self-efficacy [12, 13], maladaptive giate cycle students in a selected educational zone in Sri perfectionism [14], negative temperament [15], negative Lanka. self-image [16], and depressive symptoms [17] are found to have positive associations with burnout. Educational Methods environment related factors such as positive school class Study design and setting atmosphere [18], support from school [18], and academic This school-based cross-sectional study was conducted achievement [12, 15, 17] are negatively associated, while in a selected educational zone in the Kegalle district, high course/work load is positively associated with burn- Sabaragamuwa Province, Sri Lanka from January 2015 out [12, 15]. to April 2015. Altogether there are 144 schools in the Even though the global literature suggests that there is selected educational zone with a total student population a multitude of student burnout correlates, the evidence in of approximately 51,000. the South Asian context is sparse. This existing research gap in the South Asian context needs to be addressed, Participants as differences in the socio-cultural backgrounds, educa - The study population consisted of all grade thirteen stu - tional systems, and the level of education might limit the dents studying in Sinhala medium government schools generalisability of previous study findings. in the educational zone at the time of the study exclud- The scanty evidence base pertaining to student burnout ing students who were unable to read or write in Sinhala in the South Asian context reveals that the prevalence of Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 3 of 11 language. There were 1690 grade thirteen students in The clinically validated cut-off values for the subscale seven category 1AB and 31 category 1C government scores of the MBI-SS were developed by computing schools studying in four main subject streams, viz., Sci- Receiver Operating Characteristic curves using the clini- ence, Arts, Commerce and Technology. Category 1AB cal diagnosis made by a Consultant Psychiatrist as the schools have classes in all subject streams, whereas 1C reference standard in a sample of grade thirteen students schools have classes in Arts and Commerce streams only. in a similar educational setting. The clinical assessment The required sample size was calculated with 95% of burnout by the Consultant Psychiatrist was based on confidence level, 5% absolute precision for an antici - the clinical criteria for the diagnosis of work related neu- pated prevalence of burnout of 30% (as 32% of colle- rasthenia according to the ICD 10 classification [30–35]. giate cycle students were categorised as having burnout Diagnostic accuracy of the MBI-SS test results based on in the validation study of the Sinhala version of the “exhaustion + 1” criterion, which is an accepted crite- MBI-SS) based on standard sample size calculation for- rion used in burnout research [30, 34–37], was assessed mula [25] using OpenEpi software version 3.01. The comparing with the results of the clinical diagnosis. The final sample size was computed as 790, after adjusting clinically validated cut-off values for EX, CY and rPE for a design effect of 2.2 and an anticipated level of non- subscale scores were 12.5, 7.5 and 10.5, respectively. The response of 10%. Stratified cluster sampling technique sensitivity, specificity, positive and negative predictive was used to select the study sample in which a clus- values of the Sinhala version of the 15-item MBI-SS were ter was defined as a grade thirteen class with a median 91.9, 93.2, 86.4 and 96.1%, respectively. Further, the posi- number of 21 students. Hence, the study was conducted tive and negative likelihood ratios were 13.48 and 0.09, in 38 clusters from 15 schools selected according to the respectively [38]. probability-proportional-to-size. The second component of the SAQ, which was intended to gather information related to the correlates Measures of burnout, was developed following an extensive litera- A pre-tested, validated self-administered questionnaire ture search and with the inputs from the experts in the (SAQ), which consisted of two components, was used in fields of psychiatry, psychology, public health, teaching, the present study. student counseling, and medical education. The ques - The first component of the SAQ included the Sin - tionnaire was pre-tested in a sample of grade thirteen hala version of the 15-item MBI-SS in order to assess students in a similar educational setting. The question - the burnout status. The MBI-SS has been used to assess naire consisted of ten sections focused on information burnout status among high school students in several set- related to personal and family characteristics, residence, tings [3, 26–28] and the Sinhala version of the 15-item school environment, curriculum, the pattern of study, MBI-SS was found to be a valid and a reliable instrument support for studies, study enthusiasm, future expecta- to assess burnout among collegiate cycle students in Sri tions, personal behaviours/personal life factors, and Lanka [29]. Out of the total 15 items of the MBI-SS, five, behaviours of others. four and six items are targeted at identifying exhaustion (EX), cynicism (CY) and reduced professional efficacy Procedure (rPE) subscales, respectively. The frequency in which Ethical clearance to conduct the study was obtained the respondents experience feelings related to each sub- from the Ethics Review Committee of the Faculty of scale was assessed using a seven-point, fully anchored Medicine and Allied Sciences, Rajarata University of Sri response format, ranging from 0 (never) to 6 (every day). Lanka (Reference no: ERC/2014/057). Prior to data col- During the validation and cultural adaptation of the lection, administrative clearance was obtained from the Sinhala version of the MBI-SS, a multi-disciplinary panel Zonal Director of Education and the principals of all the of experts in the fields of psychiatry, psychology, public selected schools. The dates for data collection in different health, teaching, student counseling, and medical edu- schools were selected according to the logistic conveni- cation has assessed each item of the questionnaire on ence of the schools in order to minimise the disturbance its relevance, appropriateness, and acceptability in the to the routine academic and other endeavours. Informed local context for assessing burnout among grade thir- written consent was obtained from all the students in teen students. In the confirmatory factor analysis, the each selected classroom. Data collection was took place 15-item three-factor model emerged as an acceptable fit - inside the classrooms and each participant was given a ting model. In addition, the 15-item MBI-SS showed high copy of the printed SAQ to be filled independently. Con - internal consistency (Cronbach’s α > 0.8) and high test– fidentiality of data collected was adhered to strictly and retest reliability (p < 0.001) [29]. the anonymity of the participants was maintained. Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 4 of 11 Data analysis Table 1 Basic characteristics of  the  sample of  grade Data analysis was done by using the SPSS version 17.0. thirteen students (n = 796) After entering, double independent check of entries Characteristic Number Percentage (%) Cumulative was carried out to identify any incompatible entries. percentage (%) The dataset was examined for univariate and multivari - ate outliers using box plots and Mahalanobis distance, Sex respectively. Female 440 55.3 55.3 Scoring of the MBI-SS was carried out according to Male 356 44.7 100.0 the MBI manual instructions [39]. Based on the clini- Religion cally validated cut-off values for the three subscales, a Buddhist 774 97.2 97.2 participant was categorised as having burnout according Hindu 2 0.3 97.5 to the “exhaustion + 1” criterion, in which a participant Christian 16 2.0 99.5 who is having a high score on EX in combination with Islam 1 0.1 99.6 a high score on either of the CY or rPE subscale was Other 3 0.4 100.0 regarded as having burnout [30, 37]. As cluster sampling Monthly family income technique was used to select the study participants and LKR 10,000–20,000 251 31.5 31.5 the number of participants in the different clusters was LKR 20,001–30,000 118 14.8 46.4 not uniform, weighted analysis was conducted to deter- LKR 30,001–40,000 98 12.3 58.7 mine the prevalence of burnout. Each observation in the LKR 40,001–50,000 143 18.0 76.6 sample was assigned a particular weight, which was cal- > LKR 50,001 186 23.4 100.0 culated as the product of inverse selection probabilities Subject stream at each stage of sampling [40, 41]. Incorporating the val- Science 235 29.5 29.5 ues of diagnostic accuracy of validated Sinhala version Arts 276 34.7 64.2 of the MBI-SS [38], the adjusted prevalence of burnout Commerce 229 28.8 93.0 with 95% confidence interval (CI) was calculated in Technology 56 7.0 100.0 grade thirteen students [42]. Total 796 100.0 For the assessment of correlates of burnout, a two-step LKR Sri Lankan Rupees procedure was followed. First, a bivariate analysis was performed to identify potential correlates of burnout. Then, a multivariable analysis using binary logistic regres - The mean age of the grade thirteen students in the sam - sion was conducted to identify the relevant predictors of ple was 18.4  years (SD = 0.32  years). The majority of the burnout and to control for potential confounding among participants were females (n = 440, 55.3%) and 276 stu- the various predictor variables. Categorical data related dents (34.7%) were studying in the Arts subject stream. to predictor variables were amalgamated rationally as dichotomous variables where necessary for the bivariate analysis and the crude odds ratios (OR) were calculated Descriptive statistics of the Sinhala version of the MBI‑SS as the measures of effect with 95% CI. Correlates that subscale scores showed statistical significance at p value less than 0.05 in Table  2 summarizes the mean total scores and the mean the bivariate analysis were included in the multivariable item scores of the three subscales of the MBI-SS Sinhala analysis using backward stepwise elimination method. version. The model produced adjusted odds ratios (AOR) and 95% CI with the significance level for variables of interest. In order to conduct logistic regression, it is recommended to Prevalence of burnout have at least ten observations per an independent variable The prevalence of burnout based on the clinically vali - [43] and the dataset met this criterion. dated cut-off values for each subscale score and the “exhaustion + 1” criterion was 36.8% (95% CI = 33.5– 40.2%). The weighted analysis conducted to compensate for the complex sampling design resulted in a weighted Results prevalence estimate of 31.3% (95% CI = 28.1–34.6%). Characteristics of the sample According to the sensitivity and the specificity of the Sin - Out of the total of 872 grade thirteen students in the 38 hala version of the 15-item MBI-SS, the adjusted preva- identified clusters, 796 students completed the SAQ; lence of burnout among grade thirteen students in the hence, the response rate was 91.3%. Table  1 summarises study was 28.8% (95% CI = 25.0–32.7%). the basic characteristics of the study sample. Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 5 of 11 Table 2 Descriptive statistics of  the  subscale scores Having to encounter disturbances while studying and of  the  MBI-SS Sinhala version among  grade thirteen being subjected to bullying at school emerged as statisti- students (n = 796) cally significant positive associations with burnout. Subscale Mean total score SD Mean item SD score Discussion The present study, which was designed with the objec - EX 11.98 7.16 2.40 1.43 tive of determining the prevalence and correlates of CY 6.80 5.98 1.70 1.49 burnout among grade thirteen students in Sri Lanka, rPE 10.56 6.46 1.76 1.08 addresses an important research vacuum in relation to burnout research in high school students in the South Correlates of burnout in the bivariate analysis Asian context. The study findings reveal that almost one In the bivariate analysis, 35 factors emerged as signifi - in four grade thirteen students is likely to have burnout cant predictors of burnout. These included a number of and burnout is significantly associated with a multitude factors related to the study environment, curriculum, of academic environment related correlates. and behaviours. Table  3 presents the summary of sta- Having a high response rate and computing adjusted tistically significant independent predictor variables of prevalence by compensating for the sampling complexity burnout emerged in the bivariate analysis. and diagnostic uncertainty of the assessment tool pro- vide valid and precise estimates for burnout among grade thirteen students in this study. The prevalence of burnout among the collegiate cycle Multivariable analysis of correlates of burnout students in this study is higher than the reported values All 35 independent predictors identified at bivariate of prevalence of burnout among high school students analysis were included in the multivariable analysis. conducted in different study settings, such as 10.9% None of these predictors had categories with very few among students in Finnish public junior high schools [9], observations, both the dependent and the independ- 14% among Finnish high school students [10], and 12.6% ent variables were dichotomous in nature, and there among middle school and regular secondary school in were no outliers in the data set. Table 4 summarises the Northern China [44]. According to the World Bank sta- results of the multivariable analysis of the correlates of tistics, the total enrolment in tertiary education in Fin- burnout retained in the final model. Out of the 14 fac - land and Sweden are substantially higher than that of Sri tors retained in the final model, 12 factors made unique Lanka; based on which, it can be argued that the com- statistically significant contributions at a p value less petitiveness of the tertiary education enrolment exami- than 0.05. nations and academic endeavour related stress could be Multivariable analysis elicited several statistically sig- higher in the Sri Lankan context as opposed to the study nificant associations with burnout when controlled for settings in other highlighted countries, contributing to other factors included in the model (p < 0.05). Perceived the observed difference of the magnitude of the prob - satisfaction about the facilities available in the classroom, lem of burnout. However, other possible explanations about the library facilities, and about the scope of the such as the crucial differences in the educational context, subject content covered in the curriculum showed statis- burnout assessment tool, assessment cut-off values, and tically significant negative associations with burnout. assessment criteria have to be taken into consideration in Students who thought the content covered in the sub- critically evaluating the research findings for comparative jects are relevant to the curriculum and students who purposes. Provided that there is no universally accepted easily understood the subject content taught in the cur- diagnosis method of burnout, it is important to appreci- riculum had statistically significant lower likelihood of ate that the prevalence estimates reported in the present having burnout in comparison to their counterparts. study is dependent upon the burnout assessment criteria Satisfactory support from parents to satisfactory sup- used in the study. port from teachers were found to have statistically signif- The multivariable analysis revealed that the students icant negative associations with burnout. who encountered disturbances while studying were more The students who selected the subject stream based likely to have burnout in comparison to their counter- on their own decision were less likely to have burnout parts. It can be assumed that the disturbances at study as opposed to those who have selected the current sub- place causing distractions would make the students una- ject stream for other reasons. Students who felt that both ble to study or concentrate on their studies. The result - their own future expectations and parental expectations ant frustration may provoke a feeling of indifference or are encouraging their studies were less likely to have distant attitude towards work. The ultimate result would burnout as opposed to those who did not feel so. Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 6 of 11 Table 3 Statistically significant independent predictor variables of  burnout in  grade thirteen students in  the  bivariate analysis Characteristic Burnout No burnout Total Odds ratio (95% CI) p n (%) n (%) n (%) value Personal characteristics Sex (n = 796) Female 145 (33.0) 295 (67.0) 440 (100.0) 0.7 (0.5–0.9) Male 148 (41.6) 208 (58.4) 356 (100.0) p = 0.012 Residence related factors Satisfaction of study place (n = 705) Satisfactory 139 (32.0) 295 (68.0) 434 (100.0) 0.6 (0.5–0.9) Not satisfactory 116 (42.8) 155 (57.2) 271 (100.0) p = 0.004 Type of disturbances (n = 360) From inside 70 (49.3) 72 (50.7) 142 (100.0) 1.5 (1.0–2.3) From outside/both inside and outside 85 (39.0) 133 (61.0) 218 (100.0) p = 0.054 Disturbances to studies at residence (n = 796) Yes 155 (43.1) 205 (56.9) 360 (100.0) 1.6 (1.2–2.2) No 138 (31.7) 298 (68.3) 436 (100.0) p = 0.001 Duration of travel to school (n = 796) Less than 30 min 131 (33.0) 266 (67.0) 397 (100.0) 0.7 (0.5–0.9) More than 30 min 162 (40.6) 237 (59.4) 399 (100.0) p = 0.026 School environment related factors Facilities of the classroom (n = 796) Satisfactory 81 (17.0) 395 (83.0) 476 (100.0) 0.1 (0.1–0.2) Not satisfactory 212 (66.3) 108 (33.7) 320 (100.0) p < 0.001 School library facilities (n = 796) Satisfactory 194 (33.1) 392 (66.9) 586 (100.0) 0.6 (0.4–0.8) Not satisfactory 99 (47.1) 111 (52.9) 210 (100.0) p < 0.001 School health services (n = 796) Satisfactory 111 (30.4) 254 (69.6) 365 (100.0) 0.6 (0.4–0.8) Not satisfactory 182 (42.2) 249 (57.8) 431 (100.0) p = 0.001 School counseling services (n = 796) Satisfactory 98 (30.9) 219 (69.1) 317 (100.0) 0.7 (0.5–0.9) Not satisfactory 195 (40.7) 284 (59.3) 479 (100.0) p = 0.005 School recreational facilities (n = 796) Satisfactory 85 (28.3) 215 (71.7) 300 (100.0) 0.5 (0.4–0.7) Not satisfactory 208 (41.9) 288 (58.1) 496 (100.0) p < 0.001 Curriculum related factors Scope of subject content (n = 766) Satisfactory 115 (22.5) 395 (77.5) 510 (100.0) 0.1 (0.1–0.2) Not satisfactory 175 (68.4) 81 (31.6) 256 (100.0) p < 0.001 Amount of assignments/workload (n = 777) Satisfactory 103 (28.5) 259 (71.5) 362 (100.0) 0.5 (0.4–0.7) Not satisfactory 180 (43.7) 232 (56.3) 412 (100.0) p < 0.001 Relevance of subject area (n = 781) Satisfactory 141 (30.5) 322 (69.5) 463 (100.0) 0.5 (0.4–0.7) Not satisfactory 144 (45.3) 174 (54.7) 318 (100.0) p < 0.001 Difficulty in understanding (n = 796) Easily understood 79 (18.9) 338 (81.1) 417 (100.0) 0.2 (0.1–0.3) Not easily understood 214 (56.5) 165 (43.5) 379 (100.0) p < 0.001 Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 7 of 11 Table 3 (continued) Characteristic Burnout No burnout Total Odds ratio (95% CI) p n (%) n (%) n (%) value Study pattern related factors Duration of study per day (n = 796) Less than 2 h 119 (42.5) 161 (57.5) 280 (100.0) 1.4 (1.1–2.0) More than 2 h 174 (33.7) 342 (66.3) 516 (100.0) p = 0.014 Methods of study (n = 796) Interactive methods with others 271 (36.0) 482 (64.0) 753 (100.0) 0.5 (0.3–0.9) No interaction with others 22 (51.2) 21 (48.8) 43 (100.0) p = 0.045 Answering past papers (n = 796) Yes 177 (31.8) 379 (68.2) 556 (100.0) 0.5 (0.4–0.7) No 116 (48.3) 124 (51.7) 240 (100.0) p < 0.001 Study support related factors Support from parents (n = 792) Satisfactory 208 (30.4) 476 (69.6) 684 (100.0) 0.1 (0.1–0.2) Not satisfactory 84 (77.8) 24 (22.2) 108 (100.0) p < 0.001 Support from colleagues (n = 759) Satisfactory 160 (33.8) 313 (66.2) 473 (100.0) 0.7 (0.5–1.0) Not satisfactory 117 (40.9) 169 (59.1) 286 (100.0) p = 0.050 Support from teachers (n = 780) Satisfactory 231 (33.3) 462 (66.7) 693 (100.0) 0.3 (0.2–0.5) Not satisfactory 54 (62.1) 33 (37.9) 87 (100.0) p < 0.001 Study enthusiasm related factors Subject stream (n = 740) Science 105 (44.7) 130 (55.3) 235 (100.0) 1.8 (1.3–2.5) Arts/Commerce 154 (30.5) 351 (69.5) 505 (100.0) p < 0.001 Preference of subject stream (n = 796) Own preference 230 (32.5) 478 (67.5) 708 (100.0) 0.2 (0.1–0. 3) Not own preference 63 (71.6) 25 (28.4) 88 (100.0) p < 0.001 Opinion of academic performance (n = 796) Satisfactory 112 (28.6) 279 (71.4) 391 (100.0) 0.5 (0.4–0.7) Not satisfactory 121 (44.7) 224 (55.3) 405 (100.0) p < 0.001 Average marks (n = 577) Less than 50 101 (46.3) 117 (53.7) 218 (100.0) 1.6 (1.1–2.2) More than 50 126 (35.1) 233 (64.9) 359 (100.0) p = 0.007 Future expectation related factors Personal expectations (n = 774) Encouraging 249 (34.8) 467 (65.2) 716 (100.0) 0.3 (0.2–0.6) Not encouraging 36 (62.1) 22 (37.9) 58 (100.0) p < 0.001 Expectations of parents (n = 772) Encouraging 241 (33.8) 473 (66.2) 714 (100.0) 0.2 (0.1–0.3) Not encouraging 44 (75.9) 14 (24.1) 58 (100.0) p < 0.001 Expectations of teachers (n = 761) Encouraging 232 (34.6) 439 (65.4) 671 (100.0) 0.4 (0.3–0.7) Not encouraging 49 (54.4) 41 (45.6) 90 (100.0) p < 0.001 Expectations of colleagues (n = 677) Encouraging 162 (33.5) 322 (66.5) 484 (100.0) 0.7 (0.5–0.9) Not encouraging 83 (43.0) 110 (57.0) 193 (100.0) p = 0.020 Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 8 of 11 Table 3 (continued) Characteristic Burnout No burnout Total Odds ratio (95% CI) p n (%) n (%) n (%) value Personal behaviours/personal life factors Love affair (n = 796) Yes 87 (43.3) 114 (56.7) 201 (100.0) 1.4 (1.1–2.0) No 206 (34.6) 389 (65.4) 595 (100.0) p = 0.028 Eec ff t of hobbies (n = 769) Useful to studies 67 (31.8) 144 (68.2) 211 (100.0) 0.7 (0.5–1.0) Not useful to studies 219 (39.2) 339 (60.8) 558 (100.0) p = 0.055 Eec ff t of religious activities (n = 780) Useful to studies 216 (35.1) 400 (64.9) 616 (100.0) 0.7 (0.4–1.1) Not useful to studies 69 (42.1) 95 (57.9) 164 (100.0) p = 0.098 Factors related to behaviours of other people Household substance abuse (n = 796) Yes 152 (46.2) 177.(53.8) 329 (100.0) 2.0 (1.5–2.7) No 141 (30.2) 326 (69.8) 467 (100.0) p < 0.001 Colleagues substance abuse (n = 796) Yes 130 (42.9) 173 (57.1) 303 (100.0) 1.5 (1.1–2.0) No 163 (33.1) 330 (66.9) 493 (100.0) p = 0.005 Peer pressure for substance abuse (n = 796) Yes 21 (53.8) 18 (46.2) 39 (100.0) 2.1 (1.1–3.9) No 272 (35.9) 485 (64.1) 757 (100.0) p = 0.024 Subjected to bullying at school (n = 796) Yes 160 (51.8) 149 (48.2) 309 (100.0) 2.9 (2.1–3.9) No 133 (27.3) 354 (72.7) 487 (100.0) p < 0.001 Total is not equal to 796 due to responses not included in the analysis and/or missing values Table 4 Correlates of burnout in the sample of 796 grade thirteen students in the multivariable analysis Factor B SE Wald df p value Odds ratio 95% CI Disturbance to studies: yes 0.740 0.366 4.083 1 0.043 2.1 1.1–4.3 Facilities of the classroom: satisfactory − 2.151 0.327 43.156 1 < 0.001 0.1 0.1–0.2 Library facilities: satisfactory − 0.691 0.332 4.338 1 0.037 0.5 0.3–0.9 School health services: satisfactory − 0.578 0.328 3.080 1 0.079 0.6 0.3–1.1 Scope of subject content: satisfactory − 1.464 0. 357 16.818 1 < 0.001 0.2 0.1–0.5 Relevance of subject area: satisfactory − 1.900 0.421 20.333 1 < 0.001 0.2 0.1–0.3 Difficulty in understanding: easily understood − 1.380 0.340 16.444 1 < 0.001 0.3 0.1–0.5 Duration of study: less than 2 h per day 0.574 0.325 3.127 1 0.077 1.8 0.9–3.4 Support from parents: satisfactory − 1.800 0.488 13.612 1 < 0.001 0.2 0.1–0.4 Support from teachers: satisfactory − 1.084 0.449 5.832 1 0.016 0.3 0.1–0.8 Preference of subject stream: own preference − 1.995 0.486 16.825 1 < 0.001 0.1 0.1–0.4 Personal expectations: encouraging − 1.251 0.576 4.716 1 0.030 0.3 0.1–0.9 Expectations of parents: encouraging − 1.002 0.346 8.375 1 0.004 0.4 0.2–0.7 Subjected to bullying at school: yes 0.981 0.302 10.557 1 0.001 2.7 1.5–4.8 Italic values indicate significance of p value (p < 0.05) df degree of freedom; SE standard error; 95% CI 95% confidence interval Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 9 of 11 be a cynical attitude of withdrawal and detachment from the negative association elicited between burnout and academic activities, which is an important aspect of stu- the support from parents to teachers. When academic dent burnout. endeavour related stress increases, the students are in School environment related factors, such as having sat- search for social support to cope with the stress and the isfactory classroom and library facilities was associated mental pressure. with lower likelihood of having burnout and these find - Students, who selected the subject stream that they ings are consistent with the existing evidence [18]. It can are following based on their own decision, were less be assumed that the limited facilities or resources avail- likely to have burnout as opposed to those who opted able at school environment are unable to inspire students to follow the current subject stream for other reasons. to work hard, though in a paradox, the limited resources Similarly, Vasalampi, Salmela-Aro and Nurmi [49] have necessitate students to work hard to achieve better shown that self-concordance was associated with stu- results. The lack of psychological stimulation originated dent burnout. According to Ryan and Connell [50], by the lack of resources may aggravate students’ disinter- even though it is believed that personal goals are self- est towards studies. Furthermore, it may contribute to determined, they are not solely originated based on the further mental exhaustion of students owing to the idea intrinsic values and personal interests. It is believed of having to work hard. These reasons may have lead to that individuals are inclined to adopt their goals for the significant association elicited in the study. external reasons such as social pressure or expectations Curriculum related factors, such as students’ satisfac- of others [51]. These arguments have generated the tion of the scope of the subject content covered in the basis for the self-concordance model for goal selection curriculum and relevance of the subject to the curricu- [51, 52]. Furthermore, according to literature, pursuit lum, emerged as statistically significant predictors with of goals for internalised reasons promotes sustained lower likelihood of having burnout. In a similar vein, efforts leading to a better goal progress [53]. studies suggest a positive association of course work- In relation to future expectations on academic load with burnout [12, 15]. Factors such as the wide endeavours, students who felt that both their own scope covered in the subject content and the difficulty in future expectations and parental expectations are understanding the subject content demand students to encouraging their studies were less likely to have burn- make an extra effort in relation to their academic endeav - out as opposed to those who felt these expectations are ours. Students in failing to understand the deep and dif- not encouraging to their studies. As discussed above, ficult underpinning subject content resort to cram facts as per the self-concordance model for goal selection, a and theories, which hinders the opportunity to acquire certain element of external influence is incorporated in higher-order generic competencies. This results in the the selection of academic goals. In the Sri Lankan con- imbalance between the abilities of the students and the text, parental expectations play an important role in expectations demanded by the academic environment this regard. Hence, the subjective perception of encour- leading to the excess amount of stress. This progressive aging parental expectations together with personal vicious phenomenon culminates in burnout, which is expectations may bring about a better goal progress. recognised as a result of failed attempts to cope with a Thus, the students who are having such subjective per - variety of negative stress situations [45]. ceptions may be less likely to develop burnout. In relation to support received for studies from oth- In the present study, the students who were subjected ers, students who perceived that they receive satis- to bullying at school were more likely to have burnout. factory support from their parents and satisfactory This finding is congruent with existing evidence sug - support from their teachers were found to have statisti- gesting that bullying in schools has a negative impact cally significant negative associations with burnout and on students’ mental health [54, 55]. Being bullied is a these findings are consistent with few other study find - particularly intense and traumatic form of stress, which ings [12, 13, 15]. Social support is regarded as one of the usually involves a loss of control, and poses a threat to a important aspects in explaining the burnout phenom- person’s wellbeing. enon [46]. Social support of an individual includes peo- In sum, the present study revealed several signifi - ple who are caring and valuing the individual and the cant correlates of burnout and the findings are congru - people that can be relied upon [12]. Social support has ent with other study findings. Interestingly, amongst been identified as a resource that enables individuals the significant correlates, almost all were either corre - to cope with stress [47]. Pines and Maslach [48] high- lates that are directly related to the academic endeav- lighted the fact that the social support has been iden- ours or correlates that have a direct effect on academic tified as both a preventive mechanism and a remedy endeavours. against burnout. These collective evidence, elucidate Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 10 of 11 Competing interests Limitations The authors declare that they have no competing interests. There are some limitations related to the study that need to be taken into consideration in interpreting the results. Availability of data and materials The datasets used and analysed during the present study are available from Owing to the fact that the present study was conducted the corresponding author on reasonable request. in a selected educational zone in the Kegalle district, Sri Lanka the generalisation of the study findings to other Consent for publication Not applicable. study settings should be done with caution, consider- ing the educational and cultural differences. Due to the Ethics approval and consent to participate cross-sectional nature of the study, a temporal relation- Ethical clearance to conduct the study was obtained from the Ethics Review Committee of the Faculty of Medicine and Allied Sciences, Rajarata University ship between burnout and correlates could not be estab- of Sri Lanka (Reference no: ERC/2014/057). Informed written consent from all lished based on the present study findings. In the present the participants were obtained prior to data collection. (All the participants study, even though multivariable analysis accounted for were above the age of 16 years). confounding, the effect of unknown confounders could Funding not be accounted for. This work was supported by the University Grants Commission-Sri Lanka, under the Postgraduate Research Grant scheme [Grant number: UGC/DRIC/ PG/2015(I)/RUSL/01]. The funding body did not involve in the design of the study and collection, analysis, and interpretation of data and in writing the Conclusions manuscript. The prevalence of burnout among grade thirteen stu - dents in the selected educational zone, Sri Lanka is high Publisher’s Note (28.8%, 95% CI = 25.0–32.7%). Multivariable analysis elic- Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. ited multiple statistically significant correlates that are directly related to the academic endeavours, including a Received: 2 January 2018 Accepted: 18 May 2018 multitude of factors related to the school environment, school curriculum, study enthusiasm, study support, and future expectations. In light of these findings, it is recommended to encourage the positive and supportive References 1. Schaufeli WB, Martinez IM, Pinto AM, Salanova M, Bakker AB. Burnout and involvement of parents and teachers for students’ aca- engagement in university students: a cross-national study. J Cross Cult demic endeavours and to strengthen the counseling ser- Psychol. 2002;33:464–81. vices at the school level to rectify the problems related 2. Yavuz G, Dogan N. Maslach Burnout Inventory-Student Survey (MBI-SS): a validity study. Procedia Soc Behav Sci. 2014;116:2453–7. to burnout among collegiate cycle students in Sri Lanka. 3. Hu Q, Schaufeli WB. The factorial validity of the Maslach Burnout The present study findings broaden the scanty evidence Inventory-Student Survey in China. Psychol Rep. 2009;105:394–408. base pertaining to the magnitude and the associations of 4. Rostami Z, Abedi MR, Schaufeli WB, Ahmadi SA, Sadeghi AH. The psychometric characteristics of Maslach Burnout Inventory Student student burnout in the South Asian context. Survey: a study students of Isfahan University. Zahedan J Res Med Sci. 2014;16:55–8. 5. Ilic M, Todorovic Z, Jovanovic M, Ilic I. Burnout syndrome among medical Abbreviations students at one university in Serbia: validity and reliability of the Maslach AOR: adjusted odds ratio; CY: cynicism; EX: exhaustion; LKR: Sri Lankan Rupees; Burnout Inventory-Student Survey. Behav Med. 2017;43:323–8. MBI-SS: Maslach Burnout Inventory-Student Survey; OR: odds ratio; rPE: 6. Maroco J, Tecedeiro M. Maslach Burnout Inventory-Student Survey: reduced professional efficacy; SAQ: self-administered questionnaire; SE: stand- Portugal-Brazil cross-cultural adaptation. Psicol Saúde Doenças. ard error; SD: standard deviation; 95% CI: 95% confidence interval. 2009;10:227–35. 7. Galán F, Sanmartín A, Polo J, Giner L. Burnout risk in medical students Authors’ contributions in Spain using the Maslach Burnout Inventory-Student Survey. Int Arch NDW, DSD and GSA were involved in the conception and design of the study. Occup Environ Health. 2011;84:453–9. NDW collected, analysed and interpreted data. DSD and GSA made substantial 8. Oliva Costa E, Santos A, Abreu Santos A, Melo E, Andrade T. Burnout Syn- contribution to data analysis and interpretation. NDW prepared the manu- drome and associated factors among medical students: a cross-sectional script. DSD and GSA made substantial contribution to revise the manuscript. study. Clinics. 2012;67:573–9. All authors read and approved the final manuscript. 9. Virtanen T, Lerkkanen MK, Poikkeus AM, Kuorelahti M. Student behavioral disengagement, school truancy and burnout in Finnish junior high Author details schools: latent profile analysis. Presented at ECER 2014, The Past, the Department of Community Medicine, Faculty of Medicine and Allied Sci- Present and the Future of Educational Research, Porto, 2014. Portugal. ences, Rajarata University of Sri Lanka, Saliyapura 50008, Sri Lanka. Depart- 10. Tuominen-Soini H, Salmela-Aro K. Schoolwork engagement and burnout ment of Community Medicine, Faculty of Medicine, University of Peradeniya, among Finnish high school students and young adults: profiles, progres- Peradeniya 20400, Sri Lanka. Teaching Hospital-Kandy, Kandy 20000, Sri sions, and educational outcomes. Dev Psychol. 2014;50(3):649–62. Lanka. 11. Schraml K, Perski A, Grossi G, Simonsson-Sarnecki M. Stress symptoms among adolescents: the role of subjective psychosocial conditions, Acknowledgements lifestyle, and self-esteem. J Adolesc. 2011;34(5):987–96. Authors would like to acknowledge all the students who participated in the 12. Yang HJ. Factors affecting student burnout and academic achievement in study and the principals and the teaching staff of the selected schools for their multiple enrollment programs in Taiwan’s technical-vocational colleges. support. Int J Educ Dev. 2004;24(3):283–301. Wickramasinghe et al. Child Adolesc Psychiatry Ment Health (2018) 12:26 Page 11 of 11 13. Yang HJ, Farn CK. An investigation the factors affecting MIS student General Survey (MBI-DV ) in individuals with and without clinical burnout. burnout in technical-vocational college. Comput Human Behav. Stress Health. 2005;21:17–25. 2005;21(6):917–32. 35. Maslach C, Leiter MP. Understanding the burnout experience: 14. Zhang Y, Gan Y, Cham H. Perfectionism, academic burnout and engage- recent research and its implications for psychiatry. World Psychiatry. ment among Chinese college students: a structural equation modeling 2016;15:103–11. analysis. Personal Individ Differ. 2007;43(6):1529–40. 36. Kitaoka-Higashiguchi K, Morikawa Y, Miura K, Sakurai M, Ishizaki M, Kido T, 15. Jacobs SR, Dodd D. Student burnout as a function of personality, social et al. Burnout and risk factors for arteriosclerotic disease: follow-up study. support, and workload. J Coll Stud Dev. 2003;44(3):291–303. J Occup Health. 2009;51:123–31. 16. Slivar B. The syndrome of burnout, self-image, and anxiety with grammar 37. Brenninkmeijer V. How to conduct research on burnout: advantages and school students. Horizons Psychol. 2001;10:21–32. disadvantages of a unidimensional approach in burnout research. Occup 17. Salmela-Aro K, Kiuru N, Leskinen E, Nurmi JE. School burnout inventory Environ Med. 2003;60:16i–20i. (SBI) reliability and validity. Eur J Psychol Assess. 2009;25(1):48–57. 38. Wickramasinghe ND, Dissanayake DS, Abeywardena GS. Clinical validity 18. Salmela-Aro K, Kiuru N, Pietikïnen M, Jokela J. Does school matter? The and diagnostic accuracy of the Maslach Burnout Inventory-Student role of school context in adolescents’ school-related burnout? Eur Psy- Survey. Manuscript submitted for publication. 2017. chol. 2008;13:12–23. 39. Maslach C, Jackson SE, Leiter MP. Maslach burnout inventory manual. 3rd 19. Shetty A, Shetty A, Hegde MN, Narasimhan D, Shetty S. Stress and ed. Palo Alto: Consulting Psychologists Press; 1996. burnout assessment among post graduate dental students. Nitte Univ J 40. Lumley T. Analysis of complex survey samples. J Stat Softw. Health Sci. 2015;5(1):31–6. 2004;9(1):1–19. 20. Bera T, Mandal A, Bhattacharya S, Biswas NM, Ghosh A, Bera S. Burnout 41. Maletta H. Weighting. 2007. http://www.spsst ools.net/Tutor ials/WEIGH among medical students—a study across three medical colleges in TING.pdf. Accessed 17 Mar 2015. Eastern India. Ind Med Gaz. 2013;356–9. http://medin d.nic.in/ice/t13/i9/ 42. Speybroeck N, Devleesschauwer B, Joseph L, Berkvens D. Misclassification icet1 3i9p3 56.pdf errors in prevalence estimation: bayesian handling with care. Int J Public 21. Department of Examinations-Sri Lanka. Statistical handbook 2008–2010. Health. 2013;58(5):791–5. 2011. http://www.doene ts.lk/exam/docs/ebook s/Stati stica l-Handb ook- 43. Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation 2008-2010.pdf. Accessed 21 June 2016. study of the number of events per variable in logistic regression analysis. 22. UNICEF Sri Lanka. National survey on emerging issues among adoles- J Clin Epidemiol. 1996;49(12):1373–9. cents in Sri Lanka. 2004. https ://www.unice f.org/srila nka/Full_Repor t.pdf. 44. Zhang X, Klassen RM, Wang Y. Academic burnout and motivation of Accessed 12 Jan 2016. Chinese secondary students. Int J of Soc Sci Human. 2013;3(2):134–8. 23. Perera H. Mental health of adolescent schoolchildren in Sri Lanka—a 45. Farber BA. Stress and burnout in suburban teachers. J Educ Res. national survey. Sri Lanka J Child Health. 2004;33:78–81. 1984;77(6):325–31. 24. Rodrigo C, Welgama S, Gurusinghe J, Wijeratne T, Jayananda G, Rajapakse 46. Jackson SE, Schwab RL, Schuler RS. Toward an understanding of the S. Symptoms of anxiety and depression in adolescent students; a per- burnout phenomenon. J Appl Psychol. 1986;71(4):630–40. spective from Sri Lanka. Child Adolesc Psychiatry Ment Health. 2010;4:10. 47. Russell DW, Altmaier E, Van Velzen D. Job-related stress, social sup- 25. Lwanga SK, Lemeshow S. Sample size determination in health studies: a port, and burnout among classroom teachers. J Appl Psychol. practical manual. Geneva: World Health Organization; 1991. http://apps. 1987;72(2):269–74. who.int/iris/bitst ream/10665 /40062 /1/92415 44058 _%28p1-p22%29.pdf. 48. Pines A, Maslach C. Characteristics of staff burnout in mental health set - Accessed 14 Jan 2016. tings. Psychiatr Serv. 1978;29(4):233–7. 26. Shin H, Puig A, Lee J, Lee JH, Lee SM. Cultural validation of the 49. Vasalampi K, Salmela-Aro K, Nurmi JE. Adolescents’ self-concordance, Maslach Burnout Inventory for Korean students. Asia Pacific Educ Rev. school engagement, and burnout predict their educational trajectories. 2011;12(4):633–9. Eur Psychol. 2009;14(4):332–41. 27. Aypay A. Elementary School Student Burnout Scale for Grades 6–8: 50. Ryan RM, Connell JP. Perceived locus of causality and internalization: a study of validity and reliability. Kuram ve Uygulamada Egit Bilim. examining reasons for acting in two domains. J Pers Soc Psychol. 2011;11(2):520–7. 1989;57(5):749–61. 28. Lee J, Puig A, Kim YB, Shin H, Lee JH, Lee SM. Academic burnout profiles 51. Sheldon KM, Elliot AJ. Goal striving, need satisfaction, and longitu- in Korean adolescents. Stress Health. 2010;26(5):404–16. dinal well-being: the self-concordance model. J Pers Soc Psychol. 29. Wickramasinghe ND, Dissanayake DS, Abeywardena GS. Factorial struc- 1999;76(3):482–97. ture, reliability and validity evidence of the Maslach Burnout Inventory- 52. Sheldon KM, Elliot AJ. Not all personal goals are personal: comparing Student Survey in Sri Lanka. Manuscript submitted for publication. 2017. autonomous and controlled reasons for goals as predictors of effort and 30. Maslach C, Leiter MP, Schaufeli W. Measuring burnout. In: Cooper CL, attainment. Pers Soc Psychol Bull. 1998;24(5):546–57. Cartwright S, editors. The Oxford handbook of organizational well-being. 53. Sheldon KM, Ryan RM, Deci EL, Kasser T. The independent effects of goal Oxford: Oxford University; 2009. contents and motives on well-being: it’s both what you pursue and why 31. Maslach C, Schaufeli WB, Leiter MP. Job burnout. Annu Rev Psychol. you pursue it. Pers Soc Psychol Bull. 2004;30(4):475–86. 2001;52:397–422. 54. Bonell C, Fletcher A, Fitzgerald-Yau N, Hale D, Allen E, Elbourne D, Legood 32. Schaufeli WB, Bakker AB, Hoogduin K, Schaap C, Kladler A. On the clinical R. Initiating change locally in bullying and aggression through the school validity of the Maslach Burnout Inventory and the burnout measure. environment (INCLUSIVE): a pilot randomised controlled trial. Health Psychol Health. 2001;16:565–82. Technol Assess. 2015;19(53):1–10. 33. Schaufeli WB, Leiter MP, Maslach C, Michael PL, Christina M. Burnout: 55. Sampasa-Kanyinga H, Roumeliotis P, Xu H. Associations between 35 years of research and practice. Career Dev Int. 2009;14:204–20. cyberbullying and school bullying victimization and suicidal idea- 34. Roelofs J, Verbraak M, Keijsers GPJ, de Bruin MBN, Schmidt AJM. Psycho- tion, plans and attempts among Canadian schoolchildren. PLoS ONE. metric properties of a Dutch version of the Maslach Burnout Inventory 2014;9(7):e102145.

Journal

Child and Adolescent Psychiatry and Mental HealthSpringer Journals

Published: Jun 1, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off