Key words
Self-determination theory - Habitual behaviour - Motivation - Physical activity -
Maintenance
Regular physical activity is a modifiable behavior that is associated with multiple
health benefits [19]. Unfortunately, 47% of Americans do not meet national recommendations for physical
activity and 39% are completely sedentary [21]
[25]. Moreover, physical activity rates decline as an individual ages [11]
[18]. To date, it remains unclear why some individuals fail to maintain physical activity
while others succeed [11]. Specific to physical activity interventions, the majority that targeted long-term
maintenance of physical activity have been unsuccessful [20]. Research examining prevalence of successful physical activity maintenance currently
remains limited, warranting a better understanding of the associated influences.
One theoretical framework to conceptualize individuals’ motivation to sustain regular
physical activity is the self-determination theory (SDT). The SDT organizes the internal
and external factors that motivate sustainment of health behavior [7]. External or extrinsic factors motivate individuals’ physical activity in efforts
to achieve external and/or tangible rewards [3]. Internal or intrinsic factors are fueled by internal rewards (e. g., interest)
[6], and these internally derived motivations are expected to support long-term habitual
behaviors [7]. Deci and Ryan [7] theorized three basic psychological needs required to achieve optimal intrinsic
motivation, including competence, relatedness, and autonomy. Competence is an individual’s
belief they can successfully perform regular physical activity. Relatedness is one’s
sense of community when physically active and the relationships built from participation.
Autonomy is the individual’s perceived control over decisions relating to their physical
activity. With these needs met through a supportive environment, the individual has
the potential to develop autonomous and intrinsic motivation [22].
One multidimensional construct of the SDT is physical activity motives, which includes
motivations that manifest from both internal and external factors [23]. The first, interest, is an intrinsic motive relative to one’s perceptions of interest
and fun relative to physical activity. Competence is another intrinsic motive driven
by one’s motivation to improve and master new physical activity skills and challenges.
The first of three extrinsic motives is appearance, which motivates physical activity
behaviors via one’s desire to develop, improve, and/or maintain their physical physique,
muscle definition, and/or body weight. Fitness is the extrinsic motivation to be healthy,
strong, and energetic. Finally, social motives are the extrinsic drive to socially
interact and meet new people during physical activities. According to the SDT, individuals
emphasizing intrinsic motives are expected to maintain healthy behaviors more successfully
than those primarily motivated by extrinsic motives [8]. However, evidence regarding how SDT motives strengthen physical activity maintenance
remains limited and somewhat ambiguous.
Previous physical activity studies reported that successful physical activity maintenance
was supported by both intrinsic and extrinsic motives. Research found the most influential
intrinsic motives reported for physical activity maintenance were interest, competence,
and fitness [1]
[4]. Furthermore, Aaltonen and colleagues [1] reported a decrease in extrinsic motives as participants progressed toward successful
physical activity maintenance. Alternatively, similar research found certain extrinsic
motives were positively associated with physical activity maintenance [1]
[4]. Overall, sustained regular physical activity seems to involve multiple motives
that are both intrinsic and extrinsic. Research outcomes from this study provide further
understanding of adults’ motives in relation to their physical activity maintenance.
The current study objective was to compare motives between adults who have successfully
maintained regular physical activity and those who have not. Based on change from
participants’ previous and current physical activity, four samples were categorized
and compared: maintainers, improvers, decliners, and sedentary. Adults who reported
sustained regular physical activity for at least 10 years (maintainers) were expected
to have higher psychological needs (autonomy, relatedness, and competence) and intrinsic
motives (interest and competence) than participants who remained sedentary and those
who reported decreased physical activity (decliners). Similarly, participants who
increased their physical activity (improvers) were expected to be similar to maintainers,
reporting higher psychological needs and intrinsic motives than decliners and sedentary
adults. Differences between sedentary adults and decliners were not expected. Certain
extrinsic motives were also expected to be higher among maintainers and improvers,
specifically fitness and social motives. All other comparisons were deemed exploratory.
Materials & Methods
The current study used a retrospective study design. A one-time assessment was completed
by participants. Participants reported their current motives and physical activity
levels, and then retrospectively reported their same activity at least 10 years prior.
Participants were recruited through a partnership with a university alumni organization
in Kansas, USA. Participants were invited to complete the online survey via invitational
emails. Approval from the University Institutional Review and participant consent
preceded all study procedures. The current study meets the ethical standards of the
journal [14].
Data were reported over 4 months in the year 2015. The participant response rate was
20%. Only data collected from adults between the ages of 28 and 45 years were included
in the final analysis. Participants were asked to self-report basic demographic information
(age, gender, ethnicity, and education), their height, weight, and perceived health
status. Researchers used height and weight to calculate each participant’s body mass
index (BMI) using the following equation: [weight (lb)/height (in) 2] x 703. The recommended BMI range for adults is between 19.0 and 24.9 [5]. Current participants with a BMI≤25.0 were classified as normal, and participants
with a BMI>25.0 were considered overweight/obese. Perceived health was measured with
one question on a 5-point Likert scale, asking participants to report their general
health status as poor (1), fair (2), good (3), very good (4), or excellent (5). Physical
activity and SDT construct measurements are described in the following sections.
Physical activity measure
The Godin-Shephard Leisure–Time Physical Activity Questionnaire (GSLTPAQ) was used
to assess participants’ moderate and strenuous physical activity, which has been previously
validated for adults [10]
[2]. Participants reported the number of previous and current moderate and strenuous
physical activity bouts of 15 min they performed in a typical week. Previous physical
activity was described with the following statement: “Please recall your usual exercise behavior during college, prior to graduation (when
you were approximately 17 to 23 years old).” Similar to their previous physical activity, participants reported their current
weekly bouts of moderate and strenuous physical activity.
Self-reported moderate and strenuous physical activity was ultimately calculated into
metabolic equivalents (MET). MET represents physical activity in multiples of resting
oxygen consumption. Weekly frequency of moderate and strenuous physical activity bouts
were weighted using the following calculations: weekly moderate physical activity
bouts x 5, and weekly strenuous physical activity bouts x 9. The MET level required
to receive the health benefits of regular physical activity is≥24 MET per week [10]
[2]. Based on the change in participants’ self-reported previous and current physical
activity MET, each participant was categorized into one of the following four sample
subgroups: maintainers, or those who met previous and current physical activity recommendations; decliners, or those who met recommendations previously but did not maintain regular activity;
sedentary, or participants who never reported meeting physical activity recommendations; and
improvers who reported not meeting recommendations previously but currently meeting recommendations.
SDT construct measures
The Basic Psychological Needs Scale (BPNS) [8]
[9] was used to assess participants’ three basic psychological needs (autonomy, competence,
and relatedness). Participants responded to 21 statements using a 7-point Likert scale
ranging from 1 (“not at all true for me”) to 7 (“very true for me”). This scale has
been deemed reliable and valid [26]. Participants’ current physical activity motives were assessed using the revision
of the Motives for Physical Activity Measure (MPAM-R), including competence, interest,
appearance, fitness, and social motives [23]. Participants responded to 30 items on a 7-point Likert scale (1, not at all true
for me, to 7, very true for me).
Statistical analyses
Data were analyzed using SPSS Statistics version 23 (IBM, Chicago, IL). Only participants
with complete data were included in analyses. Descriptive statistics characterized
the sample and described participants’ current and previous physical activity MET,
BPNS, and physical activity motives. Analysis of variance (ANOVA) F-test was used
to test for differences in participants’ previous and current physical activity MET,
BPNS, and physical activity motives between maintenance subgroup classification (maintainers,
improvers, sedentary, and decliners). Any significant difference found between the
four subgroups groups (P<0.05) was then examined using Tukey’s HSD post hoc analyses with Bonferroni correction
(0.05/n=4, significance level=P<0.01). Unadjusted and adjusted means were calculated to account for subgroup differences
based on the following covariates: gender (male, female) and perceived health (5-point
Likert scale). Violations of normality and homogeneity of variance were not detected
(P<0.05).
Results
Participants (N=721) were mostly female (63%) and Caucasian (90%) with a mean age
of 33.37 (standard deviation (SD)=3.87). Participants’ average perceived health was
3.83 (SD=0.77), the mean BMI was 26.00 (SD=5.31), and 49% of participants were overweight
or obese. Participants were highly educated, including 37% college graduates and 58%
with a Master’s degree or higher. Differences in previous and current physical activity
MET between subgroups based on gender, ethnicity, education level, and BMI are shown
in [Table 1]. There were no significant differences in previous physical activity MET between
subgroups. Males reported significantly higher current physical activity MET than
females (mean difference (MD)=10.84, P<0.001). Overweight/obese participants reported significantly lower physical activity
MET than those with a normal BMI (MD=5.06, P<0.05).
Table 1 Adjusted mean differences in metabolic equivalent of physical activity between demographic
subgroups.
|
Sample Characteristics
|
n (%)
|
Previous PA MET
|
Current PA MET
|
|
Mean (SEM)
|
Mean (SEM)
|
|
Total Sample
|
721 (100%)
|
61.81 (1.72)
|
46.48 (1.26)
|
|
Gender
|
|
Male
|
268 (37%)
|
65.54 (2.85)
|
53.30 (2.12)
|
|
Female
|
453 (63%)
|
59.61 (2.15)
|
42.46 (1.53)
|
|
Ethnicity
|
|
White
|
650 (90%)
|
60.71 (1.82)
|
46.37 (1.34)
|
|
Mixed or Other
|
71 (10%)
|
71.93 (5.03)
|
47.54 (3.53)
|
|
Education
|
|
HS Diploma or AS
|
27 (4%)
|
57.67 (10.05)
|
38.78 (6.78)
|
|
College Graduate
|
276 (38%)
|
63.06 (2.91)
|
45.51 (2.04)
|
|
Advanced Degree
|
418 (58%)
|
61.26 (2.17)
|
46.49 (1.26)
|
|
BMI
|
|
†Normal
|
365 (51%)
|
62.90 (2.44)
|
48.99 (1.90)
|
|
§Overweight/Obese
|
356 (49%)
|
60.70 (2.42)
|
43.93 (1.64)
|
Abbreviations: PA=physical activity; MET=metabolic equivalent of moderate and strenuous
weekly physical activity; SEM=standard error of the mean; HS=high school; AS=some
college or Associates Degree; BMI=body mass index; Statistically significant differences
are bolded (P<0.05);
±
Covariates included age (continuous) and perceived health (5-point Likert scale,
1=poor, 2=fair, 3=good, 4=very good, 5=excellent); † Normal weight=body mass index≤25.00; §Overweight/Obese=body mass index>25.00.
Comparison of physical activity and self-determination constructs between maintenance
subgroups are provided in [Table 2]. Significant covariates were gender (P<0.001) and perceived health (P<0.001). As seen, previous and current physical activity MET were different between
maintenance groups (P<0.05). Pairwise comparisons were used to reveal specific subgroup differences (P<0.01). Maintainers reported higher previous weekly strenuous and moderate physical
activity MET compared to improvers (MD=64.46, P<0.001) and sedentary participants (MD=66.65, P<0.001). Participants who improved their physical activity reported lower previous
MET than those who declined (MD=53.81, P<0.001). Finally, decliners reported more previous physical activity MET than sedentary
participants (MD=56.01, P<0.001). Specific to participants’ current weekly strenuous and moderate physical
activity, maintainers reported higher MET than improvers (MD=11.96, P<0.001), decliners (MD=49.13, P<0.001), and the sedentary (MD=50.49, P<0.001). And, improvers reported higher current physical activity MET than decliners
(MD=37.17, P<0.001) and sedentary participants (MD=39.53, P<0.001).
Table 2 Unadjusted and adjusted mean differences in physical activity, basic psychological
needs, and physical activity motives by maintenance subgroups.
|
Maintainers (n=461)
|
Improvers (n=79)
|
Decliners (n=125)
|
Sedentary (n=56)
|
|
Unadjusted
|
Adjusted†
|
Unadjusted
|
Adjusted†
|
Unadjusted
|
Adjusted†
|
Unadjusted
|
Adjusted†
|
|
Mean (SEM)
|
Mean (SEM)
|
Mean (SEM)
|
Mean (SEM)
|
Mean (SEM)
|
Mean (SEM)
|
Mean (SEM)
|
Mean (SEM)
|
|
Physical Activity
|
|
Previous Weekly MET
|
75.90 (2.10) a
|
75.90 (2.10)
a
|
11.44 (0.81)
b
|
11.44 (0.81) b
|
65.26 (3.09)
ac
|
65.26 (3.09)
ac
|
9.25 (1.02)
b
|
9.25 (1.02)
b
|
|
Current Weekly MET
|
60.24 (1.45)
a
|
60.24 (1.45)
a
|
48.28 (2.76)
b
|
48.28 (2.76)
b
|
11.10 (0.62)
c
|
11.10 (0.62)
c
|
9.75 (1.06)
c
|
9.75 (1.06)
c
|
|
Basic Psychological Needs§
|
|
Competence
|
5.58 (0.04)
|
5.58 (0.04)
|
5.58 (0.10)
|
5.58 (0.10)
|
5.43 (0.08)
|
5.43 (0.08)
|
5.27 (0.14)
|
5.27 (0.14)
|
|
Autonomy
|
5.66 (0.04)
|
5.66 (0.04)
|
5.69 (0.10)
|
5.69 (0.10)
|
5.50 (0.08)
|
5.50 (0.08)
|
5.65 (0.12)
|
5.65 (0.12)
|
|
Relatedness
|
5.74 (0.04)
|
5.74 (0.04)
|
5.75 (0.11)
|
5.75 (0.11)
|
5.50 (0.07)
|
5.50 (0.07)
|
5.73 (0.11)
|
5.73 (0.11)
|
|
Intrinsic Physical Activity Motives§
|
|
Interest
|
5.08 (0.06)
a
|
5.08 (0.06)
a
|
4.59 (0.18)
ab
|
4.59 (0.18)
ab
|
4.08 (0.12)
b
|
4.08 (0.12)
b
|
3.74 (0.18)
c
|
3.74 (0.18)
c
|
|
Competence
|
4.77 (0.06)
a
|
4.77 (0.06)
a
|
4.20 (0.17)
b
|
4.20 (0.17)
b
|
3.48 (0.12)
c
|
3.48 (0.12)
c
|
3.19 (0.18)
c
|
3.19 (0.18)
c
|
|
Appearance
|
5.13 (0.06)
|
5.13 (0.06)
|
5.36 (0.13)
|
5.36 (0.13)
|
4.89 (0.13)
|
4.89 (0.13)
|
4.79 (0.20)
|
4.79 (0.20)
|
|
Extrinsic Physical Activity Motives§
|
|
Fitness
|
6.08 (0.04)
a
|
6.08 (0.04)
a
|
5.91 (0.11) ac
|
5.91 (0.11)
ac
|
5.40 (0.11)
b
|
5.40 (0.11)
b
|
5.52 (0.15)
c
|
5.52 (0.15)
c
|
|
Social
|
2.95 (0.06)
a
|
2.95 (0.06)
a
|
2.64 (0.15)
ab
|
2.64 (0.15)
ab
|
2.50 (0.11)
b
|
2.50 (0.11)
b
|
2.26 (0.16)
b
|
2.26 (0.16)
b
|
Abbreviations: SEM=standard error of the mean; MET=metabolic equivalent of moderate
and strenuous weekly physical activity; Statistically significant differences are
bolded (P<0.05); †Covariates included in the adjusted models were gender (1=male,
2=female) and perceived health (5-point Likert scale, 1=poor, 2=fair, 3=good, 4=very
good, 5=excellent);
§
7-point Likert scale (1=not at all true, to 7=very true); Any row mean within either
the unadjusted or adjusted column not followed by the same letter is significantly
different (P<0.001).
Also shown in [Table 2] are differences in physical activity motives between maintenance subgroups. Except
for the appearance motive, subgroup differences were detected for all other motives
(P<0.05). Significant covariates were gender (P<0.05) and perceived health (P<0.001). Pairwise comparisons revealed maintainers reported higher physical activity
interest than decliners (MD=1.00, P<0.001) and sedentary participants (MD=1.33, P<0.001). Improvers were higher in interest compared to sedentary participants (MD=0.84,
P<0.001). Maintainers were higher in competence compared to improvers (MD=0.57, P<0.001), decliners (MD=1.28, P<0.001), and the sedentary (MD=1.58, P<0.001), and improvers were higher in competence compared to decliners (MD=0.72, P<0.001) and sedentary participants (MD=1.01, P<0.001). Similar differences were seen in extrinsic motives between maintenance subgroups.
Maintainers reported higher fitness motive compared to decliners (MD=0.67, P<0.001) and sedentary participants (MD=0.55, P<0.001), improvers were higher in fitness than decliners (MD=0.51, P<0.001). Last, social motive was higher among maintainers compared to decliners (MD=0.45,
P<0.001) and sedentary participants (MD=0.69, P<0.001).
Discussion
The current study examined the relationship between certain SDT constructs and physical
activity maintenance. This study supports previous research and theory, and also adds
to the current literature. To begin, males in the current study reported more current
physical activity than their female counterparts, which mimics outcomes of previous
research [12]. Also as expected, participants with a BMI ≤25.00 reported more current physical
activity than those who were overweight or obese [13]. Secondly, differences in previous and current physical activity followed expected
patterns and confirmed the categorization of maintenance groups.
Relative to SDT constructs, no differences emerged in basic psychological needs between
the maintenance groups. According to theory, this indicates all current participants
had the psychological framework to facilitate successful maintenance of regular physical
activity [8]. Maintainers in the current study reported higher intrinsic motives (competence
and interest) and extrinsic motives (fitness and social) than decliners and sedentary
participants. This is comparable to evidence indicating that both aid in successful
physical activity maintenance [1]
[4]
[5]. Interestingly, no differences were seen between maintainers and improvers, which
may indicate similar motives when both adopting and maintaining physical activity.
Specific differences in physical activity motives between maintenance groups are comparable
to previous research. Physical activity maintainers were higher in competence than
all other maintenance subgroups, confirming similar research [1]
[4]. Moreover, competence was the only motive maintainers reported higher than improvers;
hence, competence-related motives for physical activity may bridge a gap between adoption
and successful maintenance. Compared to decliners and sedentary participants, maintainers
reported higher interest for physical activity; and improvers reported more interest
than the sedentary. These results are comparable to previous studies, reporting highest
interest among adults who had successfully maintained physical activity [4]
[5].
Extrinsic motives were also highest among maintainers and improvers. Both maintainers
and improvers reported a higher fitness motive than decliners, which supports evidence
that health and fitness motives predict intrinsic motivation [15]. Maintainers also reported the highest social motive, confirming evidence that social
motivations can facilitate intrinsic regulation of physical activity [16]. Interestingly, no differences in appearance motive were seen between maintenance
groups. Buckworth and others [4] found appearance motives were highest among college students maintaining physical
activity. Conversely, appearance was not a salient motivation for physical activity
maintenance among a sample of young to middle-aged adults [17]. Also, appearance motives were inversely related to female adults’ physical activity
[24]. Further research is needed to clarify certain developmental shifts that impact
one’s motives to sustain regular physical activity.
Conclusions
Given the numerous health benefits of regular physical activity, a better understanding
of the mechanisms driving long-term, habitual physical activity behavior is required.
Based on current outcomes, both intrinsic and extrinsic motivations impact habitual
physical activity. Additional research is needed to clarify possible age/developmental
differences to determine the motives most conducive to physical activity maintenance
across the lifespan. One study limitation was the homogeneous study sample. Most participants
were Caucasian and highly educated, which may not generalize to the greater population.
Also, the current study used a retrospective study design that is open to memory bias.
Prospective, longitudinal examinations are warranted.
