A scoping review of internet addiction and academic performance among health sciences students

Abstract

Introduction

Internet addiction (IA) poses a public health concern in medical education settings. Despite students’ increasing reliance on the internet for learning, evidence regarding its direct impact on quantitative academic performance indicators is conflicting and fragmented. This study aimed to map the scope, nature, and characteristics of existing research on the association between IA and academic performance among healthcare students.

Methods

This scoping review was conducted following the Arksey and O’Malley framework and the PRISMA-ScR guidelines. A systematic and comprehensive search was conducted in international and regional databases from inception to November 15, 2025. Out of 1,325 initial records, 15 eligible studies were selected for final analysis after screening.

Results

The findings indicate a concerning prevalence of IA (ranging from 6% to 84%) and a complex and predominantly indirect relationship with academic performance. Although a direct association with Grade Point Average was not statistically significant in all studies, IA was strongly correlated with impairments in learning process quality (surface learning approach, academic burnout), mental health (strong comorbidity with depression and anxiety), and even physical health. Identified risk factors included Type D personality, patterns of mobile phone multitasking, and negative academic experiences.

Conclusion

IA is a multidimensional phenomenon primarily affecting medical students’ cognitive and emotional capital. Effective management requires understanding its complex mechanisms, implementing evidence-based educational interventions, and creating supportive educational frameworks. Additionally, longitudinal and interventional studies are needed to better clarify outcomes and evaluate preventive strategies.

Background

The explosive growth of internet access and digital technologies has brought about a profound transformation in the realm of education and learning, particularly within higher medical education. Medical sciences students, as future architects of the healthcare system, are more dependent than ever on the internet for accessing up-to-date information, participating in virtual classes, interacting with instructors and peers, and conducting scientific research [1, 2]. While this unprecedented access provides countless learning opportunities, it simultaneously heightens the risk of problematic and addictive use of this virtual space [3, 4].

Internet addiction (IA), also referred to as problematic internet use (PIU) in some literature, describes a pattern of excessive, compulsive, and uncontrollable internet use that leads to clinically significant impairment or distress in personal, social, academic, or occupational functioning [5, 6].

It is important to clarify the diagnostic status of these terms. Neither IA nor problematic internet use is currently recognized as a formal diagnosis in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM-5-TR) [7] or the International Classification of Diseases, 11th Revision (ICD-11). The closest formal diagnostic entities are “Internet Gaming Disorder” (DSM-5-TR, Section III — condition for further study) and “Gaming Disorder” (ICD-11), which are specifically limited to online gaming activities and do not encompass the broader range of problematic online behaviors (e.g., social media use, compulsive web browsing, online shopping) examined in this review [8]. Therefore, throughout this manuscript, the term " IA " refers to a research-based construct operationalized by validated screening instruments (e.g., Young’s Internet Addiction Test [IAT] [9] or other validated measures), rather than a clinically diagnosed condition.

However, direct comparison of these prevalence rates is complicated by heterogeneity in measurement tools, different cut-off scores used to define IA, and varying cultural contexts across countries. These methodological and contextual factors should be considered when interpreting the wide range of reported prevalence estimates. In this review, IA is used as a broad umbrella term that encompasses ‘Problematic Internet Use’ (PIU) and related constructs. Although conceptual distinctions exist in some literature, the majority of studies and screening instruments do not reliably differentiate between them. Therefore, IA is adopted here as an inclusive terminology to capture the full spectrum of maladaptive internet use behaviors.

Among the student population, especially medical sciences students who face high academic loads, clinical pressures, and a critical need for time management, a concerning prevalence of IA has been reported. For example, a study in Bangladesh found that 84% of medical students were at risk of IA [10]. Similarly, 38.5% of pharmacy students in Egypt were identified as having potential IA [11]. These figures indicate that IA has become a major public health concern within medical education environments.

The academic performance of medical sciences students is a vital indicator for assessing educational quality and a predictor of their future professional success. This performance is typically measured by criteria such as Grade Point Average (GPA), course grades, and learning satisfaction [12, 13]. Maintaining optimal academic performance in these high-stakes fields, which are directly linked to human health, is of paramount importance.

However, despite a considerable volume of studies on IA among students, the research literature focusing specifically on medical sciences students is scattered and in need of integration regarding its scope, consistency of findings, and the identification of mediating and moderating factors. For instance, while some studies on dental [14, 15] and medical students [16] have reported a significant negative relationship between IA and GPA, others, such as a study on Colombian dental students, found no direct significant association but emphasized the important role of concurrent variables like study location and concurrent mobile phone use [14]. These discrepancies may stem from differences in assessment tools, cultural-educational contexts, or the presence of protective and risk factors.

Furthermore, with the onset of the COVID-19 pandemic and the unprecedented expansion of online education, medical students’ reliance on the internet for learning intensified dramatically. This sudden shift further blurred the lines between educational and non-educational internet use and likely exacerbated the prevalence and consequences of IA. Studies from this period suggest that stress related to online education can lead to worsening depressive symptoms and internet gaming disorder in adolescents and students by increasing academic stress and reducing social support [3, 17]. However, the specific impact of this period on the relationship between IA and academic performance in medical sciences students has not been well delineated.

Consequently, the existence of these knowledge gaps justifies the need for a systematic scoping review. Due to its exploratory nature and ability to map broad literature, this type of review is the most appropriate method for gathering, categorizing, and presenting a comprehensive picture of existing evidence, identifying common measurement tools, highlighting relational patterns, and pinpointing research gaps in this field. The primary objective of this scoping review is to map the association between IA and academic performance among healthcare students. Secondary objectives include identifying mediating or moderating variables (e.g., mental health, sleep quality, learning approaches, and academic burnout) that may explain or influence this relationship, as well as mapping the scope, nature, and characteristics of existing studies in this specific population. The findings of this review can serve as a basis for designing targeted educational-preventive interventions, informing policy in medical sciences faculties, and guiding future research utilizing longitudinal and interventional designs.

Methods

Study design

This study is a scoping review conducted to map the scope, nature, and characteristics of existing studies on the association between IA and academic performance among medical sciences students. The methodology was predefined and guided by the Arksey and O’Malley framework (2005) and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR) guidelines. No formal protocol was registered for this scoping review.

Stages of study implementation based on the Arksey and O’Malley’s framework

Stage 1: identifying the research question

In the first stage, the primary objective was to precisely define and narrow the study domain. Accordingly, the main research question focused on determining the nature, scope, and characteristics of existing studies concerning the association between IA and academic performance in medical sciences students. Subsidiary questions included: identifying the measurement tools used to assess IA and academic performance; examining the types of research designs employed in this field; extracting key findings related to the direction, strength, and mediating or moderating factors of this association; and finally, identifying existing gaps in the research literature.

Stage 2: identifying relevant studies

To achieve maximum coverage of relevant studies, a systematic and comprehensive search was conducted in international and regional electronic databases. The search encompassed databases such as PubMed/MEDLINE, Scopus, Web of Science, PsycINFO, EMBASE, as well as specialized Iranian medical sciences databases including SID and Magiran. A systematic and comprehensive search was conducted in international and regional databases from inception to November 15, 2025. Keywords were formulated using a combination of MeSH/Emtree terms and free-text keywords covering concepts such as “Internet Addiction,” “Problematic Internet Use,” “Internet Gaming Disorder” alongside “Academic Performance,” “Educational Status,” “Grade Point Average,” and “Medical Sciences Students” (Table 1).

Table 1.

Example of search keywords and MeSH terms for MEDLINE (PubMed)

Component	Search Terms
Population	MeSH Terms: Students, Medical OR Students, Health Occupations Keywords: "medical student” OR “healthcare student” OR “health professional student” OR “dental student” OR “nursing student” OR “pharmacy student” OR “medical undergraduates” OR “midwifery student” OR “public health student” OR “allied health student”
Concept (1) Internet Addiction	MeSH Term: Internet Addiction Disorder Keywords: "Internet Addiction Disorder” OR “Problematic Internet Use” OR “Compulsive Internet Use” OR “Internet Gaming Disorder” OR “Online Addiction”
Concept (2) Academic Performance	MeSH Terms: Academic Performance OR Educational Status Keywords: "Academic Performance” OR “Educational Status” OR “Grade Point Average” OR “GPA” OR “Academic Achievement” OR “Academic Success” OR “Learning” OR “academic failure”

A comprehensive search strategy was developed using a combination of MeSH terms and free-text keywords with Boolean operators (AND, OR). The full search strategy for PubMed is presented below:

("Internet Addiction Disorder"[MeSH] OR "Internet addiction" OR "Problematic Internet Use" OR "Compulsive Internet Use" OR "Internet Gaming Disorder")AND("Academic Performance"[MeSH] OR "Educational Status" OR "Grade Point Average" OR "GPA" OR "Academic Achievement")AND("Students, Medical"[MeSH] OR "healthcare students" OR "nursing students" OR "medical students" OR "pharmacy students").

The search strategy was adapted for other databases. In addition, a supplementary manual search was performed within the reference lists of relevant review articles and the selected primary articles. In total, 1,325 records were identified through database searching and manual search methods.

Stage 3: study selection

Eligibility criteria for study inclusion were developed using the Population, Concept, Context (PCC) framework (Table 2).

Table 2.

Eligibility criteria for scoping review

Component	Description
Population	Undergraduate and postgraduate students enrolled in healthcare/medical science fields (including but not limited to medicine, dentistry, nursing, pharmacy, and allied health professions). Excluded: Students in non-healthcare fields, school students, teachers, general population, and healthcare professionals who are no longer students.
Concept	The core concepts of this review are: 1. Internet Addiction: Problematic, excessive, or compulsive use of the internet, often measured by validated scales (e.g., IAT, PIUQ). 2. Academic Performance (AP): Quantifiable educational outcomes, including but not limited to GPA, exam scores, course grades, academic progress, and rates of failure or dropout. The review focuses on examining the nature, strength, and direction of the relationship between these two concepts.
Context	The review considers observational (cohort, case-control, cross-sectional) and interventional studies conducted in any higher education setting (universities, medical schools, teaching hospitals) worldwide, with no restriction on publication year. The primary language of publication is restricted to English or Persian.

Stage 3: study selection

The screening and selection process was conducted in three stages, following the PRISMA guidelines.

1. Removal of Duplicates: A total of 1,325 records were identified (1,321 through database searching and 4 through other sources). After removing 650 duplicate records using EndNote X9 and manual screening, 675 records remained for title and abstract screening.

2. Title and Abstract Screening: Two researchers independently screened the 675 articles based on the inclusion and exclusion criteria. In this process, 630 articles were excluded due to: mismatch in the study population (e.g., focusing on school students, teachers, or the general population); not investigating both core concepts of the study (addressing only IA or only academic performance); and irrelevant study design (including theoretical reviews or purely qualitative research). After resolving discrepancies through discussion and consensus, 45 articles were deemed eligible for full-text assessment.

3. Full-Text Screening: In the full-text screening stage, two researchers independently evaluated the full text of the remaining 45 articles. During this process, 30 articles were excluded for the following reasons: not reporting academic performance indicators quantitatively; combining medical sciences students with other disciplines without the possibility of data separation; unavailability of the full text; and reporting results in a language other than English or Persian. Ultimately 15 articles met all inclusion criteria and were selected for data extraction.

The identification, screening, and selection process is visually summarized in the PRISMA flow diagram (Fig. 1), which clearly illustrates the stages and the number of studies at each stage.

Fig. 1.

PRISMA flow diagram of the study selection process

Consistent with scoping review methodology, no formal critical appraisal of included studies was conducted.

Stage 4: charting the data

In the fourth stage, key data from the 15 selected articles were collated using a standardized data extraction form. This process was conducted independently by two researchers, with any discrepancies resolved through discussion and final consensus.

The extracted data included:

• General Study Characteristics: Authors, publication year, country where the study was conducted, study aim, and research design.

• Key Findings Statistical indicators related to variable relationships, such as correlation coefficients, odds ratios, and mean comparisons; level of statistical significance; and identified mediating or moderating factors.

Stage 5: collating, summarizing, and reporting the results

In the fifth stage, the extracted data were summarized and analyzed qualitatively using descriptive statistics. The findings were organized according to the main axes of the research questions and reported in the form of tables, charts, and narrative descriptions. During this stage, study characteristics such as publication year, country of origin, study design, and specialized fields were described. Patterns of association regarding the direction and strength of the relationship between IA and various academic performance indicators were synthesized and explained. Finally, relevant factors, including mediating and moderating variables, as well as research gaps identified from the analysis of the studies, were extracted and presented.

Results

This scoping review of 15 studies presents key findings on the relationship between IA and academic performance among health sciences students (Table 3). The results are organized and reported around the following major themes.

Table 3.

Characteristics and key findings of included studies in the scoping review (n = 15)

Lead Author, Year	Country	Primary Objective	Key Methodological Details	Key Findings Relevant to the Review Topic
Afrin, 2024 [10]	Bangladesh	To determine the prevalence of IA and its impact on academic performance in medical students.	Design: Cross-sectional Instrument: YIAT (Young Internet Addiction Test) Cut-off: Not reported Sample: Third-year medical students (n = 312, 51% male, mean age 21.05 ± 0.83 years, 90.1% Bangladeshi, 77.2% living in hostel) Academic measures: (1) Exam result (2) Class attendance	The prevalence of IA was 84%. A highly significant association was found between the level of IA and first professional examination results and class attendance.
Fidanci, 2024 [18]	Turkey	To evaluate the impact of IA on academic success and grit in medical students.	Design: Cross-sectional (descriptive) Instrument: YIAT -Short Form (12 items) Cut-off: Not reported (scale score range reported: 12–60, but no addiction categorization provided) Sample: n = 458, university students, mean age 21.9 years, 50.2% male, 49.8% female (230 male, 228 female), Academic measures: GPA (out of 100, range 63–86)	No significant difference in GPA or grit score was found among groups with different levels of IA. A very weak positive relationship existed between GPA and grit.
Khatony, 2024 [19]	Iran	To develop a model of IA based on academic performance, academic experience, and clinical self-efficacy in nursing students.	Design: Correlational / Path analysis Instrument: YIAT Cut-off: ≤20 (no addiction), 21–49 (mild), 50–79 (moderate), 80–100 (severe) Sample: n = 340, nursing students, 55.6% female, 44.4% male (189 female, 151 male) Academic measures: (1) GPA (2) Academic term (academic experience)	Significant negative correlations were found between clinical self-efficacy, academic experience, and academic performance with IA. Academic experience had a significant negative impact on IA.
Sadeghi Bimorgh, 2023 [20]	Iran	To investigate the relationship between mobile phone addiction and IA with academic burnout in medical students.	Design: Cross-sectional (descriptive-analytical) Instrument: YIAT − 20 items, MP IA (cell phone), MBI (burnout) Cut-off: Not reported Sample: n = 190, medical students (basic sciences, physiopathology, externship, internship), mean age 22.2 years, 49.5% male, 50.5% female (94 male, 96 female), Academic measures: Academic burnout (MBI: exhaustion, cynicism, inefficacy)	Internet use could increase academic burnout, but mobile phone addiction had no significant effect. The mean IA score was lower among clinical trainees (interns).
Mrayyan, 2023 [21]	Jordan	To assess cyberchondria, IA, and anxiety-related symptoms among nursing students during COVID-19.	Design: Cross-sectional (web survey) Instrument: YIAT Cut-off: 0–19 (no addiction), 20–39 (low), 40–69 (moderate), 70–100 (severe) Sample: n = 333, nursing students, 68.8% female, 31.2% male (229 female, 104 male) Academic measures: GPA	Students had moderate levels of cyberchondria and IA. Demographic characteristics, IA, and anxiety symptoms were significantly interrelated. GPA, age, and educational level were significant predictors of IA.
Labrague, 2023 [22]	USA	To examine the association between problematic Internet use and psychological distress in nursing students and the mediating role of coping skills.	Instrument: PIUQ Cut-off: Not reported Sample: n = 279, student nurses, 78.5% female, 21.5% male (219 female, 60 male) Academic measures: Academic performance (fair, average, good, very good) – self-reported	Problematic Internet use was significantly associated with average academic performance. Students with lower academic performance reported higher levels of problematic Internet use.
Bahrami, 2022 [23]	Iran	To investigate the effect of IA on academic performance and musculoskeletal disorders in health sciences students.	Design: Cross-sectional Instrument: YIAT, Pham & Taylor Academic Performance Questionnaire, Nordic Musculoskeletal Questionnaire Cut-off (IAT): Not reported Cut-off (Academic): <120 (weak), 121–174 (moderate), > 175 (strong) Sample: n = 400, university students, mean age 21.3 ± 2.04 years, 28.2% female, 71.8% male (113 female, 287 male) Academic measures: Pham & Taylor Questionnaire	No significant relationship was found between different levels of IA and academic achievement. IA was associated with discomfort in the neck, shoulder, and wrist.
Sayed, 2022 [11]	Egypt	To assess IA and its association with common mental disorders and academic performance in pharmacy students.	Design: Cross-sectional Instrument: YIAT Cut-off: 0–30 (normal), 50–100 (potential IA) Sample: n = 808, pharmacy students, mean age 21.16 years, 73.5% female, 26.5% male (594 female, 214 male) Academic measures: GPA	The prevalence of potential IA was 38.5%. No correlation was observed between IA and GPA. However, IA was strongly correlated with depression, anxiety, and stress.
Elbilgahy, 2021 [1]	Egypt/Saudi Arabia	To investigate the prevalence of Internet and electronic device addiction and its impact on sleep and academic performance in nursing students.	Design: Cross-sectional comparative Instrument: YIAT, MPIQ, ESS Cut-off (IAT): 20–49 (average user), 50–79 (occasional/frequent problems), 80–100 (significant problems) Sample: n = 767, all female nursing students (514 Egypt, 253 Saudi) – 100% female Academic measures: GPA	Severe IA was present in 42.7% of Saudi and 10.3% of Egyptian participants. IA was associated with poor academic performance and excessive daytime sleepiness.
Javaeed, 2020 [16]	Pakistan	To evaluate the relationship between IA and academic performance in medical students.	Design: Cross-sectional Instrument: YIAT Cut-off: <20 (no addiction), 20–49 (mild), 50–79 (moderate), 80–100 (severe) Sample: n = 316, medical students (years 2–5), median age 22 years, 59.2% female, 40.8% male (187 female, 129 male) Academic measures: Professional exam results (percentage ranges: 50–60%, 61–70%, 71–80%, > 80%)	28.2% of students had ‘severe addiction’. Students addicted to the Internet performed significantly worse in their examinations.
Díaz Cárdenas, 2019 [14]	Colombia	To determine the association between IA and academic performance in dental students.	Design: Cross-sectional Instrument: YIAT Cut-off: 20–49 (mild), 50–79 (moderate), 80–100 (severe) Sample: n = 402, dental students (semesters 2–10), mean age 20.5 ± 1.1 years, 67.6% female, 32.3% male (272 female, 130 male) Academic measures: Cumulative GPA	No direct significant association was found between IA and poor academic performance. However, factors associated with IA included studying outside home, chatting on the mobile phone, and using the phone for more than 18 min while studying.
Kumar, 2018 [15]	India	To assess the prevalence of IA and its relationship with depression and academic performance in dental students.	Design: Cross-sectional Instrument: YIAT (Young), BDI-1 (Beck’s Depression Inventory) Cut-off (IAT): ≥50 indicates IA Cut-off (BDI): ≥10 indicates depression Sample: n = 384, dental students (1st year to internship), mean age 21.19 ± 1.93 years, 87.1% female, 12.9% male (304 female, 45 male) Academic measures: Marks percentage (≤ 60% = average/regular, > 60% = good)	The prevalence of IA was 6%. Individuals with depression (OR = 6) and those with scores below 60% (OR = 6.71) were more likely to have IA.
Loredo Silva, 2018 [24]	Brazil	To evaluate smartphone use, IA, and its association with surface and deep learning approaches in medical students.	Design: Cross-sectional Instrument: YIAT Cut-off: Not reported (only classified as “problematic users” 68.2%) Sample: n = 710, medical students (all 6 years) Academic measures: Learning approaches (surface/deep - Biggs questionnaire)	68.2% were problematic Internet users. Frequent smartphone use and IA were associated with a higher level of surface learning (learning for exams) and a lower level of deep learning (conceptual understanding).
Poormirzaei, 2017 [25]	Iran	To predict IA based on Type D personality and humor styles and to examine its relationship with academic performance.	Design: Cross-sectional (descriptive correlational) Instrument: YIAT, Type D Personality (Denollet), Humor Styles (Martin) Cut-off: ≥50 = addicted group ( score 50–79 = dependent group, > 80 = addicted group) Sample: n = 225, medical students, mean age 21.49 ± 2.02 years, 71.6% female, 28.4% male (161 female, 64 male), Academic measures: GPA	Negative affectivity and self-defeating humor style predicted 23% of the variance in IA. IA was correlated with decreased academic performance.
Boonvisudhi, 2017 [26]	Thailand	To study the level of IA and its association with depression in medical students.	Design: Cross-sectional Instrument: YIAT, PHQ-9 (depression) Cut-off (YDQ): ≥5 = “possible IA” (from 0–8 scale) Cut-off (PHQ-9): ≥9 = depression Sample: n = 705, medical students (years 1–5), mean age 20.51 ± 1.91 years, gender distribution: 389 female (55.2%), 316 male (44.8%) Academic measures: GPA, academic problems (self-reported)	24.4% had probable IA and 28.8% had depression. A significant association existed between probable IA and depression. An association was found between academic problems and IA.

Heterogeneity in the prevalence of IA

The prevalence of IA among health sciences students varied considerably across studies, ranging from concerning to very high levels. In one sample, 84% of Bangladeshi medical students reported some degree of IA, indicating an epidemic within that educational setting [10]. In contrast, other studies reported lower prevalences of approximately 6% among Indian dental students [15] and 24.4% among Thai medical students [26].

A complex, non-linear relationship with academic performance

The findings of this review indicate that the relationship between IA and academic performance is not consistently direct but may be indirect and influenced by multiple factors. Most included studies were cross-sectional; therefore, causal direction cannot be determined.

Examination of existing research reveals a paradoxical pattern. Some studies reported a significant negative association between IA and academic performance. For instance, Afrin found that 84% of Bangladeshi medical students had IA, which was significantly associated with lower examination results and class attendance [10]. Javaeed reported that 28.2% of Pakistani medical students had severe IA, and these students performed significantly worse in their examinations [16]. Elbilgahy found that severe IA was present in 42.7% of Saudi and 10.3% of Egyptian nursing students, and IA was associated with poorer GPA [1]. Poormirzaei reported a correlation between IA and decreased GPA among Iranian medical students [25].

In contrast, several other studies failed to find a direct significant association between IA scores and GPA. Díaz Cárdenas found no direct association between IA and cumulative GPA among Colombian dental students [14]. Fidanci reported no significant difference in GPA across different levels of IA among Turkish medical students [18]. Sayed observed no correlation between IA and GPA among Egyptian pharmacy students (38.5% potential IA prevalence) [11]. Bahrami found no significant relationship between IA and academic achievement among Iranian health sciences students [23].

These contradictory findings suggest that the association between IA and academic performance may be indirect, potentially mediated by factors such as learning approaches, academic burnout, sleep quality, and mental health.

Strong impact on mediating learning processes and outcomes

Although a direct association with GPA was not found in all studies, the reviewed evidence suggests that IA may be associated with poorer quality of learning processes and academic well-being indicators. These should be interpreted as potential mediating pathways rather than equivalent outcomes to academic performance.

• Impairment of Learning Approaches: Emerging evidence suggests IA may be linked to deficits in executive functions, sustained attention, and working memory, all crucial for learning complex medical concepts [27]. Among Brazilian medical students, problematic Internet and smartphone use were associated with a higher level of surface learning (learning for tests) and a lower level of deep learning (conceptual understanding) [24].

• Increased Risk of Academic Burnout: Prolonged and compulsive Internet use can lead to emotional exhaustion, cynicism, and reduced personal efficacy in studies. Academic burnout itself can directly increase the risk of IA and create a vicious cycle, potentially mediated by reduced self-control and increased peer rejection [28, 29]. Multiple studies highlight the role of IA as a significant risk factor for academic burnout. For instance, among Iranian medical students, Internet use was a significant predictor of increased academic burnout [20]. Notably, the same study found mobile phone addiction was not necessarily linked to burnout, indicating the importance of digital content and context.

• Negative Impact on Academic Engagement: The model proposed for Iranian nursing students suggests that academic experience (which can include commitment and engagement) has a significant negative impact on the level of IA, indicating that students with a more positive connection to their academic environment are less likely to fall into addiction [19].

• Sleep Quality Disruption: Internet use, particularly via mobile phones late at night, is associated with impaired sleep quality, duration, and regularity, which in turn negatively affects concentration, memory, and learning capacity during the day [30].

Strong comorbidity with mental health problems

The strongest and most consistent finding across studies was the association between IA and mental health conditions, particularly depression. While some studies suggest the possibility of a bidirectional relationship, cross-sectional designs cannot establish causal direction. These mental health variables should be interpreted as potential mediators in the relationship between IA and academic performance. Among Indian dental students, individuals with depression were six times more likely to have IA [15]. Among Thai medical students, the likelihood of depression was 1.58 times higher in the group with probable IA [26]. Among Egyptian pharmacy students, IA was strongly correlated with higher levels of depression, anxiety, and stress [11]. This strong, bidirectional relationship with depression [15, 26] suggests IA may be both a cause and a consequence of psychological distress. Students may retreat into the online space to escape academic stressors or negative emotions (the “self-escape” model), thereby perpetuating a vicious cycle of avoidance and diminished performance. This perspective explains the apparent paradox in GPA-related findings; the negative impact may not be direct but exerted through the deterioration of mental health and learning quality, which may ultimately (though not always immediately reflected in GPA) affect final outcomes.

Identification of personality and behavioral risk factors

The reviewed studies indicate that certain individual factors and behavioral patterns render health sciences students more vulnerable to IA. Among Iranian medical students, Type D personality (a research-based personality construct not classified in DSM or ICD, characterized by negative affectivity and social inhibition) with its traits of negative affectivity and social inhibition, combined with a self-defeating humor style, collectively predicted approximately 23% of the variance in IA [25]. Furthermore, among Colombian dental students, problematic mobile phone use patterns, including chatting while studying and spending over 18 min answering phone calls during study time, were significantly associated with higher levels of IA [14].

Consequences extending beyond the academic realm

IA is also associated with adverse physical health outcomes. A study on Iranian health sciences students showed IA was linked to increased reports of discomfort in the neck, shoulder, and wrist areas, likely due to poor ergonomics and prolonged use [23]. Additionally, among Egyptian and Saudi nursing students, IA was associated with excessive daytime sleepiness [1].

Comparison across disciplines, populations, and contexts

Cross-study comparisons reveal interesting differences across disciplines and geo-cultural contexts. Regarding academic discipline, although IA was reported across all health sciences fields, some studies suggest differences in patterns. For example, the mean IA score was lower among medical interns (senior students) compared to basic sciences students [20]. This difference may be related to factors such as increased clinical responsibilities or different patterns of internet access during rotations. However, this interpretation should be made with caution, as it remains speculative and requires further investigation.

Culturally, the very high prevalence reported in Bangladesh [10] versus moderate rates in Iran or Turkey may point to the role of socioeconomic factors, access to alternative leisure activities, or different cultural norms in technology use. Furthermore, the Colombian study’s emphasis on smartphone distraction during study as a key risk factor [14] highlights the issue of “media multitasking,” perhaps relevant in educational environments with high in-class engagement. In contrast, other studies focused on intra-individual factors like Type D personality [25].

Methodology of the reviewed studies

The predominant research design across all 15 reviewed studies was cross-sectional. Regarding measurement tools, the vast majority of studies utilized Young’s Internet Addiction Test (IAT) or its short form. Furthermore, the assessment of academic performance was largely limited to GPA or self-reported grades.

Knowledge gaps

This review reveals several critical gaps in the existing knowledge base. The absence of longitudinal or cohort studies capable of explaining temporal sequences and inferring causal relationships was identified as the most prominent research gap. Furthermore, no interventional study was found that evaluated the effectiveness of strategies such as digital time management training or online cognitive-behavioral interventions in reducing IA and improving academic performance among health sciences students. Additionally, the role of mediating and moderating variables—including social support, academic self-efficacy, teaching quality, and access to university resources—has not been systematically investigated. Moreover, the focus of most research has been on “general Internet addiction,” with very limited specific investigation of phenomena like “Internet gaming disorder” within this population. Finally, all studies emphasized the negative consequences of Internet use, neglecting to explore its potential positive capacities for balanced use and enhancement of the learning process.

Discussion

The findings of this scoping review indicate that the relationship between IA and academic performance among health sciences students is complex, multifaceted, and predominantly indirect. While a direct, negative correlation with increased Internet use was anticipated, emerging evidence suggests that adverse effects often manifest through the erosion of learning processes, mental health, and academic engagement.

Prevalence heterogeneity and contextual factors

The substantial heterogeneity in reported prevalence rates, ranging from 6% to 84% [10, 15], underscores the significant influence of contextual factors. These variations likely stem from differences in cultural norms surrounding Internet use, access to alternative recreational facilities, and diversity in screening tools and diagnostic cut-offs. The alarmingly high prevalence in certain settings (e.g., 84% in Bangladesh [10]) serves as a serious warning, indicating that IA can become a normative unhealthy behavior in some medical colleges. A global meta-analysis of general (2025) university students reported a prevalence of 41.84% [31], which is lower than some medical student studies like Afrin [10], yet still represents an epidemic-level concern.

In addition to these contextual differences, it is important to consider the broader evolution of the digital environment over time. The rapid expansion of digital accessibility, widespread smartphone ownership, and the emergence of algorithm-driven platforms (e.g., social media and personalized content delivery systems) may have contributed to increasing engagement and potentially more compulsive patterns of Internet use among students. These platforms are intentionally designed to maximize user engagement through continuous feedback and reinforcement mechanisms, which may increase the risk of habitual and excessive use [32, 33].

However, the available evidence in this review is primarily based on cross-sectional studies conducted at different time points and in diverse settings. Therefore, it is not possible to determine whether the prevalence or severity of Internet addiction has systematically increased over time. The lack of longitudinal and trend-based studies represents a critical gap, and future research is needed to examine temporal patterns and better understand how evolving digital ecosystems influence Internet use behaviors among healthcare students.

Inconsistent direct association between IA and academic performance

Regarding its association with academic performance, the findings of this review are inconsistent. While some studies reported a direct decline in GPA or examination results [1, 10, 16, 25], others failed to find such a direct association [11, 14, 18, 23]. This apparent paradox can be explained by considering the indirect and mediating effects of IA on the learning process. The pattern observed in our review is not unique to health sciences and aligns significantly with research in non-medical populations. External evidence beyond the scope of this review aligns with this finding. For instance, among university students in Peru, IA was directly identified as a strong predictor of academic procrastination. In that study, “academic self-efficacy” emerged as a crucial mediating variable, whereby IA, by diminishing students’ confidence in completing tasks, leads to delays in coursework and consequently reduced performance quality [34]. This finding reinforces the theoretical framework of the results from Khatony (2024) concerning the impairment of clinical self-efficacy [19], suggesting that underlying psychological mechanisms are shared across different student groups.

Indirect pathways: mediating and moderating variables

In this review, a key finding of this review is that the association between IA and academic performance among healthcare students is largely indirect. Rather than a direct effect on grade-based indicators such as GPA, IA appears to be associated with a range of intermediary factors that may ultimately influence academic outcomes. In this section, we distinguish between mediating variables (factors that may explain how or why IA could affect academic performance) and moderating variables) factors that may influence the strength or direction of this association (.

Potential mediators

Several variables emerged as potential mediators across the included studies. First, mental health conditions, particularly depression and anxiety, showed consistent associations with both IA and academic impairment [11, 15, 26]. While cross-sectional designs preclude causal inference, it is plausible that students with higher levels of IA experience greater psychological distress, which in turn may be associated with reduced study motivation, concentration, and academic engagement. Second, academic burnout — characterized by emotional exhaustion and reduced self-efficacy — was associated with IA in multiple studies [20, 28, 29], suggesting a possible pattern whereby compulsive internet use is correlated with burnout, which may further correlate with lower academic performance. Third, poor sleep quality was identified as another candidate mediator, as excessive internet use, especially at night, may disrupt sleep patterns and subsequently be associated with impaired daytime cognitive functioning [1, 30]. Fourth, surface learning approaches (learning primarily for test performance rather than conceptual understanding) were associated with problematic internet use, indicating that IA may be correlated with less effective learning strategies [24].

The potential mediating role of psychological factors is further supported by evidence from non-healthcare student populations. Comparable findings have been reported outside the healthcare student population. For instance, among Turkish university students, loneliness and rumination serially mediated the relationship between social anxiety and IA [35]. This parallel finding suggests that the indirect pathways identified in our review may reflect broader psychological mechanisms common across diverse student groups, rather than phenomena unique to healthcare education.

Potential moderators

Fewer studies examined moderating effects. However, Type D personality (negative affectivity and social inhibition) was identified as a potential moderator in one study, predicting a substantial proportion of variance in IA [25]. Academic self-efficacy and coping skills were also suggested as factors that may buffer the negative associations of IA [19, 24], though evidence remains limited. Additionally, behavioral patterns such as mobile phone multitasking during study time appeared to be associated with higher levels of IA and poorer academic outcomes [14].

Overall, the evidence suggests that the relationship between Internet addiction and academic performance is largely mediated and moderated by a set of interrelated psychosocial factors. Key mediators include mental health variables (e.g., depression and anxiety), academic burnout, sleep disruption, and maladaptive learning approaches, all of which may impair cognitive functioning and academic engagement. In addition, factors such as academic self-efficacy, coping strategies, and perceived social support may act as moderators, influencing the strength and direction of this relationship. Although these variables are consistently reported across studies, their exact causal roles remain uncertain due to the predominance of cross-sectional designs.

It is important to emphasize that due to the predominance of cross-sectional designs, these variables should be interpreted as correlates and potential pathways rather than confirmed causal mechanisms. Longitudinal research is needed to test the directionality of these relationships.

Proposed pathway: IA, self-care disruption, and academic impairment

Although causal inference is limited by cross-sectional designs, a plausible pathway can be proposed: IA may impair academic performance indirectly by disrupting core self-care behaviors. First, late-night internet use reduces sleep quality and duration, which impairs attention and memory consolidation [1, 30]. Second, excessive online time erodes structured daily routines (e.g., irregular study schedules, missed classes) [10, 14]. Third, prolonged sedentary screen use often displaces physical activity, contributing to fatigue and low mood [23]. These three pathways are likely interactive, potentially creating a vicious cycle of worsening self-care and academic decline. However, this model requires testing through longitudinal mediation studies.

Knowledge gaps and future research directions

This scoping review identified several key gaps in the literature on internet addiction and academic performance among health sciences students. The evidence is largely based on cross-sectional studies, which limits the ability to determine causal or temporal relationships. Therefore, it remains unclear whether internet addiction leads to poorer academic performance or whether academic difficulties increase vulnerability to problematic internet use.

In addition, although some studies report direct associations between internet addiction and GPA, findings from this review suggest that the relationship is likely indirect. However, potential mediating factors such as mental health problems (e.g., depression, anxiety, and burnout), learning approaches, and sleep quality have rarely been examined in integrated or longitudinal models.

Another major gap is the absence of interventional studies targeting internet addiction in health sciences students. No experimental studies were found evaluating strategies such as behavioral or educational interventions to reduce problematic internet use or improve academic outcomes.

Furthermore, individual factors such as personality traits and behavioral patterns (e.g., mobile phone multitasking during studying) have been understudied in comprehensive predictive models. Academic performance has also been narrowly assessed, mainly using GPA, while broader outcomes such as clinical competence and learning engagement have been largely ignored.

Future research should therefore focus on longitudinal designs to clarify causal pathways and examine mediating mechanisms over time. Interventional studies are also needed to test the effectiveness of educational and behavioral strategies for reducing internet addiction. In addition, future studies should develop integrated models combining psychological, behavioral, and contextual factors, and adopt multidimensional measures of academic performance beyond GPA. Finally, qualitative and mixed-methods research is recommended to better understand students’ experiences and contextual determinants of internet use behaviors.

Practical/policy implications

The findings of this review highlight the need for a comprehensive, multi-level approach to address Internet addiction among health sciences students. Interventions should extend beyond general awareness and incorporate structured, evidence-informed strategies at the individual, peer, and institutional levels.

At the institutional level, universities are encouraged to develop and implement clear digital wellbeing policies that promote balanced and purposeful use of technology within academic environments. Such policies may include guidelines on appropriate use of digital devices during learning activities, recommendations for screen time management, and integration of digital wellbeing principles into academic regulations and student support frameworks.

At the educational level, incorporating structured psychoeducational programs into the curriculum can play a critical role. These programs may focus on self-regulation skills, time management, healthy coping strategies, and critical awareness of online behaviors. Workshops and short training modules on digital literacy and responsible Internet use can further enhance students’ ability to manage their online engagement effectively.

Peer-led initiatives represent another promising strategy. Training student mentors or peer facilitators to lead group discussions and share practical strategies for managing Internet use may increase acceptability and engagement, particularly among younger students. Such approaches can foster a supportive social environment and reduce stigma associated with seeking help.

At the service level, universities should strengthen student counseling systems by integrating routine screening for problematic Internet use and related mental health conditions. Providing brief behavioral interventions, counseling services, and, where necessary, referral pathways can facilitate early identification and timely support for at-risk students.

In addition, the use of digital tools themselves can be leveraged as part of the solution. For example, app-based monitoring systems that provide feedback on screen time, as well as voluntary “digital detox” campaigns, may help students develop greater awareness and control over their online behaviors.

Overall, combining policy development, educational interventions, peer support, and accessible services may offer a more sustainable and context-sensitive approach to reducing problematic Internet use in academic settings.

Strengths of the scoping review

This scoping review, adhering to the Arksey and O’Malley methodological framework and the PRISMA-ScR guidelines, benefits from a high level of methodological transparency and coherence. Conducting a systematic search in international and regional databases provided relatively comprehensive coverage of the existing scientific literature. The focus on the specific population of health sciences students, rather than students in general, resulted in the extraction of more targeted findings with higher applicability in health sciences education. Moreover, the conscious attention to existing contradictions in study results, including those concerning GPA, and their analysis instead of dismissal, contributed to a deeper understanding of the complexities of the subject.

Limitations of the review

Several limitations should be considered when interpreting the findings of this scoping review. First, the search was limited to articles published in English and Persian; therefore, relevant studies published in other languages may have been missed. Second, the predominance of cross-sectional designs among the included studies precludes causal inference regarding the relationship between IA and academic performance. Third, significant heterogeneity in the definitions and measurement tools used to assess IA and academic performance (primarily GPA) makes quantitative synthesis, including meta-analysis, difficult and inappropriate. Fourth, publication bias is possible, as studies with negative or non-significant findings may be less likely to be published and thus underrepresented in this review. Fifth, consistent with standard scoping review methodology (Arksey & O’Malley, 2005; PRISMA-ScR), we did not conduct a formal risk of bias or quality assessment of the included studies. Although this is not a requirement for scoping reviews — which aim to map the existing literature rather than critically appraise evidence quality — we acknowledge this as a limitation, as studies with methodological weaknesses were not excluded from synthesis. Finally, we acknowledge that relying primarily on GPA — as most included studies did — oversimplifies academic performance in health sciences education. This construct also includes clinical competencies, professional behaviors, academic engagement, and learning quality (e.g., deep vs. surface learning). The absence of such multidimensional measures in the reviewed literature may limit a full understanding of IA’s impact. Future studies should incorporate broader assessments (e.g., OSCEs, clinical competence ratings, engagement scales) beyond GPA alone.

Conclusion

This scoping review confirms that IA is a prevalent and multifaceted health and educational challenge among health sciences students. Although its relationship with traditional grade-based indicators (such as GPA) may not be direct and uniform, it has a strong and documented negative impact on the quality of the learning process (by promoting surface approaches and increasing burnout), mental health (particularly through comorbidity with depression), and even physical health. The relationship between these two constructs is not linear and simple but is complex, reciprocal, and mediated by numerous individual and environmental factors. An effective response to this challenge requires moving beyond a purely statistical perspective toward a deeper understanding of psychological dynamics, designing evidence-based interventions, and creating supportive changes in educational environments. Investment in future longitudinal and qualitative research is essential to complete this roadmap.

Abbreviations

IA

Internet addiction

GPA

Grade Point Average

YIAT

Young Internet Addiction Test

PRISMA-ScR

Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews

Authors' contributions

HK: Conceptualization, Methodology, Writing- Original draft preparation, Data curation, Investigation. NB: Conceptualization, Methodology, Data curation, Investigation, Writing- Original draft preparation.

Funding

The authors declare that no financial support was received for the research reported in this article.

Data availability

Data sharing not applicable as no datasets generated and/or analysed for this study.

Declarations

Ethics approval and consent to participate

This study was a review of previously published literature and did not require ethical approval.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.