|ORIGINAL RESEARCH ARTICLE
|Year : 2018 | Volume
| Issue : 2 | Page : 80-86
A cross-sectional examination of the relationship between approaches to learning and perceived stress among medical students in Malaysia
Alireza Behzadnia, Daniel R Smith, Michaela L Goodson
Newcastle University Medicine Malaysia, Iskandar Puteri, Johor, Malaysia
|Date of Web Publication||30-Nov-2018|
Newcastle University Medicine Malaysia, No. 1 Jalan Sarjana, Kota Ilmu, [email protected], Iskandar Puteri, 79200 Johor
Source of Support: None, Conflict of Interest: None
Background: Learning approaches have been proposed to affect the experience of psychological stress among tertiary students in recent years. This relationship becomes important in stressful environments such as medical schools. However, the relationship between stress and learning approaches is not well understood, and often studies done cannot be generalized due to different sociocultural differences. In particular, no study in Malaysia has looked at learning approaches among medical students. Aims: To address this gap, we examined the relationship between perceived stress and learning approaches by considering sources of stress. Methodology: The Perceived Stress Scale (PSS-10), Medical Student Stressor Questionnaire, and the Revised Two-Factor Study Process Questionnaire were answered by the preclinical and final-year students studying MBBS in a Malaysian campus of British University. Results: Deep learning was positively and surface learning negatively associated with perception of coping with stress. In this study, neither approaches were associated with psychological stress as opposed to previous reports. We found surface learners to report higher level of stress associated with social stressors. We found students' self-perception of feeling incompetent and feeling they need to do well to be significant sources of stress. Discussion: Deep learning promotes psychological resilience. This is of paramount importance in learning environments where stress is highly prevalent such as medical school. Promotion of deep learning among medical students is required at earlier stages as they tend to solidify their approach through their university years and carry that approach beyond school into their workplace.
Keywords: Coping, learning approaches, medical students, stress, stressors
|How to cite this article:|
Behzadnia A, Smith DR, Goodson ML. A cross-sectional examination of the relationship between approaches to learning and perceived stress among medical students in Malaysia. Educ Health 2018;31:80-6
|How to cite this URL:|
Behzadnia A, Smith DR, Goodson ML. A cross-sectional examination of the relationship between approaches to learning and perceived stress among medical students in Malaysia. Educ Health [serial online] 2018 [cited 2022 Jan 16];31:80-6. Available from: https://www.educationforhealth.net/text.asp?2018/31/2/80/246750
| Background|| |
The way we learning can be broadly thought as either “deep” or “surface” based on the Approaches to Learning theoretical framework. Svensson and Marton in 1976 conceptualized this framework to reflect the level of processing and engagement that goes into learning, where comprehending subject matter beyond memorization is considered to be “deep,” while memorizing (almost always without further analysis) is the “surface” approach to learning. Adoption of either approaches can be influenced by different factors categorized as contextual (e.g., teaching and assessment methods), perceived contextual (e.g., workload and assessments), and student factors (e.g., age and gender). This means we may deem either approaches fit for the task based on the context that learning is happening.
We can think of psychological stress as a perceived contextual factor influencing learning approaches; however, perception of stress is also fluid and context dependent. This relationship becomes of particular interest in high-pressure learning environments such as medical schools. Medical students have reported to experience a significantly higher level of stress compared to the age-matched population.,, Self-reported sources of stress among medical students can be divided into academic and nonacademic; academic sources such as workload and assessments (including mode of assessment) and nonacademic sources such as competition, self-doubt, and financial problems.,,,
The apparent similarities between stressors and learning approaches influencers among medical students raise the question: Is there an interplay between psychological stress, sources of stress and learning approaches, amongst medical students?
Few studies in the recent years have attempted to answer this question. In a 5-year long longitudinal study of medical students, Sandover et al. (2015) found student's response to stressors impacts their approaches to learning. Similarly, Chen et al. (2015) report that surface learning is linked to a higher level of perceived stress in medical students. Few studies have explored this, and the majority of reports are from North America, making the results difficult to generalize.
To address this research gap, we decided to carry out a cross-sectional study of a cohort of medical students from our own institution which offers a British MBBS curriculum in Malaysia. We hypothesized that significant stressors are associated with surface learning approach. Hence, the focus of our study was on perceived stress, sources of stress, and learning approaches among medical students. Specifically, our research questions were as follows:
- What are the predominant sources of stress in our student cohort?
- How are the learning approaches related to the perceived distress and coping?
- How do the sources of stress associated with the adoption of a particular learning approach?
| Methodology|| |
A quantitative, cross-sectional study was carried out using a survey among preclinical (year 1 and 2) and final-year (year 5) medical students at the Newcastle University Medicine Malaysia in 2016. The questionnaire comprised four sections, namely, Demographics, Perceived stress, Sources of stress, and Learning approaches. Each section compromised a validated Likert scale-based questionnaire. Ethical approval for this study was granted by the University Ethics Committee.
Basic demographic questions gathered information on gender, date of birth, year of study, relationship status, ethnicity, and history of mental illness.
The Perceived Stress Scale-10 item (PSS-10) was used to assess perceived stress. This is validated psychometric instrument widely used in undergraduate studies that measures two factors: “perceived distress” and “perceived coping.”
Self-reported severity of stressors was measured using the Medical Students Stressor Questionnaire (MSSQ). MSSQ was designed and validated in Malaysia based on the Asian medical students' reports and studies in 2011. It contains 20 items answered on a four-point Likert scale, grouped to form six principal domains of stressors: (1) academic, (2) interpersonal and intrapersonal, (3) teaching and learning, (4) social, (5) drive and desire, and (6) group activities. The mean score of each domain is then stratified: causing “mild stress” (M = 0–1.00), “moderate stress” (M = 1.01–2.00), “high stress” (M = 2.01–3.00), or “severe stress” (M = 3.01–4.00).
Biggs' Study Process Questionnaire (SPQ) is a well-known instrument used to assess learning approaches at the tertiary level.,,, The latest revision of the SPQ, a Revised Two Factor SPQ (R-SPQ-2F), was used to assess learning approaches. Either approach is measured through 10 items that are answered on a five-point Likert scale.
R-SPQ-2F produces two separate scores for deep (DA) and surface approach (SA). DA and SA are further divided by motive and strategy, thus establishing the categories as deep motive (DM), deep strategy, surface motive, and surface strategy.
The R Environment for Statistical Computing was used for all statistical analyses. To check for the validity of questionnaires, Cronbach's alpha was used to check for internal consistency and explanatory factor analysis was used specifically for the MSSQ due to the limited evidence of its validity in the literature. Outliers were identified using Mahalanobis distance (P < 0.001) and Bonferroni outlier test (P < 0.05). Regression models were used to answer our questions with Type II ANOVA test in tests involving multiple predictors. Regression analyses were each checked for homogeneity of variance (Levene's test), normality of residuals (Shapiro–Wilk test), and multicollinearity (variance inflation factor). Statistical significance was set at the conventional 5% threshold (α = 0.05). Effect sizes were estimated with 95% confidence intervals (CIs), however, partial eta squared (η2) was estimated at 90% CI (1-2α).
| Results|| |
All the questionnaires had an acceptable internal consistency with Cronbach's alpha values of >0.70 (15): PSS with a raw Cronbach's alpha of 0.85 (95% CI = 0.82–0.88), MSSQ assessed in each domain yielded Cronbach's alpha values ranging from 0.65 (95% CI = 0.57–0.73) to 0.94 (95% CI = 0.92–0.95), and R-SPQ-2F with a Cronbach's alpha of 0.76 (95% CI = 0.71–0.81) for the deep and 0.83 (95% CI = 0.80–0.86) for the surface approach.
PSS-10 and the SPQ questionnaires have been subjected to rigorous assessment; however, MSSQ-20 is a new instrument, developed in 2011. To check the validity of MSSQ-20 factor structure, we carried out an exploratory factor analysis using principle-axis factor extraction. The principal component analysis suggested five factors with an eigenvalue >1 (K-1 rule). The five factors were confirmed using parallel analysis and maximum-likelihood factor analysis with varimax rotation, thus confirming the presence of five factors as opposed to the suggested six (Chi-squared: χ2 (100, N = 210) = 127.76, P < 0.05).
The factor loading and the explained variance of the factors were Factor 1 – Interpersonal-related stressors (loading: 3.533, variance: 0.177); academic-related stressors (ARS) with two items from the group activity-related stressors (GRS) (which were “feeling of incompetent ” and “need to do well”) (3.473, 0.174); social-related stressors and ” participation in class presentation” (the only other item from GRS) (1.849, 0.092); teaching- and learning-related stressors (1.814, 0.091); and Drive- and desire-related stressors (1.550, 0.078). Accordingly, we used the five-domain model in our further analyses.
The overall response rate was 85.60% (214 out of 250). Eighty-seven (40.65%) students were in the 1st year, 96 (44.85%) in the 2nd year, and 31 (14.50%) in the final year of their studies. The mean age of participants was 20.35 years (standard deviation [SD] = 0.94) with a range of 18–24. Majority were female (n = 130; 60.80% vs. n = 74 [34.60%]). Two stated an “unspecified gender” and the remaining (n = 8; 3.70%) did not disclose their gender. Malaysian students came from different ethnic background: Chinese (n = 96; 44.9%), Malay (n = 77; 36.0%), and Indian (n = 15; 7.0%). The rest of the participants were international students (n = 9; 4.2%).
Perceived Stress Scale
The mean PSS score was 19.51 (SD = 5.92; total possible score = 40). More than half of the students (n = 121; 57.60%) often felt “nervous and stressed” in the last month out of which, 58.70% (n = 71) “never” or “only sometimes” felt confident during the same period. However, 52.40% (n = 110) of all students reported to have often felt confident. We found no significant demographic factor associated with the total PSS score: gender (F (1, 190) = 2.937, P = 0.088), year (F (1, 190) = 1.027, P = 0.360), and age (F (1, 190) = 0.404, P = 0.526).
Medical Student Stressor Questionnaire
[Table 1] summarizes the frequency of self-reported levels of stress caused by each domain. ARS (M = 2.61, SD = 0.75) were reported to be causing high or severe level of stress (M > 2.00). Frequently academic stressors reported were”tests and examinations” (72.30%), “large amount of content to be learnt” (62.90%), and “lack of time to review” (59.50%).
|Table 1: Frequency of self-reported levels of stress based on different stressor domain|
Click here to view
Age and the year of study were linked to drive- and desire-related stressors: age (F (1, 190) =13.985, P < 0.001, η2 = 0.091, 90% CI = 0.022–0.132]) and year of study (F (2, 190) = 12.020, P < 0.001, η2 = 0.112, 90% CI = 0.047–0.180), both with a small effect size.
Study Process Questionnaire
Deep learning score was significantly higher than surface learning (DA: M = 31.70, SD = 6.13 vs. SA: M = 26.00, SD = 7.18) and paired t-test (t (209) = 8.825, P < 0.001, d = 0.853, 95% CI = 0.655–1.052). Male students scored higher than females in both approaches: SA (F (1, 190) = 6.426, P = 0.012, η2 = 0.033, 90% CI = 0.004–0.084) and DA (F (1, 190) = 5.344, P = 0.0219, η2 = 0.027, 90% CI = 0.002–0.076). Final-year medical students had a higher deep learning approach score than the preclinical years (year 1 and 2) (F (2, 190) = 6.108, P < 0.003, η2 = 0.060, 90% CI = 0.013-0.116) [Figure 1].
|Figure 1: Mean scores with standard error of the mean bars. Left plot: Learning approaches and gender differences. Right: Deep learning approach score in different study years|
Click here to view
What are the predominant sources of stress in our student cohort?
We found the new ARS domain to be significantly associated with perceived distress and perceived coping scores. To answer this question, we carried out a general linear regression analysis to determine the association between the new MSSQ domains on the two constructs of PSS-10. Perceived stress regression revealed: ARS (F (3, 194) =16.482, P < 0.001, η2 = 0.203, 90% CI = 0.116-0.274) and perceived coping: ARS (F (3, 194) = 4.366, P = 0.005, η2 = 0.063, 90% CI = 0.011–0.114). [Figure 2] shows the frequency of students who reported ARS items causing “high” or “severe level of stress.”
|Figure 2: Number of students stating “high level of stress” or “severe stress” caused by the stressors compromising the newly formed academic-related stressors domain|
Click here to view
To identify specific stressors (i.e., MSSQ item), we assessed the significance of the ARS items in a regression model with items fitted as categorical explanatory variables. We found “tests and examinations” and “feeling of incompetence” to be associated with perceived distress: regression model (F (28, 181) = 4.553, P < 0.001, R2 = 0.413, 95% CI = 0.327–0.510, R2adjusted = 0.323); “tests and examinations” (F (4, 181) = 2.660, P = 0.034, η2 = 0.056, 90% CI = 0.002–0.100); and “feeling of incompetence” (F (4, 181) = 6.389, P < 0.001, η2 = 0.123, 90% CI = 0.045–0.185). [Figure 3] shows the linear relationship between “feeling of incompetence” and the perceived distress and coping scores.
|Figure 3: Perceived coping (max score = 16) and perceived distress (max score = 24) scores at different levels of perceived incompetency|
Click here to view
”Need to do well” and “feeling of incompetence” were negatively associated with perceived coping: “need to do well” (F (4, 181) = 2.560, P = 0.040, η2 = 0.053, 90% CI = 0.001–0.100) and “need to do well” (F (4, 181) = 3.515, P = 0.008, η2 = 0.072, 90% CI = 0.011–0.122).
Are the learning approaches related to the perceived distress and coping?
We found deep approach (DA) to positively and surface approach (SA) to negatively be related to perceived coping [Table 2]. Two separate multiple linear regression analyses were carried out to with learning approaches fitted as continuous explanatory variables with the PSS constructs: “perceived distress” and “perceived coping” which yielded perceived coping model to be significant only: perceived coping (F (2, 207) = 5.566, P = 0.004, R2 = 0.051, 95% Cl = 0.005–0.098, R2adjusted = 0.042) and perceived distress (F (2, 207) = 1.743, P = 0.178) [Figure 4]. In assessing the subclasses, perceived coping associated with DM approach (F (1, 205) = 4.598, P = 0.033, η2 = 0.022, 90% CI = 0.001–0.065).
|Table 2: ANOVA table of the general linear model for the total perceived distress and perceived coping and the learning approaches|
Click here to view
|Figure 4: Effect plot of the learning approaches on perceived coping score. DA=Deep learning approach, SA=Surface learning approach|
Click here to view
Are the stressors associated with adoption of a particular learning approach?
We found social-related stressors (SRS) and teaching- and learning-related stressors (TLRS) to be associated with surface learning approach: SRS (F (3, 194) = 4.660, P = 0.003, η2 = 0.067, 90% CI = 0.014–0.012) and TLRS (F (3, 194) = 4.888, P = 0.002, η2 = 0.070, 90% CI = 0.015–0.123) although it should be noted that the cumulative effect size of the associations is rather small (η2 = 0.140). Further, we repeated the analysis with the MSSQ items in these two domains (similar to the first step): no significant stressor (MSSQ item) was identified. Repeated models with the surface learning subclasses (motive vs. strategy) did not yield any significant results either.
| Discussion|| |
In this study, we aimed to answer the question ”is there an interplay between stress, stressors, and learning?”. Our results have shown that there are in fact a relationships between these factors however this interplay is more complex than a three-way cause and effect relationship. In short, deep learning was associated with better coping and less stress, while the opposite was true for surface learning. In addition, our hypothesis that stressors encourage surface learning was somewhat true. We found social stressors (including teaching and the learning environment stressors) to be associated with surface learning.
It has been suggested that surface learning is linked to psychological stress;, however, we did not find a significant link between the two – despite its negative relationship with perceived coping. This raises the question that whether the proposed association is merely due to surface learners perceiving themselves to be unable to cope with the stress they experience, regardless of the severity of stress.
In this cohort, perception of coping was related to deep learning, specifically, the “DM” approach; supporting previous findings of Sandover et al. and Baeten et al., Medical students' coping strategy has an impact on their learning approach approaches to learning. In addition, deep learners experience positive emotions such as pride, confidence, and hope. Positive self-perceptions help individuals to cope with stress and bounce back from setback – as the folk theory goes: “staying positive makes you feel better.” Tugade et al. put this theory to test and found it to be true: positive emotions were associated with psychological resilience.
Behavioral psychology may further explain the relationship between coping and deep learning. Fuente et al. found deep learners to be more likely to have a competitiveness-overwork behavioral trait which has a buffering effect against stress.
Self-reported social-, teaching-, and learning-related stressors were associated with surface learning among participants; however, we suggest this finding to be interpreted contextually. Our study was conducted not long before a major examination; in fact, we excluded year 3 and four students due to their different examination schedule as well as the nature of their off-campus studies. Given the motif behind surface learning, to memorize and learn only the essentials to pass an examination, adoption of this approach seems rather appropriate. Furthermore, this could partly be due to students' misinterpretation of “understanding” of their studies as “factual recall” when it comes to assessments. Previous reports also favor surface learners in regards to examination performance; surface learning results in better outcome, especially in multiple choice question examiantions.,
Surface learning is also associated with greater hours of independent studying. This could as well explain the social-related stress (e.g., participation in class) association with surface learning in our study.
Problem-based learning, progress testing, and structure of the curriculum influence how students deal with learning – as these approaches promote deep approach.,,, Our students are enrolled in a spiral curriculum (a form of problem-based learning) with progress testing as the mode of assessment – the influence of these environmental factors (contextual) is reflected on the higher deep learning scores than surface learning in our cohort.
Finally, just as we speculated that the cross-cultural nature, the course (delivery of a British curriculum in the Malaysian setting) would impact students' experience – our findings were different from previous reports. A review of Malaysian studies on stress and stressors among tertiary students found interpersonal, emotional disturbances, and social relationship with parents, siblings, lecturers, and peers to be prevalent sources of stress; this was not true in our study. Contrary to their review and previous reports, we found female and male students to report the same level of perceived stress.,,
| Conclusion|| |
In this study we addressed learning approaches among medical students by considering contextual, perceived contextual, and student factors. Based on our findings, we suggest medical educators to consider perceived contextual factors reported by students such as feeling incompetent, stressed, or inability to cope with what they face, as important indications that require appropriate attention. Such factors not only influence the psychological health of the student but also how they approach their studies and subsequently their medical training.
We also suggest to other educators to consider students' approach to learning as an important factor which may impact their psychological resilience. We believe encouraging students to learn about learning approaches and their implications as well as being aware of their own particular approach may positively impact their academic performance as well as helping them to cope better in face of stress.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Baeten M, Kyndt E, Struyven K, Dochy F. Using student-centred learning environments to stimulate deep approaches to learning: Factors encouraging or discouraging their effectiveness. Educ Res Rev 2010;5:243-60. Available from: http//www.dx.doi.org/10.1016/j.edurev. 2010.06.001
. [Last accessed on 2016 Oct 25].
Dyrbye LN, Thomas MR, Shanafelt TD. Systematic review of depression, anxiety, and other indicators of psychological distress among U.S. And canadian medical students. Acad Med 2006;81:354-73.
Mosley TH Jr., Perrin SG, Neral SM, Dubbert PM, Grothues CA, Pinto BM, et al.
Stress, coping, and well-being among third-year medical students. Acad Med 1994;69:765-7.
Schwenk TL, Davis L, Wimsatt LA. Depression, stigma, and suicidal ideation in medical students. JAMA 2010;304:1181-90.
Dyrbye LN, Shanafelt TD. Commentary: Medical student distress: A call to action. Acad Med 2011;86:801-3.
Dyrbye LN, Thomas MR, Shanafelt TD. Medical student distress: Causes, consequences, and proposed solutions. Mayo Clin Proc 2005;80:1613-22.
Tartas M, Walkiewicz M, Majkowicz M, Budzinski W. Psychological factors determining success in a medical career: A 10-year longitudinal study. Med Teach 2011;33:e163-72.
Walkiewicz M, Tartas M, Majkowicz M, Budzinski W. Academic achievement, depression and anxiety during medical education predict the styles of success in a medical career: A 10-year longitudinal study. Med Teach 2012;34:e611-9.
Chen Y, Henning M, Yielder J, Jones R, Wearn A, Weller J. Progress testing in the medical curriculum: Students' approaches to learning and perceived stress. BMC Med Educ 2015;15:147. Available from: http//www.biomedcentral.com/1472-6920/15/147
. [Last accessed on 2016 Jun 20].
Youssef FF. Medical student stress, burnout and depression in Trinidad and Tobago. Acad Psychiatry 2016;40:69-75.
Lee EH. Review of the psychometric evidence of the perceived stress scale. Asian Nurs Res (Korean Soc Nurs Sci) 2012;6:121-7.
Tavakol M, Dennick R. Making sense of cronbach's alpha. Int J Med Educ 2011;2:53-5.
Steiger JH. Beyond the F test: Effect size confidence intervals and tests of close fit in the analysis of variance and contrast analysis. Psychol Methods 2004;9:164-82.
Tooth D, Tonge K, McManus IC. Anxiety and study methods in preclinical students: Causal relation to examination performance. Med Educ 1989;23:416-21.
Sandover S, Jonas-Dwyer D, Marr T. Graduate entry and undergraduate medical students' study approaches, stress levels and ways of coping: A five year longitudinal study. BMC Med Educ 2015;15:5.
Tugade MM, Fredrickson BL, Barrett LF. Psychological resilience and positive emotional granularity: Examining the benefits of positive emotions on coping and health. J Pers 2004;72:1161-90.
Fuente JD, Martínez-Vicente JM, Salmerón JL, Vera MM, Cardelle-Elawar M. Action-emotion style, learning approach and coping strategies, in undergraduate university students. Anales de Psicología/Annals of Psychology, 2016;32:457-65. ISSN 1695-2294.
Segers M, Nijhuis J, Gijselaers W. Redesigning a learning and assessment environment: The influence on students' perceptions of assessment demands and their learning strategies. Stud Educ Eval 2006;32:223-42.
Weller JM, Henning M, Civil N, Lavery L, Boyd MJ, Jolly B, et al.
Approaches to learning for the ANZCA final examination and validation of the revised study process questionnaire in specialist medical training. Anaesth Intensive Care 2013;41:631-40.
Kember D, Ng S, Tse H, Wong ET, Pomfret M. An examination of the interrelationships between workload, study time, learning approaches and academic outcomes. Stud High Educ 1996;21:347-58.
Bamuhair SS, Al Farhan AI, Althubaiti A, Agha S, Rahman SU, Ibrahim NO. Sources of stress and coping strategies among undergraduate medical students enrolled in a problem-based learning curriculum. Biomed Educ 2015;2015:1-8. Available from: http//www.hindawi.com/journals/jbe/2015/575139/
. [Last accessed on 2016 Jun 24].
Salam A, Yousuf R, Muhammad S, Bakar A, Haque M. Stress among medical students in Malaysia: A systematic review of literatures. Int Med J 2013;20:649-55.
Cohen MJ, Kay A, Youakim JM, Balaicuis JM. Identity transformation in medical students. Am J Psychoanal 2009;69:43-52.
Holmes CL, Harris IB, Schwartz AJ, Regehr G. Harnessing the hidden curriculum: A four-step approach to developing and reinforcing reflective competencies in medical clinical clerkship. Adv Health Sci Educ Theory Pract 2015;20:1355-70.
Al-Dubai SA, Al-Naggar RA, Alshagga MA, Rampal KG. Stress and coping strategies of students in a medical faculty in Malaysia. Malays J Med Sci 2011;18:57-64.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2]
|This article has been cited by|
||Trajectories of learning approaches during a full medical curriculum: impact on clinical learning outcomes
| ||Giovanni Piumatti,Sissel Guttormsen,Barbara Zurbuchen,Milena Abbiati,Margaret W. Gerbase,Anne Baroffio |
| ||BMC Medical Education. 2021; 21(1) |
|[Pubmed] | [DOI]|
||Abordagens de aprendizado e sua correlação com ambiente educacional e características individuais em escola médica
| ||Giulia Zanata Rossi,João Marcos da Silva Fischer,Sheyla Ribeiro Rocha,Gabriel Avila Casalecchi,Lucimar da Silva Retto de Avó,Carla Maria Ramos Germano |
| ||Revista Brasileira de Educação Médica. 2021; 45(3) |
|[Pubmed] | [DOI]|
||Perceived Stress and Associated Factors in Russian Medical and Dental Students: A Cross-Sectional Study in North-West Russia
| ||Sergei N. Drachev,Lina Stangvaltaite-Mouhat,Napat Limchaichana Bolstad,Jan-Are K. Johnsen,Tatiana N. Yushmanova,Tordis A. Trovik |
| ||International Journal of Environmental Research and Public Health. 2020; 17(15): 5390 |
|[Pubmed] | [DOI]|