Study selection
A total of 691 related studies were identified by searching seven electronic databases, which were reduced to 493 articles after removing duplicates; 421 were excluded after reading the titles and abstracts. The reasons for exclusion included reviews, dissertations, conference papers, and research objects other than college students. After carefully reading the full text of the remaining 72 articles, 46 studies were excluded for the following reasons: (1) a lack of OR values and 95% CI or unable to convert (n = 24); (2) reporting outcomes with constituent ratio or frequency (n = 15); and (3) no full-text articles (n = 7). Finally, 26 studies were included, 11 in English and 15 in Chinese. A PRISMA flow diagram is illustrated in Fig. 1.
Study Characteristics And Quality Assessment
The basic characteristics of the included studies and quality assessment results are summarized in Table 1 (At the end of text). A total of 26 studies were included [12,30−54], including 21 cross-sectional studies, 1 cohort study, and 4 case-control studies. Of these, 14 were rated as high quality and 12 as medium quality, with an overall average score of 7.38. The overall quality of the study was good.
Table 1
Baseline characteristics of all included studies in our meta-analysis.
Literature | Publication (Year) | Geographic region | Study design | Sample size | Mean age | Risk factors reported | Quality score |
Li YW[30] | 2020 | China | Cross-sectional study | 4848 | 19. 16 ± 1. 39 | 4、13、14 | 7 |
Hu YY[31] | 2018 | China | Cross-sectional study | 392 | NR | 2、3、7 | 8 |
Meng FY[32] | 2020 | China | Cross-sectional study | 836 | 20.15 ± 1.96 | 9、13 | 9 |
Ye F[33] | 2016 | China | Cross-sectional study | 1117 | 20.62 ± 1.266 | 4、14、17、20 | 8 |
Sun Z[34] | 2019 | China | Cross-sectional study | 448 | 19.23 ± 1.05 | 2、3、5、15、21、22 | 6 |
Su JT[35] | 2013 | China | Case-control study | NG:180 CG:180 | NG:23.77 ± 2.44 CG:23.62 ± 2.33 | 1、3、9、12、14、20、23 | 8 |
Zhu XTg[36] | 2021 | China | Cross-sectional study | 358 | 19. 91 ± 1. 13 | 1、3、6、9、12、13、16 | 8 |
Ren YC[37] | 2013 | China | Cross-sectional study | 2026 | NR | 3、13、14、17、24 | 7 |
Wang CL[38] | 2014 | China | Cross-sectional study | 303 | NR | 12、24 | 7 |
Tian ZY[39] | 2019 | China | Cross-sectional study | 478 | NR | 1、2、3、6、14、29 | 6 |
Jing LN[40] | 2021 | China | Cross-sectional study | 803 | 20.40 ± 1.30 | 2、14、18 | 8 |
Wang Z[41] | 2021 | China | Cross-sectional study | 408 | 19.54 ± 1.066 | 6、13、25 | 6 |
Hashim R[42] | 2021 | United Arab Emirates | Cross-sectional study | 202 | NR | 6 | 7 |
Ayhualem S[43] | 2021 | Ethiopia | Cross-sectional study | 808 | 21.9 ± 2.15 | 7、9、13、26、27、30 | 8 |
Daher A[44] | 2021 | Israel | Cross-sectional study | 295 | 27.7 ± 8.32 | 15、16 | 6 |
Behera P[45] | 2020 | India | Cross-sectional study | 331 | 20.5 ± 1.7 | 7、8、14 | 7 |
Weleslassie GG[12] | 2020 | Ethiopia | Cross-sectional study | 419 | 22 ± 2.215 | 8、9、14、25 | 8 |
Dighriri YH[46] | 2019 | Saudi Arabia | Cross-sectional study | 440 | 22.4 ± 1.6 | 6、11、12 | 8 |
Haroon H[47] | 2018 | Pakistan | Cross-sectional study | 360 | 20.77 ± 1.47 | 6、13、15 | 8 |
Algarni AD[48] | 2017 | Saudi Arabia | Cross-sectional study | 469 | 21.4 ± 1.3 | 6、11、12 | 7 |
Alshagga MA[49] | 2013 | Malaysia | Cross-sectional study | 232 | 20.6 ± 2.2 | 6、7、10、13 | 7 |
Ndetan HT[50] | 2009 | Dallas | Cross-sectional study | 252 | 25.6 ± 4.5 | 10 | 8 |
Kanchanomai S[51] | 2011 | Thailand | Cohort study | 524 | 19.4 ± 1.1 | 7、19、28 | 8 |
Liu HT[52] | 2014 | China | Case-control study | NG:66 CG:66 | NR | 14 | 8 |
Zhang JL[53] | 2009 | China | Case-control study | NG:120 CG:120 | NG:20.03 ± 0.41 CG:21.01 ± 0.42 | 3、13、14 | 8 |
Huang ZH[54] | 2016 | China | Case-control study | NG:111 CG:337 | 19.2 ± 1.0 | 2、5 | 6 |
Risk factors: 1-improper use of the pillow; 2-staying up late; 3-improper sitting posture; 4-female; 5-frequent alcohol consumption; 6-history of neck and shoulder trauma; 7-senior grade; 8-history of pain; 9-lack of exercise; 10-obesity; 11-history of psychosomatic symptoms; 12-emotional problems; 13-long-time electronic product usage daily; 14-long-time to bow head; 15-high pressure; 16-sedentariness; 17-poor sleep quality; 18-throat inflammation; 19-improper keyboard position; 20-heavy schoolbag; 21-age; 22-sleeping on a bus or car; 23-feeling cold and wet wind; 24-neck fatigue; 25-poor head and neck posture;26-cigarette smoking; 27-lack of rest; 28-computer screen is not positioned at a level horizontal with the eyes; 29-desks and chairs not matching height; 30-many social media. |
Risk factors for neck pain in college students
Overall, 30 potential risk factors were extracted from the 26 eligible studies, including improper use of the pillow, staying up late, improper sitting posture, female, frequent alcohol consumption, history of neck and shoulder trauma, senior grade, history of pain, lack of exercise, obesity, history of psychosomatic symptoms, emotional problems, long-time electronic product usage daily, long-time to bow head, high pressure, sedentariness, poor sleep quality, throat inflammation, improper keyboard position, heavy schoolbag, age, sleeping on a bus or car, feeling cold and wet wind, neck fatigue, poor head and neck posture, cigarette smoking, lack of rest, computer screen is not positioned at a level horizontal with the eyes, desks and chairs not matching height, and many social media. There were ten risk factors, including improper use of the pillow, lack of exercise, improper sitting posture, history of neck and shoulder trauma, senior grade, staying up late, long time electronic product usage daily, long time to bow head, high stress, and emotional problems that met the criteria for inclusion in the meta-analysis. The combined results and evidence levels are presented in Table 2.
Table 2
Risk factors' pooled analysis and level of evidence.
Risk factors | Number of included studies | Pooled effects | Level of evidence |
OR[95%CI] | P value | I2 |
Improper use of the pillow | 3 | 2.20 [1.39, 3.48] | 0.0008 | 66% | Strong |
Lack of exercise | 5 | 1.88 [1.53, 2.30] | < 0.00001 | 0% | Strong |
Improper sitting posture | 7 | 1.97 [1.39, 2.78] | 0.0001 | 78% | Strong |
History of neck and shoulder trauma | 8 | 2.32 [1.79, 3.01] | < 0.00001 | 22% | Strong |
Senior grade | 5 | 2.86 [2.07, 3.95] | < 0.00001 | 5% | Strong |
Staying up late | 6 | 1.80 [1.35, 2.41] | < 0.0001 | 55% | Strong |
Long-time electronic product usage per day | 9 | 1.45 [1.27, 1.65] | < 0.00001 | 75% | Strong |
Long-time to bow head | 10 | 2.04 [1.58, 2.64] | < 0.00001 | 82% | Strong |
High pressure | 3 | 1.91 [1.50, 2.42] | < 0.00001 | 0% | Moderate |
Emotional problems | 5 | 2.09 [1.66, 2.63] | < 0.00001 | 0% | Strong |
Improper Use Of The Pillow
Three studies [35–36, 39] reported the effect of improper use of pillows on neck pain in college students, investigating a total of 1016 participants. Our results depicted that improper use of the pillow was one of the risk factors for neck pain in college students (OR = 2.20, 95% CI: 1.39 to 3.48, P = 0.0008). However, because the heterogeneity between studies was high (I2 = 66%) (Fig. 2A), a sensitivity analysis was performed. When excluding study of SuJ et al., 2013, the heterogeneity decreased significantly (I2 = 0%) (Additional File 2A), while the study conclusions remained statistically significant (P < 0.05).
Lack Of Exercise
Compared to regular exercise, five studies [12, 32, 35–36, 43] found that physical inactivity was associated with a higher prevalence of neck pain in college students, with minimal heterogeneity (OR = 1.88, 95% CI: 1.53 to 2.30, P < 0.00001, I2 = 0%) (Fig. 2B).
Improper Sitting Posture
Seven studies [31, 34–37, 39, 53] analyzed the influence of improper sitting posture on neck pain in college students, investigated all 4002 participants, and found evidence of a positive correlation between this factor and neck pain. The combined analysis results revealed that improper sitting posture caused statistically significant harm to neck pain (OR = 1.97, 95% CI: 1.39 to 2.78, P = 0.0001). However, due to the high heterogeneity between the studies (I2 = 78%) (Fig. 2C), a sensitivity analysis was performed. When excluding the study by RenY et al., 2013, the heterogeneity decreased significantly (I2 = 0%) (Additional File 2B), while the study conclusions remained statistically significant (P < 0.05).
History Of Neck And Shoulder Trauma
Eight studies [36,39,41–42,46−49] investigated all 2947 participants, and all reported the impact of a history of neck and shoulder disease on neck pain in college students. Our pooled results demonstrated that college students with a history of neck and shoulder trauma were more likely to develop neck pain, with a high prevalence and less heterogeneity (OR = 2.32, 95% CI: 1.79 to 3.01, P < 0.00001, I2 = 22%) (Fig. 2D).
Senior Grade
Five studies [31, 43, 45, 49, 51] analyzed the effect of grade differences on neck pain among 2287 college students. The combined results depicted that a higher probability and frequency of neck pain occurred at grade level (OR = 2.86, 95% CI: 2.07 to 3.95, P < 0.00001, I2 = 5%) (Fig. 2E).
Staying Up Late
Five studies [31, 34, 39–40, 54] that evaluated the association between staying up late and neck pain in college students were included in the meta-analysis. The combined results revealed that a high prevalence of neck pain in college students was associated with often staying up late (OR = 1.80, 95% CI: 1.35 to 2.41, P < 0.0001). However, due to its high heterogeneity (I2 = 55%) (Fig. 2F), it was reduced by excluding the articles one by one. Finally, the heterogeneity decreased significantly when excluding HuY et al., 2018 (I2 = 0%) (Additional File 2C), while the results remained statistically significant (P < 0.00001).
Long-time Electronic Product Usage Per Day
Nine studies [30, 32, 36–37, 41, 43, 47, 49, 53] analyzed the impact of a long-time electronic product used daily on neck pain in 9996 college students. The results of the combined meta-analysis demonstrated that prolonged use of electronics was significantly associated with neck pain in college students (OR = 1.45, 95% CI: 1.27 to 1.65, P < 0.00001). Subgroup meta-analysis revealed that both smartphone and computer use could cause neck pain in college students (P < 0.00001 and P = 0.0002, respectively), but due to the combined heterogeneity (I2 = 75%) (Fig. 2G), the sensitivity analysis found that the overall heterogeneity was still relatively large regardless of which study was removed.
Long-time To Bow Head
Ten studies [12, 30, 33, 35, 37, 39–40, 45, 52–53] reported the impact of long bowing on neck pain in college students, investigating a total of 10388 participants. The summary results demonstrated that a long time to bow head would lead to frequent neck pain among college students (OR = 2.04, 95% CI: 1.58 to 2.64, P < 0.00001). The results of the subgroup analysis revealed that both a single time and total time to bow the head had a certain impact on neck pain in college students (P = 0.0007 and P < 0.00001). However, due to the large heterogeneity of the combined results (I2 = 82%), the heterogeneity could not be reduced by excluding studies individually (Fig. 2H). The funnel plot depicted bilateral asymmetry and, therefore, a possible publication bias (Fig. 3).
High Pressure
Three studies [34, 44, 47] analyzed the effects of high pressure on neck pain among college students. The combined results showed that the greater the pressure, the more likely they were to cause neck pain, with minimal heterogeneity (OR = 1.91, 95% CI: 1.50 to 2.42, P < 0.00001, I2 = 0%) (Fig. 2I).
Emotional Problems
Five studies [35–36, 38, 46, 48] that evaluated the relationship between emotional problems and neck pain in college students were included in the meta-analysis. The combined results demonstrated that college students with emotional problems were more likely to have neck pain than happy people (OR = 2.09; 95% CI: 1.66,2.63, P < 0.00001, I2 = 0%) (Fig. 2J).
Other Risk Factors
Although other risk factors, such as female sex, obesity, and frequent alcohol consumption, were found in their respective studies to be neck pain among college students, it is worth noting that these factors were only evaluated in one or two studies, indicating that there may not be enough evidence to prove the association with neck pain; therefore, they are not in summary analysis, but the data are presented in the form of a table (Additional File 3).