Musculoskeletal symptoms and pain-response behaviors among university students in Palestine: a cross-sectional study

Abstract

Background

Musculoskeletal (MSK) symptoms are a major contributor to disability worldwide, and university students may be increasingly affected because of prolonged sedentary behavior, intensive digital-device use, and sustained academic workloads. In Palestine, these risks may be compounded by psychosocial stressors and constrained healthcare access. This study assessed the prevalence of MSK symptoms and hand discomfort and examined associated behavioral, ergonomic, and device-use factors among Palestinian university students.

Methods

A cross-sectional study was conducted among 1,052 university students in the West Bank, Palestine, between February and April 2026. MSK symptoms were assessed using the Nordic Musculoskeletal Questionnaire, and hand discomfort was evaluated using the Cornell Hand Discomfort Questionnaire. Multivariable logistic and ordinal regression analyses were performed. Additional model diagnostics, false-discovery-rate correction, sparse-cell checks, and sex-weighted sensitivity analyses were conducted to assess model stability and the influence of the female-predominant sample.

Results

Twelve-month prevalence was highest for neck (82.4%), shoulders (75.0%), lower back (73.4%), and upper back (65.8%) symptoms. Wrist/hand symptoms affected 49.7% of participants. Moderate pain was the most commonly reported severity level (56.3%). Female participants reported higher MSK symptom prevalence across most body regions. In ordinal regression, younger age, male sex, and shorter uninterrupted study sessions were associated with lower odds of higher pain severity. Continuing study activity without behavioral modification after pain onset was associated with higher odds of several regional symptoms, including upper back symptoms (aOR 9.21, 95% CI 2.91–29.22), although estimates involving pain-response categories were interpreted cautiously because of sparse cells and wide confidence intervals. In hand-pain models, preventive hand/neck stretching was associated with higher odds of hand pain, likely reflecting symptom-driven behavior rather than a harmful effect of stretching. Sex-weighted sensitivity analyses showed that the direction of the main adjusted associations was generally unchanged, although several marginal associations were attenuated after weighting.

Conclusions

MSK symptoms were highly prevalent in this convenience sample of Palestinian university students and were associated with behavioral, ergonomic, and device-use factors. Because of the cross-sectional design and non-probability sampling, causal inference and population-level generalizability are limited. These findings may help inform university-based ergonomic education and behavioral strategies, while prospective and more representative studies are needed to clarify temporality and generalizability.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12889-026-28713-2.

Introduction

Musculoskeletal (MSK) disorders are a leading cause of disability worldwide, accounting for approximately 149 million years lived with disability (YLDs) in 2019 and representing the largest contributor to non-fatal health burden in the Global Burden of Disease (GBD) study [1]. Among MSK conditions, neck pain, low back pain, and upper extremity disorders contribute the greatest share of global morbidity [1, 2]. Although MSK disorders were historically considered conditions of middle-aged and older working populations, recent evidence indicates a rising burden among young adults, particularly university students [3, 4]. This shift parallels the global expansion of digital learning technologies and increased reliance on screen-based academic activity [5].

University students are exposed to prolonged sedentary academic behaviors that increase cumulative biomechanical loading on the musculoskeletal system [3, 6]. Studies across Asia, Europe, the Middle East, and North Africa report 12-month MSK symptom prevalence exceeding 50% among university students, with the neck, lower back, and shoulders most commonly affected [3, 7, 8]. A systematic review and meta-analysis identified prolonged screen time, poor posture, emotional stress, and sustained forward head flexion as key risk factors for neck pain in this population [9].

Repetitive strain injury (RSI) refers to a group of cumulative trauma conditions affecting muscles, tendons, nerves, and connective tissues as a result of repeated mechanical loading, sustained non-neutral postures, and insufficient recovery time [10]. Among university students, the widespread use of smartphones, laptops, and tablets for academic and communication purposes has introduced important ergonomic exposures. Sustained downward gaze during smartphone use increases cervical spine loading in a posture-dependent manner, with biomechanical models estimating approximately 18 kg of force at 30° cervical flexion, 22 kg at 45°, and 27 kg at 60° [11]. Prolonged exposure to such loading patterns provides a plausible explanation for the high burden of neck and shoulder symptoms reported in student populations. In parallel, repetitive smartphone interaction requires sustained wrist positioning and repeated intrinsic hand muscle activity, which may contribute to upper-extremity discomfort and RSI-related symptoms [10, 12]. Despite this clinical relevance, hand and wrist symptoms remain less frequently examined in university-student musculoskeletal research, particularly in low-resource settings [13].

Beyond exposure duration and posture, students’ behavioral responses to pain during academic activities may also be relevant. Many studies have focused on ergonomic exposures such as device use, sitting posture, and screen level, whereas fewer have examined what students do when pain occurs during studying, such as continuing without modification, taking a rest, stretching, or stopping study activity. These behavioral responses may reflect pain coping, symptom severity, or delayed self-management, and their association with musculoskeletal symptoms remains insufficiently explored in student populations.

Evidence from the Eastern Mediterranean region, and particularly Palestine, remains limited. Palestinian university students are exposed to global behavioral risk factors, including intensive digital-device use and prolonged sedentary study, alongside context-specific structural and psychosocial stressors. The Palestinian context is characterized by prolonged conflict, socioeconomic instability, displacement, and sustained strain on the healthcare system [14, 15]. Access to preventive and rehabilitative services is additionally constrained by movement restrictions and damage to health infrastructure [16].

Chronic psychosocial stress may also be linked to musculoskeletal symptom perception through dysregulation of the hypothalamic-pituitary-adrenal axis and central sensitization mechanisms [17, 18], although these pathways were not directly measured in the present study.

The novelty of this study lies in its combined focus on regional musculoskeletal symptom prevalence, hand discomfort, device-use and ergonomic factors, and pain-response behaviors among Palestinian university students. To our knowledge, no large-scale study in Palestine has examined these domains together in this population. This combined approach may provide a broader understanding of potentially modifiable behavioral and ergonomic correlates of musculoskeletal symptoms beyond traditional prevalence estimates.

Therefore, this study aimed to assess the prevalence of musculoskeletal symptoms and hand discomfort among Palestinian university students and to examine their associations with digital-device use, ergonomic characteristics, and pain-response behaviors during study activities. We hypothesized that longer device-use duration, suboptimal ergonomic characteristics, and continuing study activities despite pain would be associated with higher prevalence of musculoskeletal symptoms and hand discomfort.

Methods

Study design, setting and population

A cross-sectional study was conducted among university students in the West Bank, Palestine, between February and April 2026. Participants were recruited from multiple universities using a convenience sampling approach through both paper-based and online questionnaire distribution. Because probability sampling was not feasible, the sample should be interpreted as a large convenience sample rather than a statistically representative sample of all Palestinian university students. This manuscript was prepared in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline for cross-sectional studies. A completed STROBE checklist was submitted as supplementary material (Supplementary file 1).

Eligibility criteria included being enrolled in a university, aged 18 years or older, and using digital devices (computer, laptop, tablet, or smartphone) for academic purposes for at least five hours per week. Participants were required to be able to read Arabic or English and to provide written informed consent.

Students younger than 18 years, non-university attendees, and individuals with a history of significant musculoskeletal trauma or surgery within the past six months were excluded, as were those with neurological or systemic conditions known to affect musculoskeletal symptoms (including diabetic neuropathy, stroke with residual deficits, multiple sclerosis, and rheumatoid arthritis). Participants who did not complete the questionnaire or declined consent were also excluded.

Sample size

As participants were recruited from multiple universities in the West Bank and no unified registry of eligible students across participating institutions was available, the sample-size calculation used An-Najah National University, the largest contributing institution, as a pragmatic reference population of approximately 25,000 students [19]. Using a 95% confidence level, 5% margin of error, and a 50% expected response distribution, the minimum required sample size for prevalence estimation was calculated as 379 participants using an online sample-size calculator [20]. To increase precision, improve statistical power, and account for incomplete responses, the final analytic sample was expanded to 1,052 participants. Because the initial calculation was designed for prevalence estimation rather than multivariable regression modeling, regression-model stability was evaluated after model fitting using events-per-variable and sparse outcome-by-category cell checks.

Data collection

Data were collected using a structured questionnaire consisting of five sections developed based on previously published literature and adapted tools [21–23]. The first section assessed sociodemographic characteristics, including age, sex, marital status, academic discipline, year of study, residence, and physical activity. The second section assessed digital-device use patterns, including type of device, daily duration of use, and posture during use.

The third section included the Cornell Hand Discomfort Questionnaire [24], which was used to assess hand discomfort by anatomical hand region, frequency, severity, and interference with study or work. Because a formally validated Arabic version of the hand-specific Cornell questionnaire for Palestinian university students was not available, the questionnaire items were adapted for this study after review by the research team and expert panel, with pilot pretesting to assess clarity and comprehension. The fourth section included the Arabic version of the Nordic Musculoskeletal Questionnaire [25], which was used to assess musculoskeletal symptoms across nine body regions.

The fifth section assessed RSI awareness, perceived relationship between prolonged device use and musculoskeletal pain, preventive behaviors, and actions taken when pain occurred during study activities. These items were developed based on the study objectives and relevant literature and were reviewed for content validity by the expert panel. They were analyzed as individual descriptive and behavioral variables rather than as a composite knowledge, attitude, or behavior scale. The questionnaire was administered in both paper-based and online formats.

Validity and reliability

Content validity was reviewed by an expert panel with relevant clinical, public health, and research-methodology expertise. The panel assessed item relevance, clarity, cultural appropriateness, and coverage of the study objectives. Minor wording and formatting revisions were made based on their feedback. A pilot pretest involving 10 students was then conducted to assess questionnaire clarity, comprehension, administration time, and feasibility. Pilot data were excluded from the final analysis. Internal consistency was assessed using the final analytic sample.

The Nordic Musculoskeletal Questionnaire demonstrated good internal consistency for the 12-month and 7-day symptom items, with Cronbach’s alpha values of 0.801 and 0.828, respectively. Among participants who completed the hand-discomfort items, the Cornell Hand Discomfort Questionnaire demonstrated acceptable to excellent internal consistency. For the right hand, Cronbach’s alpha values were 0.726 for frequency, 0.724 for severity, 0.844 for effect on study/work, and 0.873 for the total 18-item scale. For the left hand, corresponding alpha values were 0.878, 0.849, 0.862, and 0.942, respectively.

Ethical approval

Ethical approval was obtained from the Institutional Review Board of An-Najah National University (Med. Nov. 2025/169). The study was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to participation.

Confidentiality and anonymity were strictly maintained, and data were accessible only to the research team. Data collection was conducted in private settings to ensure participant comfort and privacy.

Statistical analysis

Data were analyzed using IBM SPSS Statistics version 29.0. Data cleaning and coding were performed prior to analysis. Questionnaire items were set as mandatory in the online form, and paper-based questionnaires were reviewed for completeness before analysis. Participants who declined consent, did not meet eligibility criteria, or submitted incomplete questionnaires were excluded from the final analytic sample. Analyses were therefore conducted using complete-case data.

Categorical variables were presented as frequencies and percentages, while continuous variables were presented as means and standard deviations. Chi-square tests were used to examine bivariable associations between participant characteristics, study habits, device-use patterns, ergonomic factors, and musculoskeletal outcomes, including regional symptoms and hand discomfort.

Variables associated with each outcome in bivariable analyses at p < 0.05 were considered for inclusion in outcome-specific multivariable binary logistic regression models. This screening approach was used to reduce model complexity across multiple regional outcomes; however, because bivariable screening may exclude potentially relevant confounders, adjusted estimates were interpreted as exploratory associations rather than causal effects. Model interpretation was further guided by theoretical plausibility, events-per-variable values, sparse outcome-by-category cell checks, and model diagnostics. Adjusted odds ratios (aORs) with 95% confidence intervals (CIs) were reported.

Model diagnostics were performed for the multivariable models. Multicollinearity was assessed using variance inflation factors, model calibration using the Hosmer–Lemeshow goodness-of-fit test, discriminatory performance using the area under the receiver operating characteristic curve, and explanatory performance using Nagelkerke pseudo-R². Events-per-variable and sparse outcome-by-category cells were reviewed to identify potentially unstable estimates.

Benjamini–Hochberg false-discovery-rate correction was applied to account for multiple comparisons across musculoskeletal outcomes. Ordinal logistic regression was used to assess factors associated with overall pain severity, categorized as none, mild, moderate, or severe. To assess the influence of the female-predominant sample, sex-weighted sensitivity analyses were conducted for the primary multivariable logistic regression models. Weights were calculated as the target sex proportion divided by the observed sex proportion, allowing male and female participants to contribute equally to the weighted models. Weighted adjusted odds ratios were compared with the primary unweighted estimates. Model-performance indicators for full-sample and sex-weighted models were also compared. These analyses were interpreted as sensitivity checks and not as corrections for non-probability sampling or evidence of population representativeness.

RSI awareness, perceived device-related risk, preventive behaviors, and pain-response behaviors were analyzed as individual categorical variables. No composite knowledge, attitude, or behavior score was calculated; therefore, factor analysis and internal-consistency testing were not applied to these items.

All tests were two-tailed, and statistical significance was defined as p < 0.05.

Results

Participant characteristics

A total of 1,052 university students participated in the study. The majority were aged 20–24 years (58.2%), followed by those under 20 years (30.9%). Females constituted 77.8% of the sample, and most participants were single (86.9%). Because female students were overrepresented in the sample, prevalence estimates were interpreted with caution and sex-weighted sensitivity analyses were performed to assess the influence of sample imbalance on selected regression models.

Regarding academic characteristics, health and medical sciences represented the largest group (38.3%), followed by engineering and information technology (22.5%). First-year students accounted for 27.9% of participants. In terms of residence, 41.8% lived in villages and 39.1% in cities (Table 1).

Table 1.

Demographic characteristics of participating university students (n = 1,052)

Variable	Category	n (%)
Age	Under 20	325 (30.9)
	20–24	612 (58.2)
	25–29	43 (4.1)
	30 or older	72 (6.8)
Gender	Male	234 (22.2)
	Female	818 (77.8)
Marital status	Single	914 (86.9)
	Not single	138 (13.1)
Faculty	Engineering & IT	237 (22.5)
	Health and medical sciences	403 (38.3)
	Humanities & social sciences	88 (8.4)
	Other	324 (30.8)
Academic year	1st	293 (27.9)
	2nd	209 (19.9)
	3rd	259 (24.6)
	4th	194 (18.4)
	5th or more	97 (9.2)
Residence	City	411 (39.1)
	Refugee camp	31 (2.9)
	Town	170 (16.2)
	Village	440 (41.8)

Study habits and device use

Approximately 51.0% of participants reported studying less than 3 h per day, while 35.3% reported 3–5 h. Nearly half (46.1%) reported study sessions lasting 30–60 min.

Smartphone use was highly prevalent, with 72.6% reporting more than 5 h daily use. Laptop use of 2–4 h per day was reported by 33.6% of participants.

Smartphones were the most commonly used study device (68.3%), followed by laptops (55.5%). Nearly half of participants reported variable posture (47.4%), and 52.9% used screens below eye level. The majority were right-handed (84.6%), and 20.7% reported regular physical activity (Table 2).

Table 2.

Study habits, device use patterns, and ergonomic characteristics of participating university students (n = 1,052)

Variable	Category	n (%)
Daily study hours	< 3	537 (51.0)
	3–5	371 (35.3)
	> 5	144 (13.7)
Session duration without break	< 30 min	187 (17.8)
	30–60 min	485 (46.1)
	1–2 h	234 (22.2)
	> 2 h	146 (13.9)
Daily laptop/desktop hours	< 2	313 (29.8)
	2–4	353 (33.6)
	5–7	236 (22.4)
	> 7	150 (14.3)
Daily smartphone hours	< 2	54 (5.1)
	2–4	235 (22.3)
	5–7	453 (43.1)
	> 7	310 (29.5)
Device used for study	Laptop	584 (55.5)
	iPad/Tablet	467 (44.4)
	Desktop	64 (6.1)
	Smartphone	718 (68.3)
	Paper books	10 (1.0)
	Other/unclear	1 (0.1)
Control method	Mouse	309 (29.4)
	Touchpad	444 (42.2)
	Touchscreen	299 (28.4)
Sitting posture	Upright on office chair	268 (25.5)
	Variable/not fixed	499 (47.4)
	Sitting on sofa/bed with device on lap	235 (22.3)
	Lying down	50 (4.8)
Screen level	Below eye level	556 (52.9)
	At eye level	477 (45.3)
	Above eye level	19 (1.8)
Dominant hand	Right	890 (84.6)
	Left/Both	162 (15.4)
Regular physical activity	No	834 (79.3)
	Yes	218 (20.7)

For device used for study, participants could select multiple devices; percentages do not sum to 100%

Hand discomfort (Cornell Hand Discomfort Questionnaire)

Right-hand discomfort was reported by 45.2% of participants, compared with 13.5% for the left hand.

The most frequently affected regions in the right hand were area F (35.2%) and area E (33.6%). A similar but lower pattern was observed in the left hand (area F: 8.9%, area E: 8.6%).

Higher reported frequency of discomfort generally coincided with greater severity and functional impact across both hands. Detailed findings are presented in Fig. 1; Table 3.

Fig. 1.

Frequency of discomfort, pain severity, and functional impact by hand area: right versus left hand (Cornell Hand Discomfort Scale)

Table 3.

Cornell Hand Discomfort Scale findings: frequency, severity, and functional impact by hand area and laterality (n = 1,052)

Measure	Hand	Area A	Area B	Area C	Area D	Area E	Area F
Frequency (any discomfort)	Right	234 (22.2)	235 (22.3)	241 (22.9)	265 (25.2)	353 (33.6)	370 (35.2)
	Left	60 (5.7)	63 (6.0)	71 (6.7)	68 (6.5)	90 (8.6)	94 (8.9)
Severity (any discomfort)	Right	234 (22.2)	200 (19.0)	244 (23.2)	265 (25.2)	341 (32.4)	368 (35.0)
	Left	65 (6.2)	67 (6.4)	70 (6.7)	75 (7.1)	94 (8.9)	99 (9.4)
Effect on study/work	Right	128 (12.2)	128 (12.2)	153 (14.5)	149 (14.2)	205 (19.5)	232 (22.1)
	Left	42 (4.0)	47 (4.5)	49 (4.7)	50 (4.8)	69 (6.6)	74 (7.0)

Musculoskeletal symptoms by body region and functional impact

Musculoskeletal symptoms were highly prevalent across multiple body regions. The 12-month prevalence was highest for neck pain (82.4%), followed by shoulder pain (75.0%), lower back pain (73.4%), and upper back pain (65.8%). Other regions were affected at lower rates, including wrists/hands (49.7%), knees (49.1%), and hips/thighs (41.0%).

Seven-day prevalence showed a similar distribution, with neck pain reported by 67.0% of participants, followed by shoulders (61.2%), lower back (57.6%), and upper back (51.8%) (Fig. 2; Table 4).

Fig. 2.

Twelve-month and 7-day musculoskeletal symptom prevalence across nine body regions

Table 4.

Twelve-month and 7-day musculoskeletal symptom prevalence by body region and functional impact (n = 1,052)

Variable	Category	n (%)
Neck	12-month prevalence	867 (82.4)
	7-day prevalence	705 (67.0)
Shoulders	12-month prevalence	789 (75.0)
	7-day prevalence	644 (61.2)
Upper back	12-month prevalence	692 (65.8)
	7-day prevalence	545 (51.8)
Lower back	12-month prevalence	772 (73.4)
	7-day prevalence	606 (57.6)
Elbows	12-month prevalence	359 (34.1)
	7-day prevalence	238 (22.6)
Wrists/Hands	12-month prevalence	523 (49.7)
	7-day prevalence	383 (36.4)
Hips/Thighs	12-month prevalence	431 (41.0)
	7-day prevalence	311 (29.6)
Knees	12-month prevalence	517 (49.1)
	7-day prevalence	366 (34.8)
Ankles/Feet	12-month prevalence	403 (38.3)
	7-day prevalence	285 (27.1)
Functional impact	Reduced activity	624 (59.3)
	Absence from study/work	271 (25.8)

Reduced daily activity due to musculoskeletal symptoms was reported by 59.3% of participants, while 25.8% reported absence from study or work.

Pain severity and predictors of overall pain severity

Overall pain was common among participants. Moderate pain was the most frequently reported level, affecting 592 students (56.3%), followed by mild pain in 312 students (29.7%) and severe pain in 86 students (8.2%). Only 62 participants (5.9%) reported no pain.

In the multivariable ordinal logistic regression analysis, younger age, male sex, and shorter uninterrupted study sessions were associated with lower odds of higher overall pain severity after false-discovery-rate correction. Compared with students aged 30 years or older, students aged < 20 years had lower odds of higher pain severity (aOR 0.35, 95% CI 0.19–0.63), as did those aged 20–24 years (aOR 0.33, 95% CI 0.19–0.60). Male students also had lower odds of higher pain severity compared with female students (aOR 0.40, 95% CI 0.29–0.55). Shorter uninterrupted study sessions were similarly associated with lower odds of higher pain severity, including sessions lasting < 30 min (aOR 0.43, 95% CI 0.27–0.68) and 30–60 min (aOR 0.52, 95% CI 0.35–0.77), compared with sessions lasting > 2 h. Associations that remained statistically significant after false-discovery-rate correction are presented in Supplementary Table S10.

RSI awareness, preventive behaviors, and pain-response behaviors

RSI awareness was reported by 657 participants (62.5%), while 196 students (18.6%) had not heard about RSI and 199 (18.9%) were unsure. Awareness of the possible role of prolonged device use was high, with 978 participants (93.0%) reporting that extended digital-device use could contribute to musculoskeletal pain.

Preventive behaviors were also common. Overall, 826 students (78.5%) reported using at least one preventive measure. The most frequently reported strategy was changing posture or modifying the way devices were used, reported by 661 students (62.8%), followed by taking regular breaks (490 students, 46.6%), performing hand or neck stretching exercises (325 students, 30.9%), and adjusting the screen or study setup (247 students, 23.5%).

Students varied in how they responded to pain during studying. Nearly half of participants reported taking a rest when pain occurred (n = 525, 49.9%). Other responses included performing stretching exercises (n = 253, 24.0%), continuing studying without changing behavior (n = 140, 13.3%), and stopping or postponing studying because of pain (n = 103, 9.8%). Only 31 participants (2.9%) reported no pain during studying. These items were analyzed descriptively rather than as a composite knowledge, attitude, or behavior scale.

Sex differences in musculoskeletal symptoms

Female participants reported higher musculoskeletal symptom prevalence across most body regions compared with males. For example, 12-month shoulder pain was reported by 80.7% of females versus 55.1% of males, and lower back pain by 76.2% versus 63.7%. Similar patterns were observed across most regions in both 12-month and 7-day analyses (Fig. 3). Because females were overrepresented in the sample, sex-weighted sensitivity analyses were performed. These analyses showed that the direction of the main adjusted associations was generally unchanged after weighting male and female participants to contribute equally, and most primary associations remained statistically significant. Several marginal associations, including selected age, academic-year, residence, and marital-status comparisons, were attenuated after weighting. Complete comparisons of primary and sex-weighted adjusted estimates are presented in Supplementary Table S11, with model-performance diagnostics shown in Supplementary Table S12.

Fig. 3.

Sex differences in 12-month musculoskeletal symptom prevalence across nine body regions

Factors associated with hand pain

In multivariable analysis, younger age and smartphone use of 2–4 h per day were associated with lower odds of right-hand pain, while male sex was associated with lower odds compared with female sex. Behavioral responses to pain during studying were associated with higher odds of right-hand pain, with the highest estimates observed among students who continued studying without change or stopped/postponed studying. Preventive hand/neck stretching was associated with higher odds of both right- and left-hand pain, which may reflect symptom-driven behavior rather than a harmful effect of stretching. For left-hand pain, shorter uninterrupted study sessions and right-hand dominance were associated with lower odds. Because some hand-pain models had sparse cells and lower events-per-variable values, these findings were interpreted cautiously (Table 5).

Table 5.

Multivariable logistic regression analysis of factors associated with right- and left-hand pain in the preceding 7 days (n = 1,052)

Variable	Category	Right-hand pain aOR	95% CI	p-value	Left-hand pain aOR	95% CI	p-value
Age	> 30 years (Ref)	1.00	—	—	1.00	—	—
	< 20 years	0.354	0.166–0.758	0.008	1.153	0.487–2.728	0.746
	20–24 years	0.402	0.194–0.831	0.014	1.082	0.473–2.473	0.853
	25–29 years	1.341	0.517–3.478	0.546	2.330	0.774–7.019	0.133
Gender	Female (Ref)	1.00	—	—	1.00	—	—
	Male	0.653	0.459–0.930	0.018	0.779	0.504–1.206	0.263
Marital status	Single (Ref)	1.00	—	—	—	—	—
	Not single	0.819	0.490–1.367	0.444	—	—	—
Faculty major	Others (Ref)	1.00	—	—	—	—	—
	Engineering & IT	1.143	0.773–1.690	0.503	—	—	—
	Health and medical sciences	0.802	0.567–1.136	0.215	—	—	—
	Humanities & social sciences	0.837	0.503–1.393	0.493	—	—	—
Place of residence	Village (Ref)	1.00	—	—	—	—	—
	City	0.760	0.566–1.019	0.066	—	—	—
	Refugee camp	1.388	0.630–3.058	0.416	—	—	—
	Town	0.890	0.610–1.300	0.548	—	—	—
Daily smartphone hours	> 7 h (Ref)	1.00	—	—	1.00	—	—
	< 2 h	0.685	0.367–1.282	0.237	1.499	0.714–3.149	0.285
	2–4 h	0.625	0.429–0.911	0.014	0.851	0.506–1.430	0.543
	5–7 h	0.860	0.623–1.186	0.357	0.723	0.458–1.139	0.162
Session duration without break	> 2 h (Ref)	1.00	—	—	1.00	—	—
	< 30 min	0.716	0.443–1.157	0.172	0.359	0.177–0.727	0.004
	30–60 min	1.162	0.773–1.745	0.470	0.752	0.449–1.259	0.278
	1–2 h	1.072	0.687–1.674	0.760	0.724	0.408–1.285	0.270
Sitting posture	Lying down (Ref)	1.00	—	—	—	—	—
	Upright on office chair	1.085	0.553–2.127	0.813	—	—	—
	Variable/not fixed	1.291	0.675–2.469	0.440	—	—	—
	Sofa/bed with device on lap	1.251	0.633–2.474	0.519	—	—	—
Computer control tool	Direct touchscreen (Ref)	1.00	—	—	—	—	—
	Mouse	0.777	0.528–1.144	0.202	—	—	—
	Laptop touchpad	0.808	0.577–1.131	0.214	—	—	—
Action when pain occurs	No pain during studying (Ref)	1.00	—	—	1.00	—	—
	Continue without change	8.588	2.708–27.239	< 0.001	3.147	0.680–14.551	0.142
	Take a rest	4.027	1.321–12.279	0.014	1.439	0.323–6.416	0.633
	Do stretching exercises	4.475	1.439–13.924	0.010	1.681	0.368–7.688	0.503
	Stop/postpone studying	8.505	2.640–27.402	< 0.001	2.640	0.556–12.529	0.222
Preventive: hand/neck stretching	No (Ref)	1.00	—	—	1.00	—	—
	Yes	1.555	1.129–2.140	0.007	1.616	1.050–2.488	0.029
Preventive: adjust screen/setup	No (Ref)	1.00	—	—	—	—	—
	Yes	0.770	0.555–1.069	0.119	—	—	—
Dominant hand	Both (Ref)	—	—	—	1.00	—	—
	Right	—	—	—	0.482	0.285–0.816	0.007
	Left	—	—	—	1.813	0.843–3.896	0.128

aOR Adjusted odds ratio, CI Confidence interval, Ref Reference category. Categorical variables were analyzed using indicator coding. Predictors shown were included in outcome-specific multivariable models after bivariable screening; dashes indicate that the variable was not included in the corresponding outcome model or was not applicable. Estimates with wide confidence intervals should be interpreted cautiously because they may reflect sparse outcome-by-category cells and limited precision. For preventive hand/neck stretching, the observed positive association with hand pain should be interpreted cautiously because the cross-sectional design cannot distinguish prevention from symptom-driven behavior. False-discovery-rate correction was applied as a robustness check

Model diagnostics for hand-pain models are presented in Supplementary Table S9, and sex-weighted estimates are presented in Supplementary Table S11.

Factors associated with 7-day musculoskeletal symptoms

Younger age and male sex were consistently associated with lower odds of 7-day musculoskeletal symptoms across multiple body regions. Behavioral responses to pain during studying were associated with higher odds of symptoms in several regions, including upper back symptoms among students who continued activity without change (aOR 9.21, 95% CI 2.91–29.22). However, estimates involving pain-response categories were interpreted cautiously because several confidence intervals were wide and sparse-cell checks indicated limited precision in some outcome-by-category combinations. Shorter study sessions (< 30 min) were associated with lower odds of neck and lower-back symptoms. Use of a mouse or touchpad was associated with lower odds of wrist/hand symptoms compared with touchscreen use. First-year students had higher odds of lower-limb symptoms, while residence in towns was associated with lower odds of certain lower-body symptoms. After sex-weighted sensitivity analysis, the direction of the main associations remained unchanged, although several marginal associations were attenuated. Complete primary models are presented in Supplementary Tables S7a–S7c, sex-weighted sensitivity estimates in Supplementary Table S11, model diagnostics in Supplementary Tables S8, S9, and S12, and sparse-cell checks in Supplementary Table S13.

Model diagnostics were performed for the current multivariable models. Across the nine 7-day musculoskeletal symptom models, AUC values ranged from 0.649 to 0.689, indicating modest discriminatory performance, and Nagelkerke pseudo-R² values ranged from 0.117 to 0.179. Hosmer–Lemeshow testing suggested acceptable calibration for most models, although the lower-back model showed evidence of poorer calibration and was interpreted cautiously. This poorer calibration may reflect greater heterogeneity in lower-back pain etiology, as lower-back symptoms can arise from multiple unmeasured biomechanical, behavioral, psychosocial, and activity-related factors not fully captured in the model.Events-per-variable values were generally acceptable, but sparse outcome-by-category cells were identified for several predictors, particularly pain-response categories. Hand-pain model diagnostics showed modest discrimination, with lower events-per-variable for the left-hand pain model. Ordinal regression analysis of overall pain severity showed that younger age, male sex, and shorter uninterrupted study sessions were associated with lower odds of higher pain severity. Sex-weighted sensitivity analyses showed broadly similar model-performance indicators to the full-sample models. Detailed diagnostics are presented in Supplementary Tables S8, S9, and S12, and sparse-cell checks are presented in Supplementary Table S13.

Discussion

This cross-sectional study identified a substantial burden of musculoskeletal symptoms among Palestinian university students, with the highest 12-month prevalence observed for neck, shoulder, lower back, and upper back symptoms. In the Palestinian context—characterized by prolonged conflict, socioeconomic instability, and constrained healthcare access—MSK symptoms may represent a compounded burden extending beyond physical morbidity [1, 26].

A key contribution of this study is the identification of behavioral responses to pain during studying as important correlates of MSK symptoms and functional impact. Continuing academic activity without modification after pain onset was associated with higher odds across multiple body regions, with particularly large estimates for upper back symptoms. However, estimates involving pain-response categories, particularly the upper-back association, should be interpreted with extreme caution because sparse outcome-by-category cells and wide confidence intervals indicate limited precision; therefore, the magnitude of this effect may be overestimated and should be viewed as an imprecise signal of association rather than a precise effect-size estimate. While most prior research has emphasized ergonomic exposures, this behavioral dimension remains insufficiently explored and may represent an important and modifiable target for intervention.

The predominance of neck, shoulder, and upper back symptoms aligns with established global patterns in student populations, largely attributed to sustained forward-head posture during device use [5, 6]. In this study, suboptimal ergonomic practices were common, including widespread use of screens below eye level and variable posture. These findings are consistent with the “text neck” paradigm described by Hansraj, in which anterior head displacement substantially increases cervical spine loading [11]. Similarly, Hakala et al. demonstrated that prolonged daily computer use is associated with increased risk of neck and shoulder pain among adolescents [12]. The high prevalence of smartphone use in this cohort, both as a primary study tool and in terms of daily duration, provides a plausible contextual factor that may partly explain the observed cervical and shoulder symptom burden.

Upright and variable sitting positions were associated with increased odds of lower back symptoms compared with lying down. This finding should be interpreted cautiously given the cross-sectional design and the possibility of selection bias; students who study while lying down may differ in study duration, breaks, or other unmeasured behavioral factors. Further investigation with prospective data is needed to clarify this association. Marital status was associated with lower back symptoms but not with symptoms in other regions, which may reflect differences in physical or psychosocial demands [27]; however, given the marginal statistical evidence and lack of multi-region replication, this finding should be interpreted cautiously and requires confirmation in future studies.

Additionally, the use of indirect input devices, including a mouse or touchpad, was associated with lower odds of wrist/hand symptoms compared with touchscreen use, consistent with prior ergonomic research demonstrating increased strain associated with sustained non-neutral wrist positioning [13]. Wrist and hand symptoms were reported by nearly half of participants, with findings suggesting dominant-hand lateralization consistent with repetitive strain patterns. Pain-response behaviors during studying were associated with higher odds of right-hand pain. Hand/neck stretching was also associated with higher odds of hand pain in the unweighted models, which should not be interpreted as evidence that stretching increases pain. A more plausible explanation is reverse causation, whereby students experiencing hand discomfort were more likely to perform stretching as a coping or self-management behavior. This finding should be interpreted cautiously in light of the cross-sectional design, sparse-cell considerations, and sex-weighted sensitivity results [28].

Female students demonstrated consistently higher MSK symptom prevalence across most body regions, in line with existing literature documenting sex differences in pain perception and musculoskeletal burden [29, 30]. These differences are likely multifactorial, involving biological, biomechanical, and psychosocial mechanisms [31]. Sex-weighted sensitivity analyses showed that the direction of the main adjusted associations was generally unchanged after weighting male and female participants to contribute equally, although several marginal associations were attenuated. These findings support the robustness of the main patterns but do not eliminate the possibility of selection bias from the female-predominant convenience sample.

Younger age was a consistent protective factor across most models, supporting the role of cumulative exposure and duration-dependent risk in MSK symptom development [32]. In contrast, first-year students demonstrated higher odds of lower limb symptoms, which may reflect adaptation to new physical and environmental demands during the transition to university life, including increased mobility and changes in daily routines [33]. It is also possible that senior students in Palestine, particularly those approaching graduation or navigating the constrained Palestinian labor market, experience greater sedentary behaviors as they focus intensely on academic completion, paradoxically reducing lower limb loading.

Modifiable behavioral factors emerged as important correlates of MSK symptoms. Shorter study sessions and the use of indirect input devices were associated with lower symptom burden, while continuing activity despite pain was associated with higher odds across multiple regions. These findings are consistent with evidence demonstrating that prolonged static loading and insufficient recovery are associated with musculoskeletal strain, and that even brief microbreaks may help reduce discomfort during prolonged study or computer-based work [34, 35]. Despite this, a significant proportion of students reported prolonged study sessions without breaks, highlighting a potential opportunity for ergonomic and behavioral intervention.

A notable finding is the disconnect between awareness and behavior. While most participants recognized the relationship between device use and musculoskeletal pain, preventive practices were inconsistently applied. This knowledge–behavior gap has been reported in other populations and appears particularly relevant in the Palestinian context, where structural barriers, limited healthcare access, and competing academic demands may constrain behavior change [15, 36]. Because preventive behaviors were measured cross-sectionally, their associations with pain should be interpreted cautiously and may partly reflect symptom-driven behavior rather than prevention.

The functional impact of MSK symptoms in this cohort was substantial, with a majority reporting reduced daily activity and a significant proportion reporting absence from academic responsibilities. In a setting where academic performance is closely linked to future employment opportunities, these findings have broader socioeconomic implications. MSK symptoms in this population may therefore represent not only a clinical issue but also a public health concern with potential academic and socioeconomic implications.

Strengths and limitations

This study has several strengths. First, it included a large sample of Palestinian university students and assessed musculoskeletal symptoms across multiple body regions rather than focusing on a single anatomical site. Second, the inclusion of both 12-month and 7-day symptom prevalence provided a broader assessment of musculoskeletal symptom burden. Third, the study combined regional musculoskeletal symptoms, detailed hand discomfort, digital-device use, ergonomic characteristics, and pain-response behaviors, allowing a more comprehensive description of potentially modifiable correlates of musculoskeletal symptoms in this population. Fourth, the use of the Arabic version of the Nordic Musculoskeletal Questionnaire strengthened symptom assessment across the nine body regions. Additional robustness checks, including false-discovery-rate correction, model diagnostics, events-per-variable assessment, sparse-cell checks, ordinal regression for pain severity, and sex-weighted sensitivity analyses, further strengthened the analytical reporting.

Several limitations should also be acknowledged. The cross-sectional design limits causal inference, and temporality cannot be established. Reverse causation cannot be excluded, particularly for pain-response behaviors and preventive behaviors, as students experiencing symptoms may have been more likely to rest, stretch, or modify their study behavior. The convenience sampling strategy limits generalizability, and because recruitment included online distribution, the total number of students who viewed or received the invitation could not be reliably determined; therefore, a formal response rate could not be calculated. The sample was female-predominant, with women representing 77.8% of participants, and prevalence estimates may therefore not reflect the sex distribution or true musculoskeletal symptom burden among all Palestinian university students. Although sex-weighted sensitivity analyses were conducted, weighting cannot eliminate selection bias or replace probability-based sampling.

All outcomes were self-reported, which may introduce recall and reporting bias. Some regression estimates had wide confidence intervals or sparse outcome-by-category cells, particularly those involving pain-response categories, and should therefore be interpreted cautiously. Model diagnostics showed modest discriminatory performance, indicating that additional unmeasured factors may contribute to musculoskeletal symptoms. Psychosocial stressors were not directly measured, although they may be relevant in the Palestinian context. The RSI awareness, preventive-behavior, and pain-response items were study-specific and were not designed as a validated psychometric scale. In addition, although pilot pretesting helped assess clarity, comprehension, and feasibility, the small pilot size of 10 students was insufficient for a full psychometric assessment of the adapted and study-specific items. Although these items were developed from the study objectives, literature, and expert-panel review, no factor analysis, concurrent validity assessment, or test–retest reliability assessment was performed; therefore, these findings should be interpreted as descriptive and exploratory. Finally, although the Cornell Hand Discomfort Questionnaire demonstrated acceptable to excellent internal consistency in this sample, it was adapted for this study and has not been formally validated in Arabic among Palestinian university students. Findings related to hand discomfort should therefore be interpreted with appropriate caution.

Conclusion

MSK symptoms were highly prevalent in this convenience sample of Palestinian university students and were associated with behavioral, ergonomic, and device-use factors. Given the cross-sectional and convenience-sampled design, causal inferences cannot be made. Some estimates, particularly those involving pain-response categories, should be interpreted cautiously because of limited precision. These findings identify potential modifiable factors that warrant investigation in prospective studies. If future research confirms these associations prospectively, integrated ergonomic and behavioral strategies may be beneficial for reducing musculoskeletal symptom burden in university settings.

Abbreviations

aOR

Adjusted Odds Ratio

CI

Confidence Interval

GBD

Global Burden of Disease

HPA

Hypothalamic-Pituitary-Adrenal axis

IRB

Institutional Review Board

MSK

Musculoskeletal

NNU

An-Najah National University

OR

Odds Ratio

RSI

Repetitive Strain Injury

SD

Standard Deviation

YLDs

Years Lived with Disability

Authors’ contributions

AAB contributed to drafting the Discussion and Limitations and participated in data collection; MB performed the formal analysis and data curation; MAB contributed to drafting the Introduction and data collection; OAM led the study conceptualization, coordinated major data collection, and drafted the Results and Conclusion; LM contributed to research conceptualization, supervision, and critical revision of the manuscript; and JMD conducted the formal analysis, led the preparation of tables and figures, and contributed to drafting the Results. All authors reviewed and approved the final version of the manuscript.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Institutional Review Board (IRB) of An-Najah National University (Protocol No. Med. Nov. 2025/169). All procedures were conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants prior to inclusion in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.