Abstract
This study aimed to describe the gender roles of people interested in an exercise program done on outdoor exercise structures and test if gender roles were associated with studied outcomes. Older adults aged 65+ who were not currently performing resistance training were invited to participate. Gender roles were quantified using the Bem Sex Role Inventory 30-item questionnaire (−60 [feminine] to +60 [masculine]). Outcomes included completing the 6-week intervention (Y/N) and changes in physical function (one leg stance, 30-s chair stand), strength (predicted maximal chest press and leg press, grip strength), power (knee extensor power), and overall health via the SF-36 questionnaire. Twenty-nine adults (65.5% female; median 72 years old) participated in the study, and 17 completed the intervention (58.6%). The median (interquartile) gender role score was −13.0 (−19.5 to −8.5), with no gender role difference (p = .62) between completers and non-completers. These results suggest that older adults interested in such a program portray themselves as more feminine. No association was found between gender role scores and changes in any study outcomes. In this setting, gender roles did not seem to impact the study outcomes and therefore may not need to be considered when designing an outdoor exercise structure program.
Introduction
Very few older adults adhere to the international physical activity guidelines, especially regarding the muscle-strengthening activity recommendations (Bennie et al., 2019). There are many reasons why older adults may not engage in muscle-strengthening activities. A systematic review found that one of the largest barriers older adults face is a need for access to the required facilities and resources for physical activity (Spiteri et al., 2019). Outdoor exercise structures have been proposed as a possible solution to some of the barriers older adults face when engaging in muscle-strengthening activities because they are free, rely on body weight, and are typically installed in residential areas (Jansson et al., 2019; Lee et al., 2018).
Men tend to engage in muscle-strengthening activities more than women throughout their lifetime, partially due to the masculine connotations of the activity and how it relates to gender roles (Dworkin, 2001; Howe et al., 2017; Salvatore & Marecek, 2010). Gender refers to the socially constructed roles, behaviors, identities, and expressions of people (Government of Canada, 2018). In the 1970s, the term “gender” was introduced as an alternative to sex to challenge the idea at the time that femininity and masculinity were unidimensional traits, suggesting instead that they were two ends of a continuum, and that people could have varying degrees of both (Agbayani & Min, 2006; Krieger, 2003). Since then, tools have been developed to quantify and interpret gender more broadly. One of these tools, validated with older adults (Lyons et al., 1994), is the Bem Sex Role Inventory (BSRI) questionnaire to quantify stereotypical gender roles (Agbayani & Min, 2006).
There is some existing literature pertaining to how gender roles may have an impact on health behaviors (Fleming & Agnew-Brune, 2015). Hyuk et al. (2012) found that within a population of female university students, masculine gender roles were correlated to higher exercise participation frequency and higher intrinsic motivation (Hyuk et al., 2012). More specific to older adults, Ahmed et al. (2016) reported that feminine and androgynous gender roles were independent risk factors for determining mobility and low physical performance amongst a group of older adults (Ahmed et al., 2016).
This is a secondary analysis of a crossover trial of a 6-week exercise program on an outdoor gym in older adults. The study objectives were to, (1) describe participants’ gender roles who participate in a 6-week intervention using outdoor exercise structures, and (2) test if gender roles are associated with completing the 6-week intervention (Y/N) or changes in overall health, physical function, muscle power, and muscle strength.
Methods
Study Design
This study is a secondary analysis of a nonrandomized crossover design study (Leadbetter et al., 2024). All participants participated in a 6-week control period and then a 6-week intervention. Data collection occurred at baseline, 6 weeks (post-control), and 12 weeks (post-intervention). The main study aimed to quantify the potential changes in lower body relative muscle strength using outdoor exercise structures in older adults. Ethics approval was sought and obtained for the original study prior to the commencement of the study.
Participants
The inclusion criteria for this study were to be at least the age of 65 years old, live in a community setting, not currently be engaged in any structured muscle-strengthening activities, be able to grip objects in day-to-day life without pain (self-reported) and be cleared to exercise using the Get Active Questionnaire (Canadian Society for Exercise Physiology, 2021). Participants were excluded if they had any uncontrolled medical conditions, current musculoskeletal injuries, or vacations lasting more than 1 week during the intervention. This study aimed to recruit at least 28 people based on the original study objective. An attempt was made to recruit an equal number of men and women. Participants were recruited for the original study using virtual flyers posted on social media and physical flyers placed around the community.
Intervention
The intervention consisted of a 6-week control period followed by a 6-week active intervention. During the active intervention, from August to October 2022, participants met at an outdoor exercise structure facility for 6 weeks, three times per week. The exercise routine was designed to last 45 min, consisting of a warm-up, a resistance training program, and a cool-down. The resistance training program was developed using the American College of Sports Medicine’s recommendations and Resistance Training for Older Adults: Position Statement From the National Strength and Conditioning Association (Fragala et al., 2019). The resistance training program was the same for each participant and consisted of eight exercises focusing on major muscle groups. Participants started with the initial exercises and would continue through three exercise progressions. Participants were instructed to start with 1 set of 6 to 12 repetitions, ensuring they could complete all 12 repetitions before increasing to 2 sets of 6 to 12 repetitions and then 3 sets of 6 to 12 repetitions. Participants were instructed to take a 1–3-min rest between sets, as needed. Modifications were made by the certified exercise professional when deemed necessary to accommodate the participant’s abilities and conditions.
Participants were given a booklet with photo demonstrations to log their exercises, sets and repetitions. An exercise professional supervised sessions and participants were instructed to independently complete their exercise routine. Participants were instructed to complete their workout while the exercise professional was present at the outdoor exercise structures between 10 AM and 2 PM every Monday, Wednesday, and Friday. Participants were asked to attend a total of 18 sessions. Still, in the case of bad weather, a holiday or a personal reason, a participant could not attend a class, make-up sessions were provided at the end of the 6-week intervention as needed.
Exposure Variable
Gender roles were self-reported using the 30-item BSRI (Bem, 1974a). As the scoring instructions for the 30-item BSRI were unavailable, the scoring technique for the original 60-item BSRI was used. The 30-item BSRI items were categorized as masculine, feminine, or neutral based on the categorizations of the 60-item BSRI (Carver et al., 2013). The 60-item BSRI was scored by subtracting the sum of the feminine traits from the sum of the masculine traits; therefore, that is the general format followed for scoring the BSRI in this secondary analysis. The BSRI has 10 questions considered feminine, 10 questions considered masculine, and 10 questions considered neutral. Each item on the BSRI has a range of seven possible responses, from never or almost never true (1), to always or almost always true (7). To create the gender role scores, the feminine and masculine sections were tallied up, and the sum of the feminine traits was subtracted from the sum of the masculine traits. With ten items per gender role, and each item scored from 1 to 7, participants got a score ranging from 10 to 70 in both the masculine and feminine categories. As a result, a positive score (up to +60) shows a more masculine score and a negative score (as low as −60) is considered more feminine. In this secondary analysis, the BSRI was scored using a continuous method. The continuous method was chosen instead of a median-split method of scoring, as it has been considered a more reliable and generalizable scoring method and is proposed to limit type one error (DeCoster et al., 2011).
Outcome Measures
Baseline self-reported information was collected to answer the first objective, to describe participants’ gender roles who participated in a 6-week intervention using outdoor exercise structures, including age, sex, ethnicity, marital status, employment status, education level, and yearly family income. In addition to self-reported information, body weight and height were measured using the Canadian Society for Exercise Physiology protocols to calculate body mass index (Canadian Society for Exercise Physiology, 2021). Finally, aerobic physical activity level was assessed via a wrist accelerometer (AX3, Axivity Ltd, Newcastle Helix, UK) during a full week to determine the participant’s weekly minutes spent performing moderate to vigorous intensity activities. To be included in the analysis for the physical activity level, a minimum of four valid days (10+ hrs/day) was required (Chen et al., 2009).
The outcomes, also used to describe the characteristics of participants, included physical function, muscular and functional strength, muscle power, and overall health. Physical function was measured using the 30-s chair stand test and the one-leg stance test with eyes open, using standardized protocols (Hurvitz et al., 2000; Jones et al., 1999). Muscular and functional strength was measured through grip strength and an estimated 1-repetition maximum on the chest press. Grip strength was assessed using a JAMAR hand dynamometer, using a standardized protocol (Canadian Society for Exercise Physiology, 2021). To estimate the 1-repetition maximum for leg press and chest press, a 1- to 10-repetition maximum test was performed and estimations were made based on the Brzycki prediction equation (Brzycki, 1993). Muscle power was measured using isokinetic dynamometry testing on a HUMAC NORM Isokinetic Extremity System by CSMi (Computer Sports Medicine Inc., Stoughton, MA.) (Habets et al., 2018). Overall health was reported using the Short-form 36 Health Status Questionnaire (SF-36) (Lyons et al., 1994).
Statistical Analysis
Because of the relatively small sample, descriptive characteristics are presented using median (25th–75th) or N (%). Gender role score between participants completing or not the intervention was assessed using a Mann-Whitney U test. Any associations with gender role scores were tested via Spearman’s correlation tests with the whole group and explored by sex.
Results
As presented in Table 1, the sample included 29 White Caucasian older adults (median 72.0 years old), the majority of which were female (65.5%).
Participant Characteristics at Baseline (n = 29).
Note. Data presented as median (25th; 75th) or N (%).
All participants reported their gender (male/female/other) to be the same as their biological sex (male/female/intersex). No participants chose the “other” option for gender or the “intersex” option for their biological sex, making the sample 100% cisgender. Participants scored a median (interquartile) score of −13.0 (−19.5; −8.5) on the BSRI-30. Participants spent a median of 324.0 min performing moderate-to-vigorous activity per week at baseline. Most participants were married (65.5%), retired (79.3%), and highly educated (62.1%).
Table 2 presents the physiological and functional measures collected at baseline for both males and females.
Objective Baseline Variables Stratified by Self-Reported Gender (n = 29).
Note. Data presented as median (25th–75th).
Of the 29 participants (10 males, 19 females) enrolled in the program, 17 completed the intervention (5 males, 12 females). Figure 1 presents the gender role scores for participants who completed the intervention and those who did not complete the intervention. Participants’ scores on the BSRI-30 were not different between the two groups (r = .10, p = .62).
BSRI-30 score (masculinity-femininity) of participants who completed the program and those who did not (n = 29).
The percent differences of changes observed during the program are presented by gender in Table 3.
Percent Change in Studied Outcomes During the Intervention Stratified by Self-Reported Gender (n = 17).
Note. Data presented as median % (25th; 75th).
When men and women were analyzed together, no significant correlations were found between the score on the BSRI-30 and any of the percent differences. However, looking at men alone (n = 5), significant correlations were found between the BSRI-30 score and the percent change for the one-leg stance test (r = .90, p = .04) and the percent change for the estimated 1-repetition maximum chest press (r = −.96, p = .04). Looking at women alone (n = 12), significant correlations were found between the BSRI-30 score and the percent change for the leg extension peak power (r = −.62, p = .03), and the percent difference for the pain domain on the SF-36 (r = .63, p = .03).
Discussion
This study suggests that older adults engaging in outdoor exercise structure programs present more feminine gender traits, regardless of sex. In this relatively small sample, gender roles were not related to completing the program or any studied outcomes following a 6-week outdoor exercise structure program. These results are important because of the need to understand more about the role of gender in responding to various interventions.
Contrary to our findings, a previous study analyzing the relationship between gender roles and exercise frequency found significant correlations performed on female university students (Hyuk et al., 2012). This lack of relationship in our study between the completion of the intervention and gender roles could be explained by the limited range of scores on the BSRI-30 for the participants in this study (ranging from −34.0 to 4.0) as well as the limited number of participants to perform analysis on (n = 29), compared to the study performed by Hyuk et al. (n = 170). Another notable difference could also be that Hyuk et al. (2012) only studied females (Hyuk et al., 2012).
The secondary outcome of this analysis was to analyze the relationship between participants’ gender scores on the BSRI-30 and their physiological and functional changes comparing baseline to post-intervention. Participants’ scores on the BSRI-30 were not correlated with any of the studied variables. Previous studies analyzing the relationship between gender roles and physiological and functional outcomes have found conflicting results (Ahmed et al., 2016). These studies, for example, include Ahmed et al. (2016), who reported that feminine and androgynous roles were independently associated with mobility disability and low physical performance in older adults (Ahmed et al., 2016). Another study by Gale-Ross et al. (2009), however, found no correlation between participants’ scores on the BSRI and life satisfaction, physical health functioning, or general wellness (Gale-Ross et al., 2009). One possible reason that Ahmed et al. (2016) found a significant correlation while both Gale-Ross et al. (2009) and this study did not could be due to the sample size tested (Ahmed et al., 2016; Gale-Ross et al., 2009) allowing for a wider range for gender roles and more power. The study performed by Ahmed et al. (2016) had a large group (n = 1,995), while Gale-Ross et al. (2009) and our study had smaller sample sizes (Ahmed et al., 2016; Gale-Ross et al., 2009).
Finally, the lack of correlation observed in this study between changes in outcomes and gender roles could be due to participants’ initial characteristics, with more than 75% of the participants meeting the aerobic physical activity guidelines of a minimum of 150 min of moderate to vigorous intensity at baseline (Canadian Society for Exercise Physiology, 2021). As the secondary outcome of this study was looking at changes in the studied outcomes after a 6-week exercise intervention, this initial high physical activity level of participants made it more challenging to see a difference in outcomes.
Interestingly, when separated by sex, correlations were found in different outcomes. For example, females’ scores on the BSRI-30 were found to be correlated to their leg extension peak power, and their SF-36 pain domain percent change. Males’ scores on the BSRI-30 were correlated to their chest press estimated 1-repetition maximum and their one-leg stance test percent change. Although interesting, an appropriate design will be needed to study this hypothesis. However, our findings suggest that researchers need to pursue gender-based analysis and design studies to test specifically the role of gender by sex.
Despite the novelty of the research questions, this study has some limitations. The first is the tool assessing gender roles, the BSRI-30. The BSRI-30 was derived from the BSRI-60, created by Sandra Bem in 1974 (Bem, 1974b). The BSRI-30 lacks validity. Recent validity studies have been performed on the 60-item and 12-item versions of the BSRI questionnaire, but we could not find any recent studies validating the BSRI-30 (23,32). Only two studies were found reporting the BSRI-30 validity, one from 1997 and the other from 1988 (Campbell et al., 1997; Holmbeck & Bale, 1988).
In addition, the recommended scoring for the BSRI-30 is not available. As a result, the scoring needs to be based on the BSRI-60, which is readily available (Bem, 1974b). The third limitation of the BSRI-30 is that although this tool was generally accurate and reliable at measuring masculine and feminine sex role traits at the time of release, society and perceptions of gender and gender norms have changed considerably since its creation (Bem, 1974b; Choi & Fuqua, 2003). This shift in perception of gender has been thought to decrease the accuracy of the BSRI, as femininity and masculinity and their associated stereotypical norms are no longer defined in the same ways (Agbayani & Min, 2006; Choi & Fuqua, 2003). Although the BSRI-30 is not ideal for today’s society and standards, it is the best-accredited tool currently available to assess gender and was therefore used in this study. The final limitation regarding the BSRI in this study is that there was very little variation in the scores since most participants scored more feminine and had a very limited range.
Conclusion
Overall, there is still much to be explored about gender, gender roles, and their implications for older adults’ participation in exercise programs and health outcomes. This study suggests that people participating in such programs tend to have more stereotypical gender roles considered to be feminine. However, the gender roles of participants were not associated with the odds of completing the intervention or any changes in the studied outcomes.
Footnotes
Acknowledgements
No additional acknowledgments to report.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Ethics
Before the study began, this project was reviewed and approved by the University of New Brunswick Research Ethics Board under the file REB #2022-080
