Introduction

The population ageing process is a significant challenge in the field of health policy, leading to an increasing demand for medical specialists in the field of geriatrics and geriatric nursing. Demographic forecasts and the increasing prevalence of chronic diseases require not only the adaptation of the healthcare system, but also the development of appropriate attitudes towards old age among future medical professionals [1]. Healthcare professionals have significant potential to raise the standard of care for older people. Therefore, it is important to foster appropriate attitudes towards aging, identify its determinants, and challenge misconceptions about old age [2, 3].

Anxiety about aging is a multidimensional phenomenon involving cognitive, emotional and behavioral components related to fears of aging [4]. It is an important psychosocial construct associated with attitudes and behavior toward older people among nursing students, as well as their adjustment to their own aging. This is supported by many scientific studies, whose authors suggest that the level of anxiety about aging is related to ageism and the formation of attitudes towards older people [4-7]. Insufficient preparation may be associated with negative attitudes and reluctance to work with this group. It is worth assessing the level of compassion and aptitude for caregiving during the education of medical students at all levels [1, 2].

In recent years, the number of available tools for the evaluation of attitudes towards older people has increased, but many of them do not take into account the multidirectional nature of the anxiety about aging. Because the dimensions of anxiety and its consequences are still insufficiently understood, the Anxiety about Aging Scale (AAS) warrants further empirical investigation [4, 8].

The AAS tool has been successfully tested in more than a dozen countries around the world, including Australia [8], Taiwan [9], Greece [10], Spain [11], Iran [12], Turkey [13], Malaysia [14], Croatia [15], Pakistan [16], Lithuania [17], and Mexico [18-20]. In all studies, confirmatory and/or exploratory factor analyses or their alternatives were conducted in order to understand the structure of the test factor and adapt the instrument to the given cultural and linguistic conditions. Despite the wide application of the AAS scale in international research, in Poland, so far, no cultural adaptation and evaluation of the psychometric properties of this tool has been conducted in a group of nursing students. This is a significant research limitation due to the evolving attitudes among nursing students, who will play a key role in the care of older people as future medical professionals. This study aimed to provide a preliminary evaluation of the psychometric properties of the Polish version of the AAS in a sample of nursing students.

Material and methods

Data collection

The cross-sectional survey was conducted between August and October 2024 among 235 nursing students enrolled in bachelor’s (years 1-3) and master’s (years 1-2) degree programs.

The survey was developed using the Google Forms website. A computer-assisted web interviewing (CAWI) technique was used in the study, which allows the questionnaire to be distributed via a link to the website, while offering the full anonymity of the respondents and the possibility to answer questions at a convenient time and facilitating the collection of reliable data [21]. The link to the survey was posted on social media sites used by nursing students, such as Facebook, Messenger and WhatsApp, inviting all students to participate. At the beginning of the questionnaire, there was an introduction explaining the purpose of the survey, instructions on how to fill it in, and information that the collected data would be used for scientific purposes. The questionnaire also included the researcher’s email address in case of questions or concerns and information about the possibility of withdrawing from the study at any time. Inclusion in the study was based on the criterion of providing informed consent to participate in the study, which allowed the survey to commence. Individuals who did not give such consent and students enrolled in other fields of study were excluded. The questionnaire included questions regarding the sociodemographic characteristics of the students (gender, age, marital status, field of study, and year of study), work in the profession as well as preparation and readiness to work with older people (including work experience in the profession and readiness to work in the profession in geriatric wards), contact with older people (including contact with older people at work and contact with people over 65 and over 80) and the AAS.

Measurement instrument

The Anxiety about Aging Scale (AAS) is used to evaluate anxiety about aging, which is an indicator of attitudes and behavior towards older people, as well as an element of adaptation to one’s own aging process.

The scale consists of 20 items divided into four areas: 1 – fear of older people, 2 – psychological symptoms of old age, 3 – physical appearance, and 4 – symptoms related to fear of loss. Respondents indicate their level of agreement using a 5-point Likert scale (1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree). The higher the score, the higher the level of anxiety among the respondents. The scale not only allows the identification and assessment of the sources and levels of fear related to aging, but also provides important information for developing therapeutic and educational programs that can support the development of awareness and tolerance towards the aging population in society [4].

Obtaining permission to use the AAS scale in the present study was consistent with the guidelines provided by the authors of the instrument, as outlined in the original publication describing its development and application. The cultural adaptation of the Polish version of the AAS-PL from the original version was assessed in five aspects: translation (content of questions, difficulty level of question formulation), facade/aesthetics (the Polish language version had the same graphic design, i.e., text appearance, instructions, methods of calculating results), functional (suitability for the same purposes), reconstruction (methods for reliability and validity verification, types of standards) and psychometric (whether the tool is characterized by adequate internal consistency, construct validity, and whether the measurement results are repeatable and reliable).

The process of translating the original version of the questionnaire from English into Polish was carried out by two independent expert translators. Subsequently, both versions were compared in terms of consistency of content and meaning, and appropriate corrections were made to the Polish translation in order to faithfully reflect the authors’ intentions and the meaning of individual elements of the tool. The next step involved two expert translators carrying out a back translation of the Polish version of the scale into English and then comparing the results with the original version. A comparison of the original questionnaire with the back translation indicated a high degree of correspondence between the two versions in terms of both wording and content.

Statistical analysis

Distributions of qualitative variables were described by the absolute frequency of individual categories (N) and their percentage in the distribution of the variable (%). Mean age and standard deviation (SD) were reported. Descriptive statistics were used for quantitative variables. Correlations were analyzed using Pearson’s correlation coefficient for quantitative variables without outliers and for variables measured on a Likert scale, and Spearman’s rank correlation coefficient. Student’s t-test for independent samples was used to compare groups according to marital status, while the Mann-Whitney U-test was used to compare groups according to contact with older people. The choice of non-parametric test was based on the non-uniformity of the compared groups. Statistical analysis was conducted using IBM SPSS Statistics version 29. The statistical significance level was set at a = 0.05.

Factor analysis was performed using JASP 0.18.3., with confirmatory and exploratory factor analysis. In addition, the reliability of factors was calculated based on Cronbach’s a. The validity of the tool was assessed using average variance extracted (AVE) and the heterotrait–monotrait ratio (HTMT); correlations between factors were also examined.

Results

Characteristics of the study group

The analysis covered a group of 235 nursing students (none were excluded due to missing data), including 95.3% women. The average age in the study group was 23.31 years (SD = 6.28). More than half of the respondents (54%) were single, and more than a third (37.4%) were in an informal relationship. The majority of students were in their second year of study (35.3%). Professional experience was reported by 36.6% of respondents. The sociodemographic characteristics of the study group are presented in Table 1.

One third (35.7%) of the nursing students surveyed stated that they had been prepared for working with older people, while 27.7% declared that they had not been prepared. The majority of respondents (72.3%) believed that the number of hours devoted to devoted to geriatrics and the care of older people during their studies was sufficient. Daily contact with older people was reported by most of the students surveyed (86.4%), with 66.3% reporting daily contact at work. Frequent or very frequent contact mainly concerned people aged over 65 years, while contact with people aged over 80 years was only occasional for 44.7% of respondents. Public transportation was the most frequently indicated place of contact with older people (23.5%), followed by healthcare facilities (21.3%) and the family environment (17.5%). In response to the question about attitudes towards older people, the majority of participants (71.5%) declared a positive attitude towards older people. More than half of the respondents had experience in working with older people as a part of their studies, and approximately one quarter had gained such experience through professional work or caring for family members. Among the motivating factors for working in a geriatric ward, nursing students most often identified factors unrelated to the nature of the work, such as salary, team atmosphere or workplace location (32.2%). The most common demotivating factors were a lack of interest in geriatrics and the subject of caring for older people (31.1%) and a lack of satisfaction from working with older people (15.2%). The detailed results are shown in Table 2.

Factor structure of the AAS-PL tool

Confirmatory factor analysis (CFA) of the original four-factor model, with five items per factor, indicated an imperfect fit to the data: c²(164) = 346.34; p < 0.001; CFI = 0.925; TLI = 0.913; GFI = 0.940; RMSEA = 0.069; SRMR = 0.093. In addition, some factor loadings did not meet the minimum threshold of 0.35-0.40 [22], as shown in Figure 1.

Exploratory factor analysis (EFA) showed that item skewness did not exceed |2| and kurtosis did not exceed |7| [23]. Parallel analysis suggested four factors. Maximum likelihood extraction with promax rotation was used to allow correlations between factors. Six items (2, 7, 11, 15, 16, and 17) were excluded, yielding a proposed 14-item solution. The reported KMO value of 0.796 and Bartlett’s test result, c²(91) = 1542.90, p < 0.001, pertain to 14 items, as indicated by the 91 degrees of freedom. The initial 20-item EFA results, including the loadings and communalities of the six excluded items, are not presented; therefore, the basis for their exclusion cannot be assessed from the reported data. The first factor comprised four items (loadings 0.67-0.88) and explained 19.3% of the variance. The second comprised three items (0.75-0.89; 15.1%), the third four items (0.48-0.89; 14.4%), and the fourth three items (0.72-0.85; 13.0%). Item 6 was grouped with items 18-20. The resulting four-factor structure is a proposal for further evaluation. Items were reverse-coded as needed so that higher scores indicated greater anxiety about aging.

The proposed 14-item model was subsequently examined by CFA in the same sample used for item selection. Fit indices were c²(71) = 113.51, p < 0.001; CMIN/df = 1.60; CFI = 0.972; TLI = 0.964; GFI = 0.934; RMSEA = 0.050; and SRMR = 0.048. Because the CFA used the same data as the EFA, these indices describe in-sample fit and do not independently confirm the proposed structure. AVE exceeded 0.50 for each factor and HTMT was below 0.85 for each factor pair in this sample. Correlations between factors were strongest for factors II and III, whereas several other correlations were weak or nonsignificant. Factors I and IV were less strongly related to the overall score than factors II and III; the overall score should therefore be interpreted cautiously. Table 3 presents the EFA loadings, internal consistency, AVE, HTMT, and factor correlations. CFA fit indices are reported in the text.

Basic descriptive statistics of the measured quantitative variables with the normality of distribution test

The next stage of the analysis involved examining the distributions of quantitative variables. For this purpose, basic descriptive statistics were calculated along with the Shapiro-Wilk test to assess the normality of the distributions. Age, fear of older people, and psychological symptoms showed outliers more than three standard deviations from the mean (extremely high scores). Winsorization was performed, with outliers replaced by the nearest non-outlying values. The variables were also analyzed according to marital status, daily work contact with older people, and work in the profession, with outliers being removed in the same way. Despite the presence of outliers, no transformation was undertaken for age due to their large number. Calculation results for the entire sample are presented in Table 4.

Anxiety about aging vs. sociodemographic and occupational factors

Correlation analysis between sociodemographic and professional factors showed a significant negative relationship in the following cases:

• year of study and the overall level of anxiety about aging (a higher year of study was associated with lower anxiety about aging);

• frequency of contact with people over 65 and fear of older people (people who had frequent contact with older people showed a lower level of fear). This relationship was not found for people who had contact with people over 80;

• readiness to work in a geriatric ward and fear of older people, physical appearance as well as fear of loss and general anxiety about aging. This indicates that lower levels of anxiety and anxiety about aging were associated with greater readiness to work in geriatric care;

• the perceived need for more hours devoted to geriatrics in the curriculum and fear of older people (a stronger belief that more hours were needed was associated with lower fear of older people);

• the perceived need for additional hours of geriatrics instruction and fear of older people (a stronger belief that more hours of geriatrics classes were necessary was associated with lower fear of older people).

No relationship was found between age, professional experience, frequency of contact with older people at work, or anxiety about aging (Table 5).

The analysis of differences included marital status (single and in a relationship), daily contact with older people, and employment in the nursing profession (Table 6). The independent-samples Student’s t-test showed no differences in the level of anxiety about aging according to marital status. However, the Mann-Whitney U-test indicated that daily contact with older people distinguished two subscales related to anxiety about aging: fear of older people and psychological symptoms. The median scores indicated that people without daily contact with older people were characterized by a higher fear of older people than people who had daily contact. In contrast, people who had contact with older people had higher scores for psychological symptoms. There were no significant differences in the level of anxiety about aging in relation to physical appearance, fear of loss, or overall score. In addition, employment in the nursing profession was not significantly associated with anxiety about aging.

Discussion

This study provides a preliminary psychometric evaluation of a modified Polish version of the AAS in a sample of nursing students. EFA suggested a four-factor, 14-item solution after six of the original 20 items were removed. Subsequent CFA showed favorable fit in the same sample; however, it cannot independently confirm the proposed structure because item selection and model testing involved the same respondents. The proposed subscales showed high internal consistency in this sample. Further research should confirm the factor structure in an independent sample and assess whether the retained items adequately represent the content of the original scale.

The four-factor model of the scale was supported in a study conducted in Iran involving older people [12]. An Australian study conducted among people aged 20 to 97 also identified a four-factor structure [8], as did studies conducted in Spain [11]. A study conducted among an adult Mexican population also showed the existence of four factors in the described scale [19], similarly to a study concerning Mexican students [20] and older Mexican citizens [18]. The four-factor form of the instrument was supported in a study conducted among the adult population in Taiwan [9]. In Greece, a study conducted among three groups – primary health care professionals, nursing and social work students, and community residents – also showed a four-factor structure [10]. Another study conducted in Lithuania [17] supported the four-factor model, as did studies in Turkey [13], Pakistan [16], and Malaysia [14] among menopausal women. Moreover, a study conducted among Croatian nurses also showed a four-factor model structure [15].

The Polish version of the AAS-PL scale consists of 14 items, as does the Mexican version developed for students of physical education and human kinetics, which also excludes 6 of the original items. Both the Polish and Mexican studies excluded items 7, 11, and 17 regarding having multiple jobs in old age, lying about one’s age, and worrying about one’s health with age [20]. The other items excluded in the Polish study concerned questions about visiting older relatives, appearance changing with age, and the loss of friends as people grow older. It is worth noting that in both studies, EFA assigned the retained items to factors in the same order as in the original instrument. In the Mexican version of the tool, developed in 2023, eight items were removed to achieve better model fit and greater discriminative ability. Five of these items were also eliminated in the Polish version of the scale. In addition, a study conducted in Mexico showed that the modified version of the tool was shorter and could be administered more quickly than the original version [19]. A study conducted in Taiwan showed a structure consisting of 16 items; however, only one of them was identical to items excluded in the Polish study [9]. In the Malaysian study, a 17-item structure was supported after 3 items were rejected. Two of the rejected items in the Malaysian publication were also removed in the Polish study [14]. The Turkish study, on the other hand, excluded 4 items, supporting the 16-item structure of the tool [13]. In the Greek study, based on the results of factor analysis, attention was drawn to the possible reduction of the number of factors due to repeated content, but no attempt was made to reduce them [10]. A similar situation was observed in Lithuania, where despite the problems observed with the question regarding lying about age, it was decided to leave it in the Lithuanian version of the AAS in order to maintain consistency with the original version of the tool [17].

The number of questions was not modified in the studies conducted in Spain [11], Iran [12], Australia [8], Greece [10], Lithuania [17], and Croatia [15].

The elimination of 6 items in the Polish version of the AAS-PL may indicate cultural differences in the perception of aging between Poland and the United States, where the original version of the tool was developed. It is worth noting that the specificity of the study group was also important. Nursing students are more aware of the biological and social processes related to aging, which may have been associated with their attitudes toward these issues. In the case of the excluded question “I have never lied about my age to look younger,” many studies indicate a low factor loading compared to other items. Researchers from Lithuania, Mexico, Turkey, Taiwan, and Australia had difficulties retaining this item in the adapted version of the scale [8, 9, 13, 17, 18]. This may suggest that it is not a good indicator for the evaluation of anxiety about aging. However, this item was retained in other studies, suggesting that its suitability may vary across cultural contexts.

The proposed four-factor, 14-item solution showed favorable fit indices and internal consistency in this sample. AVE exceeded 0.50 for each factor and HTMT was below 0.85 for each factor pair. These findings provide preliminary evidence of the modified scale’s factor structure and convergent and discriminant validity within this sample but do not establish its construct or content validity independently. Further assessment of the retained items and replication of the factor structure in a new sample are needed before conclusions can be drawn about the adequacy of the Polish adaptation.

The Cronbach’s a coefficient for individual factors ranges from 0.80 to 0.89, indicating high internal consistency of the tool. In the original instrument, the values range from 0.69 to 0.78 and are acceptable, but lower compared to the Polish version of the AAS [4]. These values in other studies ranged from 0.67 to 0.88 [8, 9, 11, 17]. When comparing the reliability of the Polish adaptation with studies conducted in other countries, high Cronbach’s a coefficients were also observed in analyses carried out in Spain, Iran, and Croatia [11, 12, 15]. These results cannot, on their own, establish the stability of the AAS construct across cultural contexts. Analyzing the correlations between individual factors, the Polish study found strong correlations between factors II and III (physical appearance and symptoms related to fear of loss). The correlations between factor I and IV (fear of older people and psychological symptoms of old age) and III and IV (symptoms related to fear of loss and psychological symptoms of old age) were weak. Taking into account the correlations with the overall score of anxiety about aging, factors I and IV correlated the least, while factors II and III correlated the most. This suggests that the overall score may not fully represent all dimensions of anxiety about aging, as some factors are weakly related to the total score, and the total score should be interpreted with caution. In the American version of the tool, the highest correlations were observed between factors III and IV, although the correlation coefficient was only 0.39, indicating a moderate association [4]. In other studies, the correlation values between individual factors varied greatly, but the four-factor structure was nevertheless retained.

The analyses carried out in the Polish study revealed that people who had more frequent contact with older people had lower anxiety about aging. Researchers came to similar conclusions in a Canadian study [5], as did Boswell [6] in his study on the factors influencing ageism among psychology students. This supports the theory proposed by the authors of the original AAS that anxiety about aging is a factor influencing the formation of attitudes and behaviors towards older people [4].

An analysis of associations between sociodemographic variables and anxiety about aging in the Polish study showed that a higher year of study was associated with lower anxiety about aging. A similar conclusion was reached by Greek researchers in a study of nursing students, who noted that nursing graduates showed more positive attitudes towards older people than first-year students [24]. In the paper describing the Greek adaptation of the AAS, which surveyed nursing and social work students among others, a higher level of anxiety about aging was observed among students than among the other study group (primary care professionals and community residents). The authors suggested that this may be due to knowledge deficits resulting from insufficient hours devoted to subjects in the field of geriatrics and geriatric care during higher education [10]. Harris and Dollinger presented similar conclusions in their study, reporting that participants with more knowledge about the ageing process showed more positive attitudes towards older people than the control group [7]. Researchers from Mexico also drew attention to the need for further research in this area among a group of students [20]. Furthermore, the Polish study found that a stronger belief in the need to increase the number of hours devoted to geriatrics was associated with lower anxiety about aging. These findings are noteworthy due to their practical implications.

Another important finding was that lower levels of anxiety and anxiety about aging were associated with greater readiness to work in geriatric wards. This observation is particularly important in view of unfavorable demographic projections and the anticipated increase in demand for geriatric professionals who can provide high-quality care to older people. Further research using the ASS may help clarify the factors associated with attitudes towards aging and how these attitudes develop among nursing students [2, 3].

The practical significance of assessing anxiety about aging in nursing education lies in the need to better understand the emotional dimension of preparing students to care for older people. Identifying concerns about loss of independence, changes in physical appearance, or reduced social connections may provide a starting point for reflection on how perceptions of one’s own aging relate to the way older patients are perceived. Nursing education should provide opportunities to discuss these emotions and beliefs and link this reflection to the development of communication skills, respect for autonomy, and recognition of older people’s individual needs.

The quality of experiences gained through contact with older people may also be important. Simply increasing the frequency of such contact may not be sufficient to alleviate students’ concerns. Consideration should be given to combining clinical placements with teacher-guided reflection on these experiences and providing opportunities to engage with older people with varying levels of functional ability, including those living independently in the community.
This approach could help students develop a more diverse view of aging that extends beyond illness and care dependency. The educational needs identified could then guide the selection of course content and forms of support for students in situations that cause uncertainty or difficulties in interacting with patients. These are proposed directions for improving nursing education; their effects on competence development and the quality of subsequent care requires evaluation in longitudinal and intervention studies.

It is worth noting that the study has several limitations. First of all, the sample was recruited through social media and comprised only nursing students, limiting its diversity and the generalizability of the findings. Further analysis on a wider group of students from other academic disciplines is necessary to understand the role of the Polish cultural context in the evaluation of anxiety about aging and to understand the factors associated with the formation of attitudes toward older people. Another limitation is the lack of gender diversity, as the majority of respondents in the study were women; this may limit the representativeness of the results. The use of self-report measures also carries the risk of social desirability bias. Participants may have provided answers that were consistent with social or professional expectations, especially in the context of sensitive topics such as attitudes towards ageing. Moreover, the temporal stability (test–retest reliability), which is a standard component of the validation process, was not assessed. Due to the inability to pair the participants in a repeated measurement, the test-retest procedure, which would indicate the temporal stability of the tool, was not conducted. Furthermore, because the participants were nursing students still undergoing training, their responses may change over time as they gain further knowledge and experience. Despite the use of rigorous procedures, potential sources of error may also include subtle nuances in translation, which could be associated with differences in the interpretation of certain items. It is also important to note that the obtained results should be interpreted within a cultural context – attitudes toward aging vary across countries, which may be associated with differences in how individual items on the AAS-PL are understood and evaluated. Finally, the issue of generalizability should be addressed. The results may not be applicable to non-student populations or individuals outside of Poland without further research. Although these limitations should be taken into account in further interpretations of the study results, they do not undermine the data obtained during the analysis of the psychometric properties of the AAS-PL.

A strength of this study is that it contributes to the international development of the AAS scale. The findings may inform the development of educational interventions aimed at shaping attitudes and behaviors towards older people, especially among future healthcare professionals. In addition, the use of CAWI allows participants to remain completely anonymous, enables them to answer questions at any time, and ensures the collection of reliable information.

Clinical implications

The findings draw attention to anxiety about aging as a topic that may be relevant to nursing education and future geriatric care. At present, however, the proposed 14-item AAS-PL should not be used to assess individual students, guide clinical decisions, or evaluate educational interventions. Such applications would require independent confirmation of its structure and further evidence concerning content validity, reliability, and score interpretation.

Future directions

Future studies should assess the relevance, coverage, and comprehensibility of retained items with experts and intended respondents. They should then test the prespecified 14-item structure in an independent sample, assess test–retest reliability, and examine relationships with relevant external measures. More diverse samples would allow evaluation beyond nursing students and assessment of whether the subscales function similarly across groups.

Conclusions

In this sample of nursing students, analyses suggested a potentially useful 14-item, four-factor Polish AAS with adequate subscale internal consistency. Because item selection and the final CFA were conducted using the same data, the modified structure has not been independently confirmed. Assessment of content validity and replication in a new sample are necessary before the proposed version, particularly its overall score, can be recommended for research or practice.

Disclosures

This research received no external funding.

The study was conducted in accordance with the Declaration of Helsinki and approved by the Research Ethics Committee of the Jagiellonian University Medical College (protocol code 118.0043.1.257.2024; date of approval: 2 August 2024).

The authors declare no conflict of interest.