Study design, setting and population

This study employed a cross-sectional descriptive and analytical design. It was conducted between January and March 2024. The reporting of this study followed the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines, and the corresponding STROBE checklist is provided as Supporting Information (S1 Checklist).

The research was conducted at Faculties of Health Sciences at three universities in Turkey—one public and two private. Within these faculties, the departments of Nursing, Midwifery, Nutrition and Dietetics, Physiotherapy and Rehabilitation, Child Development, Speech and Language Therapy, Occupational Therapy, and Audiology were included in the study scope. The participants were undergraduate students enrolled in first through fourth academic years across these departments.

Approximately 4,300 undergraduate students were enrolled at the participating faculties at the time of the study. Among them, 78% were female and 22% were male. The study population consisted of students who were accessible during the data collection period and who voluntarily agreed to participate in the research.

Eligibility criteria included being 18 years of age or older, being enrolled in a Faculty of Health Sciences, and providing voluntary consent to participate. Students under 18 years of age or those with visual impairments that prevented survey participation were excluded.

Although the total number of students enrolled in the participating faculties was approximately 4,300, this number was an estimate and not a precisely defined population size. Therefore, the sample size was calculated using the standard formula for an unknown (infinite) population: n=[Z²xp(1-p)]/e² where Z = 1.96 (for a 95% confidence level), p = 0.5 (maximum variability), and e = 0.05 (margin of error). Based on this calculation, a minimum of 384 participants was required. The final sample of 896 participants exceeded this threshold, ensuring sufficient statistical power and representativeness for inferential analysis.

Sampling procedure

A non-probability, convenience sampling method was used to recruit participants. This method was chosen due to its practicality and feasibility in accessing a large number of university students during a limited data collection period, especially in an online survey.

Study variables and measurement tools

To ensure data quality, standardized procedures were followed throughout the data collection process. Informed consent was obtained digitally from all participants. Data were collected using previously validated instruments. Although no pilot study was conducted, all measurement tools were based on validated sources. Continuous monitoring of data entry was performed to minimize errors and maintain accuracy. Participant anonymity and confidentiality were strictly preserved. Data were collected online via a Google Form using a structured questionnaire composed of three sections (see Supporting Information S2 Survey Instrument). All participants who provided consent first completed the personal and demographic items, followed by the CAM and CAM-MYCS scales, which were grouped under the heading “Cancer Awareness.” The section on health and lifestyle behaviors was administered after these scales. Personal Information Form collected both socio-demographic information—including age, gender, marital status, and income level—and descriptive characteristics, such as academic department, year of study, prior coursework in oncology, and whether the participant had a family member diagnosed with cancer.

CAM, an 11-item, 5-point Likert-type scale assessing awareness of scientifically validated cancer risk factors such as tobacco use, obesity, alcohol consumption, physical inactivity, and excessive sun exposure. In line with the original scoring method, responses of “Strongly Agree” and “Agree” were coded as correct (1 point), while other responses were coded as incorrect (0 points). The total score ranged from 0 to 11, with higher scores indicating greater awareness. As no cut-off points have been established for the CAM scale in the literature, total scores were treated as continuous variables. In the current study, the scale demonstrated acceptable internal consistency with a Cronbach’s alpha of 0.732, comparable to the original study (α = 0.77) [6]. CAM-MYCS, a 12-item, 5-point Likert scale developed to assess belief in commonly held but scientifically unsupported cancer myths—such as the use of microwave ovens, stress, genetically modified foods, and mobile phones. According to the validated scoring system, “Strongly Disagree” and “Disagree” were scored as correct (1 point), and all other responses were scored as incorrect (0 points), yielding a total score between 0 and 12. Higher scores reflect lower belief in mythical causes and, thus, greater awareness. Similar to the CAM scale, the CAM-MYCS does not include predefined cut-off values, and scores were analyzed as continuous variables. In this study, the scale demonstrated acceptable reliability (Cronbach’s alpha = 0.731), consistent with the original validation by Smith et al. (2018) (α = 0.86) [7]. The Turkish version was previously adapted and validated by Coskun (2020) using back-translation and content validity indexing methods [14].

Health and Lifestyle Behaviors Form included questions on Body Mass Index (BMI) smoking (using of traditional cigarettes), physical activity levels, fruit and vegetable intake, and, alcohol consumption. This set of behavioral and lifestyle factors was collected to examine their potential associations with cancer awareness levels, as measured by the CAM and CAM-MYCS questionnaires. These variables are recognized in the literature [9] as modifiable risk factors for chronic diseases, including several types of cancer. BMI was calculated using self-reported current height and weight, based on the formula: weight in kilograms divided by height in meters squared. Participants were categorized as belonging to the normal weight (18.5–24.9), overweight (25.0–29.9), or obese (≥30) categories according to the World Health Organization (WHO) classification of BMI [15]. Smoking status was assessed by asking participants whether they currently smoked, or had never smoked, and classified accordingly. Physical activity was measured by asking how many days per week participants engaged in activities lasting at least 30 minutes per day that increased their breathing rate [9]. Physical activity levels were evaluated according to 2020 WHO recommendations, which state that a minimum of 150 minutes of moderate activity is advised per week [16]. The level of fruit and vegetable consumption was measured by asking participants how many servings (1 serving = 80g; one small apple approximately 90 g) they consumed daily [9]. Fruit and vegetable consumption of at least five servings (approximately 400g) per day was considered as meeting the recommended level [17]. Alcohol use was assessed and quantified by asking how many days per week and how many units per day participants consumed alcohol (1 unit was considered as 1 can/bottle of beer, 1 glass of wine, or 45 ml of spirits such as whiskey or vodka).

Operational definitions

In this study, cancer awareness was operationalized through two validated instruments: the CAM (Cancer Awareness Measure), which assesses recognition of scientifically established cancer risk factors, and the CAM-MYCS (Mythical Causes Scale), which evaluates beliefs in unvalidated or mythical cancer causes. Factors considered to be associated with cancer awareness were selected among sociodemographic characteristics (age, gender, marital status, income, department, year of study), oncology course attendance, family history of cancer, and health/lifestyle behaviors (BMI, smoking, physical activity, diet, and alcohol use).

Statistical analysis

Data were analyzed using SPSS version 22.0. The normality of distribution was tested using the Kolmogorov-Smirnov test. As not all variables used in inferential analyses followed a normal distribution, nonparametric tests were applied where appropriate. Descriptive data such as frequencies, percentages, and mean scores are presented in Tables 1 and 2. Mann–Whitney U tests were used for comparisons between two groups, and Kruskal–Wallis tests were used for comparisons among three or more groups. The relationship between CAM and CAM-MYCS scores was examined using Spearman’s correlation test [18].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Sociodemographic characteristics of the sample.

https://doi.org/10.1371/journal.pone.0336102.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Means scores of CAM and CAM-MYCS.

https://doi.org/10.1371/journal.pone.0336102.t002

To identify predictors of cancer awareness, two separate multiple linear regression analyses were conducted. The dependent variables were the total scores obtained from the CAM (range: 0–11) and CAM-MYCS (range: 0–12) scales. These total scores represent the number of correct responses given by each participant across all items in each scale. Independent variables included: age, gender, marital status, income level, academic department, year of study, completion of an oncology-related course, family history of cancer (first- or second-degree relatives), and behavioral/lifestyle factors (BMI category, smoking status, physical activity level, fruit and vegetable intake, and alcohol use). Before conducting the analyses, regression assumptions were assessed through residual plots and multicollinearity diagnostics (Variance Inflation Factor, VIF). Model fit was evaluated using R², adjusted R², F-tests, and standardized beta coefficients (β). A p-value <0.05 was considered statistically significant.

Patient and public involvement

Patients or members of the public were not involved in the design, conduct, reporting, or dissemination plans of this research. The study was designed by the research team based on existing literature and public health priorities.

Ethical considerations

The study received ethical approval from the Non-Interventional Clinical Research Ethics Committee of Biruni University (Decision No. 2023/85–57, dated 22.12.2023). All procedures were conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from all participants prior to data collection via a digital form.

Sociodemographic factors

Of the total participants, 85.2% were female and 14.8% male. The majority were between the ages of 18 and 45 years. Most participants were single and enrolled in the departments of Nursing, Midwifery, and Physiotherapy and Rehabilitation. More than half of the sample reported a moderate income level. First- and second-year students constituted the largest share of the study population. The sociodemographic characteristics of the participants included in the study are presented in Table 1.

Awareness for accepted and mythical cancer risk factors

More than four-fifths of the participants endorsed the established cancer risk factors listed in the CAM, including active smoking (97.2%), alcohol consumption (88.5%), and passive smoking (81.5%). At least half of the participants also identified insufficient fruit and vegetable intake (50.4%) and lack of physical activity (55%) as cancer risk factors. The least endorsed risk factor was aging, with only 40.3% recognizing it as a cancer risk factor (Fig 1a). The CAM-MYCS item most accurately identified by participants as not being a proven cancer risk factor was physical trauma, such as being hit or crushed (38.1%). Fewer than 10% of participants correctly rejected the following CAM-MYCS items as cancer risk factors: genetically modified foods (1.3%), artificial sweeteners (2.2%), foods containing additives (2.6%), stress (2.8%), electromagnetic frequencies (3.6%), aerosol sprays (8.4%), and power lines (9.2%). Overall, the percentage of correct responses for all other CAM-MYCS items, except for the physical trauma item, remained below 18% (Fig 1b).The mean score of responses to the CAM items was 70.72 ± 21.8, while the mean score of responses to the CAM-MYCS items was 10.25 ± 14.42 (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Participants’ endorsement of actual and mythical cancer risk factors. (a) Endorsement of actual causes of cancer. (b) Endorsement of mythical causes of cancer.

https://doi.org/10.1371/journal.pone.0336102.g001

Participants aged 25 years and older had significantly higher CAM median scores than younger age groups (p = .000; η² = .05). Married participants scored significantly higher on the CAM median scores compared to unmarried participants (p = .002; r = .10). Students enrolled in the midwifery department had higher CAM median scores than those in departments other than nursing and audiology (p = .000; η² = .053). CAM median scores were significantly higher among second- and fourth- year students compared to first- and third- year students (p = .000; η² = .10). No significant differences were found in CAM median scores based on participants’ gender or income level (p = .209 and p = .620, respectively). Those who had taken an oncology-related course had higher CAM median scores than those who had not (p = .000; r = .02). Participants with a first- or second-degree relative diagnosed with cancer also had higher CAM scores (p = .042; r = .06) (Table 3). Regarding behavioral and lifestyle factors, participants with a normal Body Mass Index (BMI) scored higher on the CAM median scores than those who were overweight or obese (p = .000; r = .17). In addition, those who consumed the recommended amount of fruits and vegetables (≥400g/day) had higher CAM median scores (p = .004; r = .10). However, no significant differences in CAM median scores were found based on smoking status, physical activity level, or alcohol consumption (p = .322, p = .286, and p = .872, respectively) (Table 4). Participants who were 18 years old had significantly higher CAM-MYCS median scores than those in older age groups (p = .000; η² = .03). Unmarried students had higher CAM-MYCS median scores than their married peers (p = .013; r = .08). Students from the midwifery department had lower CAM-MYCS median scores than students in departments other than audiology (p = .000; η² = .02). First-year students scored higher on the CAM-MYCS than those in the second, third, and fourth years (p = .000; η² = .02). Those who had not taken an oncology course scored significantly higher than those who had (p = .000; r = .14). No significant differences in CAM-MYCS median scores were found based on participants’ gender, income level, or having a first- or second-degree relative diagnosed with cancer (p = .455, p = .208, and p = .383, respectively) (Table 3). With respect to behavioral and lifestyle factors, overweight or obese participants had higher CAM-MYCS median scores than those with normal BMI (p = .002; r = .10), and participants who did not meet fruit and vegetable intake recommendations scored higher than those who did (p = .012; r = .08). No significant differences were found in CAM-MYCS median scores based on smoking status, physical activity, or alcohol consumption (p = .611, p = .254, and p = .656, respectively) (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. CAM and CAM-MYCS scores stratified by sociodemographic characteristics, oncology course attendance, and family history of cancer.

https://doi.org/10.1371/journal.pone.0336102.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. CAM and CAM-MYCS scores stratified by lifestyle and health behavior factors.

https://doi.org/10.1371/journal.pone.0336102.t004

Correlation between CAM and CAM-MYCS scores

A significant moderate negative correlation was found between the CAM and CAM-MYCS scales (r = –.511; p = .000). As participants’ awareness of actual cancer risk factors increased, their awareness of myths related to cancer risk factors decreased.

Result of regression analysis of CAM and CAM-MYCS associated factors

The multiple linear regression model predicting CAM scores was significant (F = 14.57, p < .001; R² = .142, Adjusted R² = .132). Significant predictors included: grade level (β = .239, p < .001), oncology course attendance (β = .106, p = .002), and normal BMI (β = −.156, p < .001). The CAM-MYCS model was also significant (F = 6.85, p < .001; R² = .072, Adjusted R² = .062). Significant predictors were: grade level (β = −.179, p < .001) and high BMI (β = .087, p = .008). Cohen’s f² was 0.165 for the CAM model and 0.078 for the CAM-MYCS model. (Table 5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Regression results for factors predicting CAM and CAM-MYCS scores.

https://doi.org/10.1371/journal.pone.0336102.t005

Awareness for evidence-based and mythical cancer risk factors

Regarding evidence-based risk factors, over half of the participants accurately identified known contributors to cancer, with the exception of aging. Among these, smoking—acknowledged as the leading cause of lung cancer and linked to 12–17 other cancer types [19]—was the most widely recognized. This finding is consistent with previous literature, both employing and not employing the CAM and CAM-MYCS scales, in which smoking has emerged as the most widely recognized cancer risk factor [9–11,20]. Similarly, a relatively high level of awareness was observed for HPV, identified by 72% of respondents as a cancer risk factor. Although HPV is primarily linked to cervical cancer, its association with head and neck malignancies is also well documented [21], suggesting participants had a commendable level of knowledge on this topic. In contrast, lifestyle-related factors such as obesity (65.7%) and physical inactivity (55%) were less frequently identified as cancer risks. This finding aligns with earlier literature showing a tendency to underestimate the impact of modifiable behaviors on cancer development [9,10,22,23]. These behaviors contribute to cancer through various biological pathways, including hormonal imbalance, metabolic dysfunction, insulin resistance, and chronic inflammation [24], and are classified among the preventable causes of cancer [25]. Notably, only 50% of participants recognized low fruit and vegetable consumption as a risk factor—making it one of the least acknowledged in our sample. This is in line with previous findings, where dietary factors often received the lowest recognition [9–11]. Aging was identified as the least known cancer risk factor, despite its fundamental biological role in carcinogenesis. Some researchers argue that while approximately one-third of cancer risk is attributable to modifiable environmental factors, the remaining majority results from intrinsic biological processes, such as cumulative stem cell divisions and random DNA replication errors [26]. From this perspective, aging represents a non-modifiable but inevitable risk factor—sometimes termed the “bad luck” component of cancer. However, more recent models posit that modifiable and non-modifiable factors are not mutually exclusive, and that decelerating the aging process through behavioral and medical interventions may reduce cancer risk by limiting DNA damage and modulating stem cell activity [27]. These findings underscore the importance of raising public awareness—particularly among younger populations—about aging as a cancer risk factor, and integrating this knowledge into prevention strategies.

Awareness of mythical cancer risk factors was notably low in our sample. Fewer than 18% of participants correctly identified all CAM-MYCS items—excluding the physical trauma item—as myths, a figure even lower than those reported in previous studies that also documented limited awareness of cancer-related misconceptions [9,10,20]. This suggests that university students in our study faced difficulty distinguishing evidence-based risk factors from commonly held but scientifically unfounded beliefs. One plausible explanation is the generalized perception that “everything causes cancer,” a cognitive bias that may lead participants to endorse items on both the CAM and CAM-MYCS scales without critical evaluation [10]. This bias is concerning, as it not only propagates misinformation but also blurs the distinction between real and mythical risks, ultimately complicating individuals’ ability to make informed health decisions. While low awareness of validated cancer risk factors is known to reduce engagement in preventive behaviors, belief in myths may be equally harmful. Misidentifying myths as real threats can divert attention from evidence-based preventive strategies—such as reducing alcohol intake or increasing physical activity—and may heighten cancer-related anxiety. Enhancing public understanding of scientifically supported risk factors is therefore essential. Educational interventions that improve individuals’ ability to discern between factual and mythical causes can foster more effective risk-reducing behaviors and mitigate misinformation. In our study, artificial sweeteners and genetically modified (GM) foods were the most commonly believed mythical cancer risk factors. This finding may be attributed to culturally specific narratives in Turkey, misinformation spread through social media platforms, and widespread skepticism toward industrial food production.

The overall mean CAM score in this study was moderately high, indicating a fair level of awareness regarding evidence-based cancer risk factors among participants. In contrast, the low mean CAM-MYCS score reflects poor ability to accurately reject mythical causes of cancer. This discrepancy suggests that while students may be familiar with established risk factors, they are considerably less adept at identifying misinformation or myths. Similar findings in previous studies have also highlighted this gap, emphasizing the need for integrated educational strategies that simultaneously promote evidence-based knowledge and critical thinking to combat cancer-related misconceptions [9,10,20].

Sociodemographic and health/lifestyle factors associated with cancer awareness

When examined in relation to sociodemographic characteristics, the sample predominantly comprised a young population. Participants aged 25 and above demonstrated significantly greater accuracy in identifying validated cancer risk factors. Conversely, those under the age of 18 showed higher accuracy in recognizing mythical cancer risk factors (p < .001). This inverse trend suggests that younger participants may have greater exposure to digital health content, online awareness campaigns, and updated curricula, which may contribute to improved recognition of misinformation. Similar findings were reported by Shahab et al. (2018), who found that individuals under the age of 30 were more likely to correctly identify both evidence-based and mythical cancer causes [9]. In contrast, Munir et al. (2022) found no significant association between age and awareness levels [10]. Marital status was also associated with differing patterns of awareness. In our study, married participants demonstrated significantly higher scores on the CAM scale, while single participants performed better on the CAM-MYCS scale. These findings differ from Munir et al. (2022), who reported greater accuracy among single individuals in identifying actual cancer risk factors, and from Shahab et al. (2018), who found no significant association between marital status and awareness [9,10]. Participants who had taken a course in oncology demonstrated significantly higher levels of awareness regarding both validated and mythical cancer risk factors (p < .001, p < .001 respectively). The regression results support these findings. The likely contributors include curriculum content focused on evidence-based learning, which strengthens both theoretical understanding and practical health literacy. Given persistent gaps—particularly in distinguishing myths from factual risks—integrating cancer literacy into university curricula appears essential. Peer-led workshops, expert seminars, and digital campaigns may support myth-correction efforts, while broader public health interventions aimed at combating misinformation could further improve cancer awareness among young adults.

Participants with a normal BMI demonstrated significantly higher awareness of validated cancer risk factors compared to those classified as overweight or obese. This finding aligns with previous research suggesting that individuals maintaining a healthy weight are more likely to engage in preventive health behaviors and be receptive to health-related information [10,28]. One possible explanation is that health-conscious individuals may actively seek and internalize evidence-based information, which contributes both to healthier weight status and greater awareness of cancer risks.

Similarly, participants who met the recommended daily intake of fruits and vegetables exhibited greater awareness of validated cancer risk factors. This may reflect greater exposure to nutritional and preventive health education among those who adhere to dietary guidelines, as dietary behaviors often correlate with health literacy levels [9]. Therefore, a higher level of cancer awareness in this group might stem from broader engagement with health-promoting knowledge and practices.

Correlation between measures of awareness for evidence-based and mythical risk factors

Participants demonstrated greater awareness of scientifically validated cancer risk factors compared to unproven myths, with a significant negative correlation observed between the two domains. This pattern is consistent with previous findings from populations in England [9], Pakistan [10], and Spain [11], suggesting that individuals who are more informed about evidence-based risks may be less adept at identifying cancer-related myths, and vice versa. One possible explanation is that individuals tend to focus on either scientific facts or popular beliefs, resulting in an imbalanced understanding of cancer risk. These findings point to a potential fragmentation in cancer-related health literacy and underscore the need for further investigation into the cognitive and informational factors contributing to this divide.

In the regression analyses, academic year, normal BMI, and oncology course attendance emerged as significant predictors of cancer risk factor awareness. Higher academic standing was positively associated with greater awareness of validated cancer risk factors, possibly reflecting the cumulative effect of academic exposure and health-related curricula. Similarly, students who had taken an oncology course demonstrated higher CAM scores, suggesting that structured education plays a crucial role in improving cancer literacy. Interestingly, normal BMI was linked to better awareness of evidence-based cancer risks, which may reflect the adoption of health-promoting behaviors informed by accurate knowledge. These findings align with previous literature that highlights the impact of education and personal health behaviors on cancer awareness levels [9,11,12]. However, the limited explanatory power of the models (R² = .142 for CAM and R² = .072 for CAM-MYCS) suggests that additional unmeasured factors—such as psychological, cultural, or informational influences—may also play a role. The regression model for CAM-MYCS revealed that lower grade level and higher BMI were significant predictors of increased awareness of mythical cancer risk factors. These findings suggest that younger students and those with higher body mass may be more susceptible to misinformation or less able to critically evaluate unvalidated health claims. The effect size, while modest (f² = 0.078), indicates a meaningful association worth addressing in future educational interventions targeting myth-related cancer beliefs.