Sex Roles

, Volume 57, Issue 11, pp 837–849

Midlife Women’s Physical Activity Goals: Sociocultural Influences and Effects on Behavioral Regulation

Authors

    • Institute for Research on Women and GenderThe University of Michigan
  • Jacquelynne S. Eccles
    • Institute for Social ResearchUniversity of Michigan
  • Stephen C. Peck
    • Institute for Social ResearchUniversity of Michigan
  • Caroline R. Richardson
    • Department of Family MedicineUniversity of Michigan
Original Article

DOI: 10.1007/s11199-007-9322-1

Cite this article as:
Segar, M.L., Eccles, J.S., Peck, S.C. et al. Sex Roles (2007) 57: 837. doi:10.1007/s11199-007-9322-1

Abstract

A better understanding of midlife women’s exercise goals could offer insight into the psychosocial facilitators and barriers to their participation. A random sample of US working women (40–60 years) was taken, and resulted in 262 participants. Cluster analysis identified participants’ most important physical activity goal. A five goal cluster solution was identified: (1) sense of well-being; (2) weight loss; (3) health benefits; (4) stress reduction; and (5) weight maintenance/toning. The goal cluster solution was validated and suggested relevant sociocultural influences and potential effects on behavioral regulation (p < .05). Data showed that weight- and health-related goals have sociocultural influences and are associated with more extrinsic and less intrinsic behavioral regulation compared to stress reduction and sense of well-being goals.

Keywords

Physical activityExerciseWomenGoalsBehavioral regulation

Introduction

Midlife women have an elevated risk of being sedentary compared to younger women and men (Scharff et al. 1999; US Department of Health and Human Services 1996). Being regularly physically active is associated with reduced risk of developing many chronic conditions (e.g., osteoporosis, diabetes, depression, cardiovascular disease; US Department of Health and Human Services 1996, 2004). In order to understand how to most effectively promote physical activity, there have been calls to conduct research targeting gender and different life stages (Booth et al. 1997; Dishman et al. 1985; Marcus and Forsyth 1998; Marcus et al. 2000; Scharff et al. 1999; Trujillo et al. 2004). This study investigates the physical activity goals of women in midlife.

Physical activity goals are important to research because behavioral goals, or the reasons for doing a particular behavior, are associated with motivation, self-regulation, and adherence (Bagozzi and Edwards 2000; Carver and Scheier 1998, 2000). This study is important because a better understanding of midlife women’s physical activity goal content offers insight into the motivational facilitators and barriers to their participation, and potentially aid in the development of better promotion strategies for this population. The primary objective of this research was to expand previous research about the reasons why midlife women participate in physical activity (Segar et al. 2006), and aimed to illuminate the specific content of the physical activity goals midlife women report having using cluster analysis. The secondary objective was to validate these clusters using variables that could highlight sociocultural influences on the content of midlife women’s physical activity goals. We validated the clusters using variables reflecting norms and pressures related to beauty/thinness, such as self-objectifying (Fredrickson and Roberts 1997) and variables representing intrinsic and introjected behavioral regulation as discussed in self-determination theory (Deci and Ryan 1985).

Taylor et al. (2006) suggested that there is a deficit in the literature because most research on goal setting has investigated the implications of goal setting, rather than the bases of goal setting. We agree with this critique, but also believe that the effects of goals on behavior cannot be understood without examining how goal content has been influenced by the surrounding cultural context. Because behavioral goals are closely connected to cultural expectations (Coole 1995; Eccles 1994; Markus and Kitayama 1991), it is important to contextualize the psychological meaning of physical activity goals within the sociocultural context. The specific physical activity goal that a woman has is likely influenced by the internalization of cultural norms and gendered pressures (Eccles 1994; Henderson and Bialeschki 1994; Markus and Kitayama 1991).

Women report exercising in order to lose weight, shape their bodies, or improve their appearance more than men (Frederick and Ryan 1993; McDonald and Thompson 1992; Silberstein et al. 1988; Tiggemann and Williamson 2000). Physical activity goals related to appearance and body shape may be manifestations of women’s internalization of sociocultural beauty norms and physical activity meanings portrayed in the media (Mutrie and Choi 2000; Theberge 1997) and reflect a woman’s attempt to achieve her “ideal self;” one that is culturally prescribed and internalized (Carver 1996; Eccles 1994). Objectification theory and self-determination theory provide frameworks to investigate how appearance and body-shape physical activity goals might detrimentally influence midlife women’s behavioral regulation and motivation for being physically active.

Objectification theory posits that women are socialized in our culture to consider themselves as objects that are evaluated on the basis of their appearance and to internalize an objectifying observer’s perspective on their own body (Fredrickson and Roberts 1997). Self-objectification (focusing on “How do I look?” to another person rather than a first-person perspective, for example, “How do I feel?”) leads to habitual body surveillance.

Research on self-objectification has suggested that self-objectifying is associated with negative consequences for women, including negative affect, shame, less awareness of internal bodily states and decreased positive motivational states (Fredrickson and Roberts 1997; Fredrickson et al. 1998).

One study investigated the relationship between self-objectification and women’s reasons for exercise. This study showed that self-objectification was positively associated with appearance motives for exercise among college-aged women (Strelan et al. 2003). These authors also reported that women who rated appearance-related reasons for exercise as important (e.g., weight control, body tone, and attractiveness) experienced increased body dissatisfaction and reduced self-esteem. Generalizing from these data and objectification theory, a midlife woman with physical activity goals related to appearance and body-shaping would be likely to self-objectify, and potentially experience extrinsic behavioral regulation.

Sociocultural thinness and beauty norms and pressures provide the context for, and are a likely source of influence in women selecting physical activity goals related to appearance and/or shaping their bodies. Predominant images of physically active women in media target improvements in body shape and appearance that can be gained from exercising (Blaine and McElroy 2002; Theberge 1997). Thus, the media, and the sociocultural norms and pressures they depict, provide a context within which women develop their physical activity goals and behavioral regulation. Importantly, the context in which goals are set influences the type of behavioral regulation that develops (Ryan and Deci 2002). According to self-determination theory, when individuals take prompts from their environment and partially internalize them, they have an introjected form of behavioral regulation. This type of regulation is characterized by an individual feeling conflicted about the behavior (Skinner and Edge 2002).

Introjected regulation represents an external regulation that has been partially, but not fully, internalized by an individual and causes them to feel pressured to perform the behavior (Reeve 2002). Behaviors that are introjected are performed to gain social approval or avoid disapproval, or to gain self-worth (Biddle and Mutrie 2001; Ryan and Deci 2000). Introjected regulation is characterized by the sentiment that ‘I should exercise’ (Ryan et al. 1992), and is associated with decreased positive affect, lower intrinsic motivation, and worse adherence (Deci and Ryan 1985; Frederick-Recascino 2002). According to self-determination theory, body-shape physical activity goals would be experienced as controlling rather than autonomous, and as a result, should undermine midlife women’s behavioral regulation and motivation (Ryan et al. 1992). In contrast, autonomously derived goals reflect a person’s true sense of volition and personal importance. These goals are more likely to be intrinsically motivating and result in better behavioral participation (Ryan and Deci 2000). Moreover, research has shown that when individuals imagine being physically active, imagery that is related to appearance is most strongly associated with introjected regulation, compared to the other types of regulation (Wilson et al. 2003). The ideas inherent in both self-determination and objectification theories suggest that physical activity goals derived out of the internalization of sociocultural norms and pressures related to appearance and body shape would result in less optimal motivation and behavioral regulation.

We hypothesize that: compared to participants with non-body-shape goals, participants with physical activity goals related to appearance and weight (body-shape goals) will: (1) have more physical activity imagery related to body shape; (2) self-objectify more; (3) have higher introjected regulation; and (4) lower intrinsic regulation.

Method

Sample and Procedure

The population of interest is midlife women who work full time. Women have less leisure time than men (Verhoef et al. 1992) and women over 50 have a substantial risk of being sedentary (US Department of Health and Human Services 1996). Women who work full time have added constraints on their ability to have leisure time, making it even more difficult to regularly participate in physical activity. Given that the majority of women work (United States Department of Labor 2003), research is needed that identifies what helps and hinders them from participating in sustained physical activity. A convenience sample was taken of the University of Michigan clerical staff to participate in a mailed survey. Inclusionary criteria were: Being between the ages of 40 and 60; working in full-time jobs that are administrative; having Internet access and an e-mail account. The age range of 40 to 60 was selected because it is considered to constitute the “core” of midlife (Keyes and Ryff 1999).

Out of the sample population of 843 employees, a random sample of four hundred women (age 40–60) was selected to participate in a mailed survey. Fifteen of the 400 were ineligible to participate because they had either taken the pilot survey, were administrators involved in the study, or were no longer employed by the university. The response rate for the survey was 71% (n = 275). Thirteen cases had incomplete data and were removed. This sample had 262 participants. However, four cases were removed prior to the analyses (i.e., two cases had incomplete data on the goal measure, and two cases were identified by the Sleipner program as multivariate outliers). The subsequent cluster analyses were conducted on 258 cases. There were no demographic differences between survey responders and non-responders. See Table 1 for the sample demographics. The sample is mostly European American and married.
Table 1

Baseline sample demographics.

Factors

Percent

 

Total Sample

N = 275

Age

49.9, SD = 5.4

BMI

28.0, SD = 6.4

Education

High School or GED

Some College

Technical College

College Degree

Grad/Prof Degree

Missing

10.5%

38.0%

5.5%

36.4%

9.1%

0.4%

Marital Status

Married

Living with partner

Separated

Divorced

Widowed

Single

62.5%

4.4%

1.1%

20.0%

2.2%

9.8%

Household Income

<$20,000

$20,000–$60,000

$60,001–$100,000

$100,001–$124,999

$125,000+

Missing

0.7%

38.5%

38.9%

10.5%

6.9%

4.4%

Ethnicity

African–American

Asian

European Amer.

Latina

Mixed ethnicities

Missing

5.1%

2.2%

89.5%

1.1%

1.5%

0.7%

The University of Michigan Human Resource Department (UMHR) conducted a database query for staff who matched the study criteria listed above. The UMHR provided mailing labels and e-mail addresses from a random sample of those who matched study criteria to the principal investigator.

The potential participants received an e-mail one week before the survey was sent, informing them that they were selected to participate in a study and that an information packet with a survey would arrive soon. One week later, potential study participants received a survey. Participants not completing their survey received e-mail inquiries on days 7, 14, and 21. After that, these participants were considered non-responders. This study received Human Subjects approval from the University of Michigan Institutional Review Board.

Measure

Physical Activity Goals

The following eighteen goals were obtained from a comprehensive review of the literature: stress reduction, weight maintenance, energy level, tone/shape body, sense of well-being, weight loss, social reasons, overall health benefits, flexibility, competition, disease prevention, family time, enjoyment, cardiovascular fitness, endurance/stamina, strength, fun, and other (Campbell et al. 2001; Cash et al. 1994; Duda and Tappe 1988; Finkenberg et al. 1994; Frederick and Morrison 1996; Gill and Overdorf 1994; Ryan et al. 1997; Trujillo et al. 2004).

To measure participants’ physical activity goals, participants were asked to answer in stages. First, they were asked to review the list of goals and “Circle the reasons below that represent your three most important reasons for being physically active.” Second, they were instructed to “Go back to the list, and underline the reasons that represent your three least important reasons for being physically active.” Third, they were asked to “Review the three reasons you circled above as being most important to you for being physically active, and write the one that you consider to be the most important reason.” Fourth, they were asked to “Review the three reasons you underlined as being least important to you for being physically active and write the one that you consider to be theleast important reason” (Block 1971; Kohn and Schooler 1983). Each of the goal scores were then recoded into the following: −2 = “least important goal”, −1 = “two second least important goals”; 0 = goals not mentioned, 1 = two second most important goals, and 2 = “most important goal.”

Physical Activity Imagery

Physical activity imagery is the first of two variables that we used to measure sociocultural influences on goals. To avoid priming participants, early in the questionnaire, they were instructed to imagine being physically active for 1 to 2 minutes. An open-ended, inductive imagination task that elicits specific thoughts, feelings, associations and goals was used. This measure has been used previously and has predictive validity (Scioli et al. 2000; Segar et al. 2006). Specifically, study participants were asked to “Take some time and imagine that you are being physically active right now. Please close your eyes, and take a minute or two to go over that experience in your mind. After you do this, please write below three to four sentences describing what you were thinking about and what you were feeling when you were being physically active. If you prefer, you can also write phrases or just words to describe your thoughts and feelings.”

Physical Activity Imagery relates to what participants imagined during the above imagery activity. This variable was created by coding participants’ narrative, described above, in addition to the following two questions: (1) “In the imagination task you just did, why were you being physically active?”; (2) “In the imagination task you just did, what would your reason(s) be for choosing to participate in that specific physical activity?” Participants were coded as having “body-shape” imagery if they wrote any of the following words (“shape”, “calories”, “toning”, and/or indicating that they wanted to lose or maintain their weight). Participants received a code of “body-shape” regardless of whether other goals were mentioned in addition to body shape goals. Participants not mentioning those words were coded as having “non-body-shape” imagery.

Self-Objectification

The second variable that we used to measure sociocultural influences on goals, self-objectification, was measured using the body surveillance subscale from the Objectified Body Consciousness Scale (McKinley 1996). Participants were asked how much they agreed with eight statements (e.g., “During the day, I think about how I look many times”), on a seven-point Likert scale from (1) Strongly disagree to (7) Strongly agree. The Self-Objectification scale was created by taking the mean of the items (including the one item that was reverse scored). The inter-item correlation of the eight-item scale in other research was reported to be .76 (McKinley 1999). Scores ranged from 1 to 7, with an average of 4.4 (SD = 1.0). Inter-item reliability was adequate, α = .86. A higher score indicates higher self-objectification (e.g., thinking about one’s body more in terms of how it looks than how it feels).

Introjected Regulation

Participants answered questions related to behavioral regulation about the most important goal that they selected. To measure introjected regulation toward participants’ most important physical activity goal, participants responded to “How important are each of the following two reasons regarding why you pursue this goal?” Possible responses ranged from (1) Not at all important to (7) Very important to the following two questions: 1. I feel that I should pursue it; and 2. I would feel guilty if I didn’t pursue it.

The Introjected Regulation scale for the most important physical activity goal was created by taking the mean of these two items. Scores ranged from 1 to 7, with an average of 5.2 (SD = 1.3). Inter-item reliability = .63. Higher scores indicate that participants had greater introjected motivation for pursing their most important physical activity goal.

Intrinsic Regulation

To measure intrinsic regulation toward participants’ most important physical activity goal, two items were used. Participants were asked: 1. How much do you “enjoy pursuing this goal?” Responses ranged from (1) Do not enjoy at all to (7) Enjoy a lot; and 2. How important is the following reason regarding why you pursue this goal? (“I like pursuing it.”) Responses ranged from (1) not at all important to (7) very important.

The Intrinsic Regulation scale was created by averaging these two items. Scores ranged from 1 to 7, with an average of 4.8 (SD = 1.6). Inter-item reliability = .85. Higher scores indicate greater intrinsic regulation.

Analyses

Identifying Homogeneous Subgroups Using Cluster Analysis

Cluster analysis permitted us to create comprehensive and in-depth motivational profiles based on the specific set of personal goals that were both important and unimportant to participants for their activity profiles. Cluster analysis was used to identify patterns of physical activity goals using the Sleipner statistical package (version 2.1) for pattern-oriented analyses (Bergman and El-Khouri 2002). Using the Sleipner Residan procedure, a small percentage of multivariate outliers were identified and removed prior to the cluster analysis. (The two cases removed exhibit profile patterns that were unique and unlikely to belong to any of the homogeneous subgroups identified using cluster analysis.)

We then used Ward’s method (with squared Euclidean distances as the dissimilarity measures) to generate 20 possible cluster solutions (Bergman et al. 2003; Everitt et al. 2001). There was a fusion (i.e., “agglomeration”) coefficient associated with each of the 20 possible cluster solutions, and we arranged these coefficients in a scree-type plot to determine the upper and lower bounds of statistically justifiable cluster solution complexity (which ranged from three to five cluster groups). As described below, we selected an optimal cluster solution by reference to both differences in the explained error sum of squares between solutions and theoretical considerations. After selecting our preferred solution, we conducted a k-means cluster relocation analysis (which used the results of the Ward’s method analysis as start values) that resulted in 27 cases being reassigned to cluster groups that better matched their profile. This step corrects for early misclassifications in the strictly hierarchical procedure of Ward’s method and maximizes within-group homogeneity.

All of the goals mentioned above, except for “competition” and those that were handwritten in as “other” goals, were used in the cluster analyses. Competition was excluded because it appeared to be irrelevant compared to the other goals as a physical activity goal for midlife women; that is, over 85% of study participants endorsed it as one of their least important reasons for being physically active, and there was no variance across preliminary cluster solutions that included this variable. The cluster analyses were conducted using the remaining 16 goals.

Validation of Cluster Solution

It is important to validate a cluster solution by showing that groups differ in predictive ways (Aldenderfer and Blashfield 1984). This can be done by demonstrating meaningful variability between clusters using variables that are directly related to the cluster group patterns but not the same variables that were used to create the cluster solution (Bergman et al. 2003; Everitt et al. 2001; Nelson et al. 2005; Reedy et al. 2005). ANOVAs, chi-square analyses, and t tests were conducted to validate the cluster group membership. Post hoc analyses were conducted using the least significant difference (LSD) procedure. Effect sizes for the analyses were calculated according to recommendations (Polit and Sherman 1990; Cohen 1988).

Results

Physical Activity Goal Cluster Solution

A final five-cluster solution was retained based on the interpretability of the profile patterns and evidence from the scree-type plot showing minimal gains in explained variance between cluster groups for higher-level solutions. See Fig. 1 for a graphical depiction of the five-cluster solution. As can be seen in Fig. 1, the five cluster group patterns was marked by a dominant goal and was named according to these marker variables. Some of the clusters can also be identified by the least important goals. It should be noted that “being social” is a dominant least important goal in all of the cluster groups and thus will not be included in the cluster definitions below. “Being social” was not removed from the cluster analysis, as was “competition”, because, although it was a dominant least important motive, it was not endorsed by the majority of the sample (44%) compared to “competition” (85%). The order of the goal clusters follows the order seen in Fig. 1. The sense of well-being cluster is characterized by participants endorsing one goal, “sense of well-being” (n = 31, 12%). This goal cluster also has “family time” as the least important exercise goal. The weight loss cluster is characterized by participants scoring high on the “weight loss” goal (n = 55, 21%) and having both “fun” and “enjoyment” as dominant least important physical activity goals. Participants in the stress reduction cluster rated “stress reduction” as their most important goal (n = 37, 14%) and “weight loss” as a least important physical activity goal. The majority of participants endorsed health reasons for being physically active. Participants in the health benefits cluster (n = 106, 40%) listed “health benefits” as their most important physical activity goal and “weight loss” as their second most important goal. Participants in the weight maintenance/toning cluster (n = 29, 11%) endorsed “weight maintenance” and “toning” equally as the most important goals, and endorsed “health benefits” as a secondary goal for participating. They rated “fun” as a least important goal for being physically active. In summary, the majority of participants (61%) reported that their most important goal for being physically active was to lose weight or to obtain health benefits.
https://static-content.springer.com/image/art%3A10.1007%2Fs11199-007-9322-1/MediaObjects/11199_2007_9322_Fig1_HTML.gif
Fig. 1

Physical activity goal cluster solution. The variables we used to cluster are on the upper right hand corner of figure. The names of each cluster are on the x-axis and the mean levels of the variables used to create the clusters are along the y-axis. As can be seen, each cluster profile has a subset of variables that distinguish it from the others. The label we gave each cluster is reflective of the meaning of the profile of each cluster group. For example, the “sense of well-being” cluster is clearly reflective of the fact that it has a very dominant goal, sense of well-being. In contrast to the other goal clusters, the fifth cluster depicted “weight maintenance/toning” has two dominant goals, weight maintenance and toning. Thus, this cluster’s name reflects the fact that there are two dominant goals and not one, as is the case with the other clusters.

Differences in demographics between cluster groups were investigated using chi-square tests and one-way ANOVAs for categorical and continuous variables, respectively. Members of different goal clusters did not vary on age [F(4, 250) = 1.4, p = .25], education level (χ2 = 22.4, df = 1, p = .13), marital status (χ2 = 13.9, df = 1, p = .84), or ethnicity (χ2 = 8.6, df = 1, p = .93). These null findings are not surprising given the homogeneity of the population from which the sample was drawn. Members of different goal clusters did differ significantly on the BMI [F(4, 249) = 6.5, p < .001], with a large effect size = .41 (Cohen 1977). Weight loss participants had a mean BMI of 31.4 (SD = 6.7), health benefits participants had a mean BMI of 29.0 (SD = 6.9), sense of well-being participants had a mean BMI of 26.9 (SD = 4.0), weight maintenance/toning participants had a mean BMI of 26.4 (SD = 5.2), and stress reduction participants had a mean BMI of 25.5 (SD = 6.0). Participants in the weight loss cluster were heavier than all of the other participants, p < .01. Participants in the health benefits cluster weighted significantly more than those in stress reduction and weight maintenance/toning, p < .05, and marginally more than those in the sense of well-being cluster, p < .10.

Sociocultural Influences on Goal Content

Physical Activity Imagery

To investigate the association between physical activity imagery and goal clusters, the clusters were divided into the two groups: (1) body-shape clusters (weight loss and weight maintenance/toning) and (2) non-body-shape clusters (stress reduction, sense of well-being, health benefits). Because four participants did not fill out the imagery task, the following analysis was conducted using 254 participants. A chi-square analysis showed a significant association between these goal clusters groupings and having body-shape imagery for being physically active from the imagination task (χ2 = 14.9, df = 1, p = .000), with a .24 effect, which would be considered between a small and medium effect size (Cohen 1988). See Table 2 for a summary of the findings. Only 22% of the non-body-shape cluster participants (n = 37) reported body-shape imagery compared to 45% of the body-shape cluster participants (n = 38). As predicted in hypothesis no. 1, the standardized residual of −3.9 indicates that a much smaller percentage of non-body-shape goal cluster participants had body-shape imagery than would be expected by chance. In addition, 78% of the non-body-shape cluster participants (n = 133) had non-body-shape imagery compared to only 55% of the body-shape cluster participants (n = 46). The standardized residual of 3.9 indicates that a much larger percentage of non-body-shape cluster participants had non-body-shape imagery than would be expected by chance. It should be noted that, despite these significant findings, the majority of the body-shape cluster participants (55%) reported non-body-shape imagery.
Table 2

Body-shape clusters vs non-body-shape clusters on imagined physical activity.

Imagined physical activity

Goal clusters

Non-body-shape

Body-shape

Total

Non-body-shape

N = 133

N = 37

N = 170

Row (%)

78.2

21.8

Column (%)

74.3

49.3

Total (%)

52.4

14.6

Adjusted residual

3.9

−3.9

Body-shape

N = 46

N = 38

N = 84

Row (%)

54.8

45.2

Column (%)

25.7

50.7

Total (%)

18.1

15.0

Adjusted residual

−3.9

3.9

Total

179

75

254

Self-Objectification

To investigate self-objectification, the goal clusters were divided into the same cluster groupings as previously described [e.g., body-shape goals (weight loss and weight maintenance/toning) and non-body-shape goals (stress reduction, sense of well-being, health benefits)]. An independent t test showed a significant relationship with Self-Objectification, t(256) = 2.9, p = .004, with a small to medium effect size = .40 (Cohen 1977). As predicted in hypothesis no. 2, the body-shape cluster group scored higher on Self-Objectification (4.8; SD = 1.0) compared to the non-body-shape cluster group (4.4; SD = 1.0).

Behavioral Regulation

Introjected Regulation

An analysis of variance showed a significant relationship between the five-category goal cluster variable and Introjected Regulation, F(4, 251) = 5.0, p = .001, and between a medium to large effect size = .36 (Cohen 1977). Partially supporting hypothesis no. 3, post hoc analyses showed that the weight maintenance/toning (5.6, SD = 1.1), weight loss (5.5, SD = 1.1), health benefits (5.4, SD = 1.1) cluster members scored significantly higher on Introjected Regulation than those in the stress reduction (4.9, SD = 1.7) and sense of well-being (4.5, SD = 1.5) clusters, p < .05.

Intrinsic Regulation

An analysis of variance showed a significant relationship between the five-category goal cluster variable and Intrinsic Regulation, F(4, 252) = 12.1, p = .001, and a large effect size = .51 (Cohen 1977). Partially supporting hypothesis no. 4, post hoc analyses showed that the participants in the weight loss cluster (3.6, SD = 1.6) scored significantly lower on Intrinsic Regulation than participants in all other clusters, p < .01). Weight maintenance/toning cluster group participants (4.7, SD = 1.3) scored significantly lower on Intrinsic Regulation than those in the stress reduction cluster (5.5, SD = 1.4), p < .05, and marginally lower than those in the sense of well-being cluster (5.4, SD = 1.4), p < .10. In addition, the health benefits cluster members (4.9, SD = 1.5) scored marginally lower on Intrinsic Regulation than those in the sense of well-being and stress reduction clusters, p < .10.

Discussion

The aims of this study were to expand previous research on midlife women’s reasons for being physically active (Segar et al. 2006) by investigating (a) the specific content of midlife women’s most important goals for being physically active; and (b) to validate the clusters using variables that highlight relevant sociocultural beauty/thinness norms and pressures that influence women’s physical activity goal selection and those that indicate intrinsic and introjected behavioral regulation.

Physical Activity Goal Prevalence Rates

We identified five distinct clusters related to midlife women’s most important goals for being physically active. It is interesting to note that only 5 dominant goal clusters emerged when there were 16 goals listed. This suggests that midlife women think about being physically active, in general, for a handful of reasons.

Almost one third of this sample reported being physically active to pursue goals related to body- and weight-related concerns. While 33% represents a significant endorsement of body-shape reasons for physical activity, it is lower than typically seen in samples of college-age women (Strelan et al. 2003; Tiggemann and Williamson 2000).

Having body-shape and/or appearance motives for physical activity might decline across the lifespan. Although women experience pressure from cultural beauty/thinness norms that result in body dissatisfaction throughout the lifespan (Tiggemann 2004; Tiggemann and Lynch 2001), the importance that women place on their body dissatisfaction, and its influence on why they exercise may decrease with age (Tiggemann 2004). Developmental perspectives on physical activity suggest that individuals in different life stages have distinct priorities and responsibilities that will be reflected in different reasons for being physically active (Trujillo et al. 2004). Compared to midlife women, most of whom are married and have families, college-age women are in a life stage where they are searching for a mate. These distinct life stages might make weight and appearance physical activity goals more salient for younger women than those in midlife.

Health benefits was endorsed by the most participants (40%). This is not surprising because health is highly endorsed as a motive for participation across age groups (Trujillo et al. 2004). Endorsing health benefits might represent a response influenced by a social desirability bias because “doing healthy things” could be considered a normative pressure in our culture. How individuals construe behavior is shaped in large part by what is learned from the professionals in the field and culture (Leventhal et al. 1998; Rothman et al. 2003). Individuals often view doing things that their healthcare providers recommend to be healthy as something they are “supposed to” or “should” do (Williams 2002). These sentiments reflect an introjected form of behavioral regulation whereby the individual acts to gain social approval, avoid social disapproval, or gain self-worth (Frederick-Recascino 2002). Not surprisingly, introjected regulation leads individuals to feel conflicted about doing the behavior, and is considered to be a regular occurrence in healthcare (Williams 2002). Our data support this idea by showing that the participants in the health benefits cluster scored high on introjected regulation.

Despite the known mental health benefits from regular physical activity (US Department of Health and Human Services 1996), only 26% of these participants are physically active for goals related to improving mental health and mood. It may be that most midlife women do not experience being physically active in a way that would improve their stress levels or sense of well-being. Physical activity and exercise have been marketed with a weight-centered and appearance-oriented focus to women (Blaine and McElroy 2002; Theberge 1997), and may influence women self-objectifying while exercising (Wolfe 1999). Because self-objectifying results in body vigilance, self-consciousness, and negatively influences mood (Fredrickson and Roberts 1997; Roberts 2002), being physically active might not be a positive mental health experience for women who self-objectify while being active. This might be especially true for women who exercise with weight-related goals because our data suggest these women self-objectify more than those who have goals that are not related to weight and toning.

In addition, many midlife women might not be aware that being physically active can have mental health benefits. Given the way exercise has been marketed to women, a woman who is not physically active might not know that regular participation can result in enhanced feelings of well-being or improved mental health. This idea is supported by study findings showing that exercise was used as a mood enhancer only by individuals who were regular, experienced exercisers, and not those who were not regular participants (Hsiao and Thayer 1998).

Another interesting thing to note is that these participants ranked “being social” and “family time” as least important goals for being physical activity. This lack of association between others/family and physical activity among this sample is important to note. Midlife women might not have this socially oriented motivation for being physically active because they were socialized before Title IX in 1972. Because of this, midlife women grew up before gender equity in sports was mandated by law, and many did not have opportunities to participate in team sports (Lutter 1994). Furthermore, because they were socialized before Title IX, much of midlife women’s socialization to being physically active may have been from marketing that promotes exercise as a vehicle for an individual woman to lose weight and sculpt her body (Theberge 1997). Both of these reasons might contribute to midlife women not having developed a socially oriented attitude toward being physically active. Research has also shown that, among midlife adults, relationships are primary for enjoying life (Brim et al. 2004). Therefore, if physical activity is not related to aspects of their social and/or family life, midlife women may be less likely to participate. The commercial success of Curves® may be related to the social benefits it provides to its female patrons (Curves 2006).

It is important to note that no clusters emerged that represented the enjoyment of being physically active in this sample of midlife women. This means that there were not a sufficient number of participants who endorsed enjoyment as an important reason for being physically active for a cluster to develop. It may be that many midlife women do not enjoy being physically active. For an individual who is middle aged, overweight, and infrequently active, when they are physically active, especially at higher intensities, it might be physically unpleasant and frustrating, rather than mood enhancing and enjoyable (Ekkekakis and Lind 2006). The lack of enjoyment of physical activity as a reason for doing it is important to note because long-term participation is thought to be related to enjoying physical activity (Biddle and Mutrie 2001; Buckworth 2000; Ingledew et al. 1998; Ryan et al. 1997; Salmon et al. 2003). Our data suggest that sense of well-being and stress reduction may be more relevant than enjoyment as intrinsic motives for midlife women.

Goal Validation: Sociocultural Influences and Effects on Behavioral Regulation

These results support the validity of the goal cluster groupings. In addition, as hypothesized, these data suggest that participants with body-shape goals have more greatly internalized cultural norms related to appearance and exercising than those with non-body-shape goals, supporting the notion that there are sociocultural influences on midlife women’s physical activity goal content. In addition, these findings suggest that distinct physical activity goals are associated with different types of behavioral regulation among midlife women.

Our findings suggest that midlife women with body-shape physical activity goals were influenced to have these particular goals by the surrounding sociocultural context. These particular participants reported having more imagery related to appearance and body shape, and self-objectified more than the participants with goals that were not related to appearance and body shape. This study is the first research we know of to suggest that there is an association between self-objectification and body-shape physical activity goals among midlife women. Our data support other research on college-age women, and a mixed sample of males and females (ages 18–35) showing that self-objectification was positively associated with exercising for appearance reasons (Strelan et al. 2003). These findings support our contention that self-objectification is associated with women exercising for body-shaping reasons, and suggest that this relationship exists at different life stages for women.

Having physical activity goals that are associated with self-objectification might be problematic because of potentially negative consequences for motivation and behavioral regulation. Self-determination theory, broadly, distinguishes between extrinsic and intrinsic motivation and types of regulation (Ryan and Deci 2000). Intrinsic regulation is an autonomous experience and represents the state of doing an activity out of interest or inherent satisfaction. According to self-determination theory there are distinct types of extrinsic regulation. However, because this study was specifically interested in the effects of internalizing sociocultural norms on midlife women’s physical activity goal development, we focused on the introjected form of extrinsic regulation; a partially internalized, self-enforced, and coercive-type of extrinsic regulation that leads women to feel like they “should” be physically active. We contend that women whose physical activity goals are influenced by sociocultural norms and pressures (e.g., weight and appearance related) should have high introjected regulation. Our data showed that participants with weight-related physical activity goals exhibited the highest introjected and the lowest intrinsic regulation levels. This is not surprising because these participants also endorsed enjoyment and/or fun as least important reasons for being physically active. Thus, participants with weight-related physical activity goals were more likely to feel that they should pursue their physical activity goal rather than enjoy pursuing it, compared to participants with goals that are more autonomous (e.g., stress reduction or sense of well-being). These results had medium to large effect sizes, and indicate robust relationships between the type of physical activity goal one has and their behavioral regulation.

Contrary to our predictions, these data showed that the participants in the health benefits cluster scored equally high on introjected regulation as did those in the weight loss and weight maintenance/toning clusters. Introjected regulation can lead people to try to coerce themselves to perform a behavior, often through guilt, shame, and internal pressure (Vallerand and Ratelle 2002). These findings suggest that being physically active for body-shape or health motives might be experienced by midlife women as being coercive, something that they “should” or “ought” to do (Carver and Scheier 1998; Ryan and Connell 1989). It is interesting to note that weight maintenance/toning cluster members reported the highest level of introjected regulation. Because their cluster profile suggests that health might be a secondary reason for their physical activity, these participants might experience introjected regulation related to both their body-shape and health-related goals. These data are in line with self-determination theory’s contention that the social environment influences the types of goals and behavioral regulation that individuals develop (Deci and Ryan 2000).

Feelings of pressure can derive from distinct sources; cultural ideals and a healthcare practitioner’s behavior change recommendation. In line with national statistics (Hill 2005), most of these midlife participants were overweight or obese. Given that most midlife women in the United States are overweight or obese, midlife women might feel pressured by their healthcare providers to exercise for weight control or health motives. Thus, midlife women may be especially vulnerable to feeling coerced to exercise, and as a result, experience high levels of introjected regulation.

We measured women’s explicit, and not their implicit, goals and processes in this study. However, women might not be conscious that their physical activity goals and motivations for participating in physical activity are influenced by sociocultural gender norms and culturally prescribed physical activity motives because non-conscious goals affect behavior without the individual’s conscious awareness (Bargh et al. 2001; Henderson and Bialeschki 1994). Socialization is an implicit process that frequently occurs outside of awareness (Coole 1995; Forgas 1981; Markus et al. 1996; Mead 1934). Given that women’s socialization about being physically active might be to some extent outside of their awareness (Forgas 1981), promotion efforts to increase physical activity in midlife women may obtain better long-term results if these efforts specifically address how midlife women have been socialized to think about physical activity and why they should participate in it.

These data suggest that social marketing campaigns and behavioral interventions might better facilitate increased physical activity among midlife women if they address and acknowledge the internalization of cultural norms and pressures (e.g., beauty/thinness and practicing health behaviors) as they impact midlife women’s relationship with (e.g., introjected regulation) and specific goals for being physically active. This suggestion seems especially relevant given that the majority of this sample reported having goals related to weight or health. This line of reasoning has been supported by longitudinal data from a physical activity intervention targeting midlife women. This 6-week, community-based program raised participants’ awareness to how they were socialized regarding exercise and physical activity, especially as it influenced their reasons and goals for participating in physical activity. Compared to their pre-program participation level (e.g., baseline), program participants were more physically active at the post-program data collection (6 weeks), which was maintained at the study follow-up (average 10 months post-program; Segar et al. 2002). That these participants maintained their higher physical activity level at the study follow-up is notable because most physical activity intervention effects do not last beyond the end of the intervention (Baranowski et al. 1998; Buckworth 2000).

Study Strengths and Limitations

The sample size and response rate are strengths of this study. Data were analyzed from participants who constitute an under-researched subpopulation. This study used person-centered methods to identify the content of midlife women’s goals for being physically active. We triangulated the data by using inductive and deductive methods, and quantitative and qualitative measures to measure the variables in this study. Triangulation is thought to offer a more comprehensive view of the phenomena under study (Biddle et al. 2001; Patton 1990).

These data are correlational and the causal inferences that we suggest are tentative, and need to be studied further using a longitudinal design. In particular, the idea that having physical activity goals related to health is associated with introjected regulation needs to be studied further. In addition, the findings cannot be generalized outside of this sample of mostly European American and university clerical staff who work full time in a Midwestern university. It is important to replicate this goal cluster solution among a similar sample, and to investigate the potentially different physical activity goals cluster solutions that women of distinct ethnicities, socio-economic statuses, and sexual orientations have.

Summary and Conclusion

These participants reported being physically active for mainly health and weight reduction goals. These data suggested that health- and weight-related physical activity goals are more related to social norms and associated with less optimal behavioral regulation compared to goals related to sense of well-being and stress reduction. These findings are important to consider because there is a body of research suggesting that autonomous regulations, intrinsic motivation and persistence in physical activity behavior may be undermined when women initiate physical activity out of compliance to sociocultural norms and pressures (Deci and Ryan 2000; Eccles 2005; Segar et al. 2006).

Acknowledgements

This research was partially funded by an American Fellowship from the American Association of University Women Educational Foundation and a grant from the Blue Cross and Blue Shield of Michigan Foundation. The authors wish to thank Lindsay Anonich, Leslie DePietro, Barbara Fredrickson, Deanna Maida, Sara Nazem, Susan Nolen-Hoeksema, Diana Perpich, Lori Rudy, and Brady West. We thank the study participants and the anonymous reviewers for their very helpful comments that improved our paper.

Copyright information

© Springer Science+Business Media, LLC 2007