The impact of somatic symptoms on depressive and anxiety symptoms among university students in central Serbia Uticaj somatskih tegoba na nastanak depresivnosti i anksioznosti kod studenata u Srbiji

Background/Aim. Depression and anxiety problems are a major public health concern due to their high prevalence rates, difficult treatment, and often chronic course. This study examined the impact of somatic symptoms on depressive and anxiety symptoms among university students in Serbia. Methods. A cross-sectional study was performed among 1,940 students using a questionnaire specially designed for this study which included presence of Somatic and Non-specific Mental Symptoms (SNMS), Beck Depression Inventory and Beck Anxiety Inventory. The presence of somatic and associated non-specific mental symptoms over the last six months served as the basis for creating a new variable called SNMS score. Results. Receiver operating characteristic (ROC) curve showed that the SNMS score might be a very good marker for the distinction of students with or without depressive symptoms (area = 0.754, p < 0.05). The threshold value was 8.50 (sensitivity 67.6%, specificity 69.4%). Binary logistic regression showed that Odds ratio was 1.052 [95% confidence interval (CI) 1.045–1.059], which means that an increase in the value of the SNMS score by 1 increases the risk of depressive symptoms by 5.2%. ROC curve showed that the SNMS score might be an excellent marker for the distinction of students with or without anxiety symptoms (area = 0.800, p < 0.05). Limit value (cut-off) was 7.50 (sensitivity 74.2%, specificity 71.6%). Binary logistic regression showed that odds ratio was 1.056 (95% CI 1.049–1.064), which means that increasing the value of SNMS score by 1 increases the risk of anxiety symptoms by 5.6%. Conclusion. The SNMS score might be a state marker for the screening and distinction of students with depressive symptoms, and excellent state marker for screening and making distinction between students with anxiety symptoms and the students who do not have these symptoms.


Introduction
Mental health problems are a serious public health problem concern due to their high prevalence rates, difficulties during the treatment, chronic course 1 and at the same time a source of immense human suffering 2 .
Mental disorders account for a large proportion of the disease burden in young people and many such illnesses have typically their first onset during the university age 3,4 .
Lack of recognition of mental health problems among young people, becomes important challenge for medical professionals in the struggle for the disorder identification and consequent treatment, as early as possible 5 .
Depression and anxiety increasingly emphasize the importance of somatic, comorbid symptoms, which increase both medical and economic burden to the society 6 .For example, shoulder or arm pain/discomfort is common among college students.Researchers reported that Nigerian undergraduates had a high prevalence of musculoskeletal pain, where shoulder pain was most common.In Australia, musculoskeletal disorders were a widespread problem for university students; and in the United States, college students reported low back pain, and the majority had musculoskeletal discomfort during/after computer use.Likewise, in Saudi Arabia, about the fourth to the fifth of students complained of headache (17%) and fatigue (24%) 7 .It is well documented that pain or aches in shoulders, arms and neck are prevalent in many societies and are an economic problem due to sickness absence and health-care costs 8 .
Several studies of the relationship among anxiety, depression, and somatic symptoms which included general population have already been conducted in many countries.Based on such experiences, three explanatory hypotheses about the nature of the relationships can be offered.Firstly, depression and anxiety disorders may influence the onset of physical symptoms in many ways 9 e.g.altered perception of physical sensations 10 .Secondly, somatic symptoms, or different types of physical limitations may be predictors of onset of depressive and anxiety disorders.Finally, according to the third hypothesis, numerous environmental, biological and psychological factors may independently influence the onset of both mental disorders and somatic complaints 11 .
In clinical practice, there is often so-called masked depression.It is characterized by somatic "mask", which is comprised of a wide range of somatic disorders.It is not rare that people suffering from masked depression seek help from doctors of different specialties, change healthcare facilities and most often they are diagnosed with hypochondria.In our culture, this projection of psychological problems to somatic symptoms is explained by, among other things, greater social acceptability of somatic disorders.Stigma and negative attitudes toward seeking help from a psychiatrist are still barriers to improving mental health 11 .
It is estimated that 7.3% to 11% of all patients suffering from depression suffer from masked depression, and on average, an interval between first signs of illness and referral to a psychiatrist, is 3-5 years.The most common motives for the referral are negative findings of somatic examinations and lack of therapy response to the applied somatic therapy.Patients who suffer from this disorder usually complain about headaches in the form of twinges, pinching, pain in the face and teeth and often closely resemble migraine.In addition, there are pains in the lumbar spine and surrounding spinal muscles, and the symptoms are often interpreted as spondyloarthrosis or discopathy.The occurrence of paresthesia in the region of extremities, the feeling of heaviness of extremities "as if they were filled with lead" are common; there is also the phenomenon of restless legs, shortness of breath and pain in various parts of the body.There are also vegetative symptoms which frequently appear as dizziness, feeling of emptiness in the head, chest pressure, pain in the heart, tachycardia, dryness in the mouth, anuria, dysphagia, meteorism or difficult breathing.Therefore, the masked depression is a disorder that can mimic different somatic diseases making its diagnosis and treatment difficult 12 .
The aim of this study was to examine the impact of somatic symptoms on depression and anxiety symptoms among a group of University students in Central Serbia.

Study design and participants
The research we have conducted was a cross-sectional survey of students attending University of Kragujevac, Serbia, during the 3-years period, that is, from 2012 to 2015.University of Kragujevac, with its twelve faculties, is a stateowned university in central Serbia.Six of its faculties are located in Kragujevac while others are located in five towns in central Serbia, covering the area with more than 2,500,000 inhabitants.All faculties were selected for the survey: Philology and Arts, Faculty of Hotel Management and Tourism.The students were randomly sampled from each study year from each faculty, in proportion to the size of the faculty in relation to the total number of students at the University.The sample for this survey, was randomly selected: 10.70% students, out of the total of 18,123 students.The students were sorted out from the University student database according to previously generated random order (random computer function).

Procedure
A self-administered anonymous questionnaire was used and it comprised Somatic and associated Nonspecific Mental Symptoms (SNMS), Beck Depression Inventory and Beck Anxiety Inventory (BAI).Ethical approval was obtained from the Faculty of Medical Sciences Ethical Committee.Participation was voluntary with no financial or other motivation.Informed consent was obtained and confidentiality of the responses was assured.The study was conducted in the participants' classrooms by the leading researcher (I.V.S).Those who were absent during the distribution of questionnaires were excluded.The research was completed within 2 years.

Instrument
A self-assessment questionnaire (which was assembled for this study) with detailed subdomain questions was used to determine SNMS.Symptoms of depression were evaluated through the Back Depression Inventory -revision of original instrument (BDI-IA) scale.This scale was developed in the 1960's and is one of the most widely used instruments for measuring the severity of depression, with the focus on behavioral and cognitive aspects of the disorder.It was designed to document a variety of depressive symptoms which the individual experienced over the preceding week.It consistes of 21 items, each answer being scored on a scale ranging from 0 to 3 12 .
Symptoms of anxiety were evaluated through the Back Anxiety Inventory (BAI) scale, a short list describing 21 anxiety symptoms which bothered them in the previous week.The scale consistes of 21 items, each answer being scored on a scale ranging from 0 to 3 13 .The presence of SNMS over the last six months served as the basis for creating a new variable called SNMS score.
Questions from the questionnaire, that were related to the presence of specified nine symptoms reported over the last six months (headache, abdominal pain, back pain, fatigue, despondency, irritability, nervousness, sleep problems, dizziness), were rated with five-graded responses (rarely or never -0, almost every month -1, almost every week -2, more than once a week -3, almost every day -4).Therefore, the maximum score for these symptoms was 36.Some clinical findings suggest that somatic and associated nonspecific mental symptoms like despondency, irritability, nervousness, sleep problems may dominate the clinical picture and develop prior to other symptoms and signs of depression 14 .

Statistical analysis and assessment
Data analysis was carried out using IBM SPSS (Statistical Package for the Social Sciences) software version 19.0.The qualitative variables were presented in number and percentage.The continuous variables (depression, anxiety and symptoms scores), were presented as means and standard deviation (SD).Descriptive statistics for somatic symptoms, depressive and anxiety symptoms were calculated, expressed as appropriately in frequencies, mean values and standard deviation.Student's t-test, Fisher's exact test, χ 2 test, were used to look for any existing differences between somatic symptoms, and depressive/ anxiety symptoms.All tests were 2-tailed, and the level of significance was set at p ≤ 0.05.Binary logistic regression analysis was used to determine associations between depressive, anxiety symptoms and somatic symptoms.The results were reported as odds ratios (OR) with 95% confidence intervals (CI).

Results
Out of 1,968 distributed questionnaires, a total of 1,940 (98.6%) students completed the questionnaire among which there were 34.7% of males and 65.3% of females) the mean age of the participating students was 21.04 (SD = ± 2.23) years with the range of 18-57 years.
Distribution of somatic and associated nonspecific mental symptoms of the sample are summarized in Table 1.The difference between the mean values of SNMS score between students with depressive symptoms and students without depressive symptoms was statistically significant (p < 0.05).The mean value of the SNMS score of students without depressive symptoms was 5.00 (range, 3.00 to 10.00), while the mean value of the new variable in students with depressive symptoms was 12.00 (range, 7.00 to 18.00).
Receiver operating characteristic (ROC) curve showed that the SNMS score might be a very good state marker for the distinction of students with depressive symptoms from students without it (area = 0.754, p < 0.05).The threshold value was 8.50, the sensitivity was 67.6% and specificity 69.4% (Figure 1).

Fig. 1 -Receives operating characteristic (ROC) curve of the sum of Somatic and associated Non-specific Mental Symptoms according to depressive symptoms.
Binary logistic regression showed that SNMS score might affect the occurrence of depressive symptoms.Odds ratio was 1.052 (95% CI 1.045 to 1.059), which means that an increase in the value of the SNMS score by 1 increases the risk of depressive symptoms by 5.2% over the baseline population rate.
The difference of mean values of the SNMS score between patients with anxiety symptoms and those without it was statistically significant, too (p < 0.05).The mean value of the SNMS score of the somatic and associated nonspecific mental symptoms in students without anxiety symptoms amounted to 4.00 (range, 2.00 to 8.00), whereas the mean value of the new variable in students with anxiety symptoms was 12.00 (range, 7.00 to 17.00).
ROC curve shows that the SNMS score might be valuable state marker for the distinction of students with anxiety symptoms, from students without anxiety symptoms (area = 0.800, p < 0.05).Limit value (cut-off) was 7.50, wherein sensitivity was 74.2% and specificity was 71.6% (Figure 2).Binary logistic regression showed that the SNMS score might affects the occurrence of anxiety symptoms.Odds ratio was 1.056 (95% CI 1.049 to 1.064), which means that an increase in the value of score of somatic symptoms by 1 increases the risk of anxiety symptoms by 5.6%.

Fig. 2 Receives operating characteristic (ROC) curve of sum of and associated
Mental Symptoms according to anxiety symptoms.

Discussion
Our study is one of the largest epidemiological studies regarding mental health status and somatic symptoms among university students in this region.The first objective of this study was to examine the distribution of same somatic symptoms among university students in central Serbia.
Our results show that headache and back pain represent common symptoms among university students, i.e. 22.3% of respondents had headache almost every month and 7.4% on daily basis over the last six months.The literature indicates that most common somatic symptoms among student population are headache, menstrual cramps, back pain, fatigue and sleeping problems 15 .Worldwide, 54.4 % of the adolescent population suffer from headache.A study conducted in Germany shown that over 80% of adolescents had headache symptoms in the last 6 months.One study showed that 30.3% of Chinese adolescents experienced headache more than once a week during the previous 3 months 16 .The data combined from 56 independent studies (analyzing a total number of 34,904 students) revealed that pooled migraine prevalence amounts to 16.1% 17 .The study conducted on a sample of 4,406 students in China showed that the prevalence of migraine among university students was 9.0% 18 , which was in accordance with the prevalence found in Chinese adults in the general population (8,9%) 19.Previous epidemiological studies linked anxiety and depression with migraine suggesting the presence of such comorbidity among 30%-84% people.It was proven that depression and anxiety increase the risk of migraine when they occur separately as well, however, if the depression and anxiety occur together, the risk becomes even higher 20 .
The results of this study suggest that 21.1% of students had back pain during the last six months almost every month, while 7.2% had back pain every day.The obtained results reveal a high prevalence of back pain among student population.The data from the current literature showed that back pain in student population varies between 20%-70%.In several studies, obtained values for back pain prevalence were Simić Vukomanović I, et al.Vojnosanit Pregl 2018; 75(6): 582-588.
higher than the results obtained by this study.For example, in a study conducted in China on the sample of 2,849 adolescents, the prevalence amounts to 41.4%, among Brazilian school children to 55.5% (n = 802) and 51.3% among 546 Danish adolescents 21.The reason may be that heavy academic pressure induces an increase in risk factors for somatic symptoms [elevated levels of mental stress, lack of sleep and changes in adolescent living habits (smoking, alcohol and coffee consumption)] which can together influence the occurrence of mental disorders.Relationship between chronic pain and depression or anxiety attract particular attention in the field.About 35% of patients with chronic pain meet the criteria for an anxiety disorder, while 30% -60% of patients with chronic pain suffer from depression 22 .It was shown that chronic pain in childhood and adolescence heightens the risk of depressive and anxiety disorders in adulthood in specific samples 23 .
The main objective of this study was to examine the impact of somatic symptoms on depressive and anxiety symptoms among University students in central Serbia.
In our study, the mean value of the sum score of somatic and associated non-specific mental symptoms among student population without depressive symptoms was 5.00 (range, 3.00 to 10.00), while the mean value of the new variable in students with depressive symptoms was 12.00 (range, 7.00 to 18.00).The mean value of the sum score of the somatic and associated non-specific mental symptoms in students without anxiety symptoms amounts to 4.00 (range, 2.00 to 8.00), whereas the mean value of the new variable in students with anxiety symptoms is 12.00 (range, 7.00 to 17.00).
Our findings are similar to those of the study conducted at two universities in Mexico, which included 506 psychology showing students that the intensity of physical manifestations is more positively correlated with scores derived from two rating scale (Beck Depression Inventory and Social Anxiety Scale for Adolescents) 15 .The indicated that pain-related complaints presented to general practitioners including undefined somatic symptoms (muscle pain, headache, and abdominal pain) are from 2.5 to 10 times more frequent in people with panic disorder, generalized anxiety disorder and major depression than in general population 15 .Patients who felt some kind of somatic pain also had worse mental functioning and higher scores on rating scales for depression, social anxiety and post-traumatic stress disorder 24 .It was suggested in current literature that somatic symptoms directly relate to difficulties in everyday functioning, disability, increased use of psychoactive substances, impaired quality of life as well as reduced use of health care services 6 .The research in the United States showed that 87% of students were feeling mild fatigue while studying.Students who were feeling moderate and severe fatigue scored higher on the BDI scale for the assessment of depression than those who were feeling mild fatigue.Students with severe fatigue also got higher scores on the scale for anxiety assessment 25 .Other studies also reported that depression disorders were in a stronger correlation with somatic problems of the musculoskeletal system 26 while cardiopulmonary symptoms were more associated with anxiety disorders 9,27 .
Our results suggest that the SNMS score might be a very good marker for the distinction of students with depressive symptoms from students without it.In other words, the sum score of somatic and associated non-specific mental symptoms might affect the occurrence of depressive symptoms, which means that an increase in the value of the sum score of somatic and associated non-specific mental symptoms by 1, increases the risk of depressive symptoms by 5.2% over baseline population rate.Several studies have reported the association between depression and pain (e.g.severity, frequency, duration and number of symptoms).Patients with some pain symptoms (e.g.back pain, headache, abdominal pain, chest pain and facial pain) are 3 to 5 times more likely to be depressed than patients without pain, and pain symptoms are associated with at least twice increased risk for coexisting depression.Additionally, one populationbased study revealed that subjects with chronic pain (defined as pain felt during most days in at least a month) were 3 times more likely to meet depression criteria than those without chronic pain.The association between depression and pain became stronger as severity of either condition increases.Additionally, more frequent pain episodes and longer duration of pain were associated with depression.An international study showed that patients with pain lasting longer than 6 months were more than four times likely to have a depression disorder than those without chronic pain 28 .
Chronic pain and depression are considered to be the harshest expressions of human suffering.Clinical experience suggests that physical pain and depression are often intertwined, and that they are in mutual interaction 22 .Some consider the pain one of the symptoms of depression, which could explain the effectiveness of antidepressants in the suppression of painful condition such as neuropathic pain 29 .
Our results indicate that the sum score of somatic symptoms might be a valuable marker for the distinction of students with anxiety symptoms, i.e. the sum score of somatic symptoms might affect the occurrence of anxiety symptoms which means that an increase in the value of the score of somatic symptoms by 1 increases the risk of anxiety by 5.6%.
Data from other studies suggest that anxiety disorders may be present in up to 60% of patients with chronic pain, i.e. more severe chronic pain was associated with more severe anxiety symptoms 30 .Anxiety is associated with pain, as a symptom, regardless of its anatomic localization 19 .The research, conducted by World Health Organization (WHO), which involved 5,447 students in 15 research centers, located in America, Europe, Asia and Africa, found that people with chronic pain were four times more likely to suffer from anxiety disorders and depression 31 .
The symptom of chronic pain can be a potential risk factor for recurrent depressive and anxiety disorders.Neuroimaging studies showed overlapping of neural networks of emotions and pain, particularly in prefrontal cortex 32 .Pain could affect regulation of the hypothalamic-pituitary-adrenal axis (increasing cortisol levels) as well as the activity of the autonomic nervous system (increased sympathetic nervous system activity and decreased parasympathetic nervous system activity) which may lead to the onset of new depressive Simić Vukomanović I, et al.Vojnosanit Pregl 2018; 75(6): 582-588.
episodes 33 .On the other side, some studies failed to find association between pain symptoms and recurrence of anxiety 32 .
Depressive and anxiety disorders, in comorbidity with physical manifestations, are associated with an increased number of suicides, decreased quality of life and significant economic burden on the health care system 24 .
The limitation of our study is the cross-sectional design, which does not permit inferences about possible causal relations between the somatic variables and depressive and anxiety symptoms.It was not possible to assess the test-retest reliability of BDI/BAI in this sample as the survey was anonymous.Another limitation was the self-reported nature of this study.Finally, the sample represents a group of students in just one university of Serbia, which may limit generalizations of the results on the other universities.
Therefore, the careful rater could rather easily screen a student with depression and anxiety with existing and validated tools if additional attention was paid to the domains related to somatic symptoms.Future studies are needed for developing the novel, valid diagnostic rating scales highly specific for detection of depressive and anxiety disorders in students suffering predominantly from non-specific physical complains without obvious causes.
The results of this study illustrated that student population talks about somatic symptoms rather than about mental problems and highlight the importance of recognizing those symptoms and connection between them and depressive and anxiety disorders.

Conclusion
The SNMS score might be a state marker for the screening and distinction of students with depressive symptoms and excellent state marker for screening and making a distinction between the students with anxiety symptoms and the students who do not have these symptoms.
Somatic symptoms could be a predictor of depression and anxiety among students, and early screening could improve their prevention and early diagnose and treatment.
Faculty of Agronomy, Faculty of Economics, Faculty of Engineering, Faculty of Mechanical and Civil Engineering, Faculty of Medical Sciences, Faculty of Education, Faculty of Law, Faculty of Natural Sciences and Mathematics, Faculty of Technical Sciences, Teachers Training Faculty, Faculty of Simić Vukomanović I, et al.Vojnosanit Pregl 2018; 75(6): 582-588.