- Journal List
- HHS Author Manuscripts
- PMC8170853
As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsem*nt of, or agreement with, the contents by NLM or the National Institutes of Health.
Learn more: PMC Disclaimer | PMC Copyright Notice
Prev Med. Author manuscript; available in PMC 2021 Jun 2.
Published in final edited form as:
Prev Med. 2018 Oct; 115: 8–11.
Published online 2018 Aug 3. doi:10.1016/j.ypmed.2018.08.002
PMCID: PMC8170853
NIHMSID: NIHMS1705704
PMID: 30081132
M Robertson,1,2 Y Liao,1 J Song,1 EJ Lyons,3 and K Basen-Engquist1
Author information Copyright and License information PMC Disclaimer
The publisher's final edited version of this article is available at Prev Med
Abstract
The aim of this study was to investigate associations between types of motivation for physical activity and self-reported weekly aerobic moderate-to-vigorous physical activity (MVPA) in the 2012 and 2014 waves of the nationally representative Health Information National Trends Survey 4 (n=7,307). We further explored differential associations between MVPA and types of motivation for physical activity by cancer survivor status.
We found that those who were more motivated by “getting enjoyment from exercise” reported 26.4% more MVPA (+49.8 min/week) than those who were less motivated by this factor, adjusting for covariates (p=.025). Conversely, those who were more motivated by “concern over the way you look” reported 22.1% less MVPA (−55.5 min/week) than those who were less motivated by this factor, adjusting for covariates (p=.002). We found no evidence for a relationship between motivation from either “pressure from others” or “feeling guilty when you skip exercising” and MVPA. We identified a significant interaction for “feeling guilty when you skip exercising” and cancer survivor status, adjusting for covariates (p=.034). Cancer survivors who reported being more motivated by “feeling guilty when you skip exercising” reported 36.2% less MVPA (−71.75 min/week) than those who were less motivated by this factor; there was no statistically reliable difference in those without a history of cancer.
Findings are concordant with previous literature highlighting the primacy of enjoyment for physical activity adherence. There is a need for further inquiry into guilt-related motivation for physical activity among cancer survivors, as it may have a unique, negative impact in this population.
Keywords: Physical Activity, Cancer Survivors, Motivation, Guilt
Introduction
Physical activity has been linked to primary prevention benefits for numerous types of cancer and can confer a host of tertiary prevention benefits for cancer survivors.1 Despite the health benefits physical activity levels both in the general population2 and among cancer survivors3 are sub-optimal4, 5. A growing literature, arising from Self-Determination Theory (SDT),6 has begun to appreciate the influence of quality of motivation in shaping physical activity patterns. At the most basic level, SDT classifies motivations as either intrinsic (i.e., behavior is inherently enjoyable) or extrinsic (i.e., behavior is a means to an end). Extrinsic motivations are further divided into discrete categories along an underlying continuum reflecting the degree to which the source of the motivation is external (e.g., rewards/punishment) vs. internal (e.g., valuing its outcomes). Compelling evidence supports the utility of understanding one’s motivation for increasing physical activity,7 but cancer survivors face challenges to physical activity that may influence the dynamic between motivation and physical activity.8, 9 There is a need to evaluate whether different types of motivation influence physical activity patterns in cancer survivors distinctly.
The aim of the present study was to investigate how various types of motivation for physical activity are associated with physical activity levels in nationally representative data, and to test whether these relationships are moderated by cancer survivor status. An exploratory aim was to similarly investigate effect modification by three other prevalent chronic conditions: arthritis, diabetes, and hypertension. We hypothesized that intrinsic motivation would be positively associated with physical activity levels in the general U.S. population, and that this relationship would be especially pronounced in cancer survivors.
Methods
Study design and population
The Health Information National Trends Survey (HINTS) 4 was a series of nationally representative cross-sectional surveys administered by the National Cancer Institute (NCI) to provide insight on cancer risk-related variables in non-institutionalized U.S. residents.10, 11 Taking into account HINTS 4’s complex, weighted survey design, we combined: HINTS 4 Cycle 2 (conducted October 2012–January 2013; response rate 40.0%; sample size 3,630), and HINTS 4 Cycle 4 (conducted August 2014–November 2014; response rate 34.4%; sample size 3,677).
Measures
Physical activity
We obtained participants’ self-reported typical weekly minutes of aerobic moderate-to-vigorous physical activity (MVPA) by multiplying (1) weekly frequency of MVPA with (2) typical time spent in MVPA. The first item asked, “In a typical week, how many days do you do any physical activity or exercise of at least moderate intensity, such as brisk walking, bicycling at a regular pace, and swimming at a regular pace?” The second asked, “On the days that you do any physical activity or exercise of at least moderate intensity, how long do you typically do these activities?” The latter was free response; to handle illogical responses on this item (e.g., >24 hours of MVPA daily) we truncated the hours of MVPA per day at 16 hours, as is recommended for a similar measure.12 We modeled MVPA as count data, using standard cut points to interpret outcomes.
Motivations for physical activity
Four types of motivation for physical activity were captured by four Likert-type items. These items were introduced with the text, “People start or continue exercising regularly for lots of reasons. How much do each of the following reflect why you would start or continue exercising regularly?” The four motivation types were “pressure from others” (which may be understood in SDT terms to reflect an extrinsic motivation from an external source), “concern over the way you look” (extrinsic motivation, somewhat external source), “feeling guilty when you skip exercising” (extrinsic motivation, somewhat external source), and “getting enjoyment from exercise” (intrinsic motivation).6 Respondents indicated whether each motivation type influenced them: “Not at all”, “A little”, “Some”, and “A lot”. Since these response choices represent ordinal level data (i.e., not equal spacing between responses), and for clarity of interpretation, we dichotomized responses into less motivation (“Not at all” or “A little”) and more motivation (“Some” or “A lot”). To investigate how this decision may have influenced the results, we repeated all analyses using orthogonal polynomials to model the ordinal-level predictor variables (thus modeling answer choices as if there was equal spacing between responses). These results evidenced concordant trends, and are not presented here.
Non-communicable disease status
Cancer survivor status was measured as a dichotomous variable that asked, “Have you ever been diagnosed as having cancer?” Similar items were used to ascertain participant’s “diabetes or high blood sugar”, “high blood pressure or hypertension”, and “arthritis or rheumatism” statuses.
Statistical methods
We combined the two HINTS waves, arraying survey and replicate weights as recommended by NCI.13 Only income range had greater than 10% missing data. We replaced this variable with a singly imputed variable provided by NCI and used casewise deletion to handle other missing data. We conducted Wald’s design-based chi-square tests of independence to test for differences by cancer survivor status, and quasi-Poisson regression models to investigate our research questions. This approach allowed us to model our outcome variable (MVPA minutes) as count data in a way that is robust to model misspecification and overdispersion. This procedure is comparable to negative binomial, but can more readily handle the HINTS 4’s complex, weighted survey design.14 We investigated the associations between MVPA and the four types of motivation for physical activity, adjusting for potentially confounding variables. We further evaluated each model by adding an interaction term for the motivator and cancer survivor status. Finally, using the whole sample, we similarly explored effect modification in separate models by arthritis/rheumatism, diabetes/high blood sugar, and hypertension/high blood pressure statuses. All models adjusted for self-reported age, gender, race (white, non-white), ethnicity (Hispanic, non-Hispanic), combined annual income (<$50,000, >=$50,000), health status (poor, fair, good, very good, excellent), and HINTS survey wave. Additionally, we adjusted for each of the other prevalent non-communicable diseases of interest. We conducted all analyses in R version 3.3.2 using the survey, psych, gmodels, and effects packages.15
Results
Participant characteristics
Wald design-based chi-square tests of independence indicated that cancer survivors were significantly older and less physically active than those without a history of cancer (Table 1). A higher proportion of cancer survivors were married, non-Hispanic white, insured, in poorer health, and retired. Health status, arthritis/rheumatism status, diabetes/high blood sugar status, hypertension/high blood pressure status were only weakly correlated (range: −.29 to .34). We observed no statistically significant differences for education level or body mass index. Weighted percentages indicated that 16.8%, 59.8%, 42.2%, and 59.0% cited more motivation to start or continue exercising regularly due to “pressure from others”, “concern over the way you look”, “feeling guilty when you skip exercising”, and “getting enjoyment from exercise”, respectively. There were no statistically significant differences in these proportions by cancer survivor status.
Table 1.
Survey weighted sociodemographic characteristics by cancer survivor status
| Cancer survivors | No history of cancer | P-valuea | ||
|---|---|---|---|---|
| (n=1,006) | (n=6,241) | |||
| Characteristic | Category | % | % | |
| Age Category (years) | <0.001 | |||
| 18–34 | 4.78 | 33.00 | ||
| 35–49 | 13.45 | 27.77 | ||
| 50–64 | 33.40 | 24.69 | ||
| 65–74 | 23.43 | 8.23 | ||
| 75+ | 24.94 | 6.31 | ||
| BMI Category | .630 | |||
| Underweight | 1.56 | 2.12 | ||
| Normal | 31.01 | 32.56 | ||
| Overweight | 34.74 | 34.46 | ||
| Obese | 32.69 | 30.86 | ||
| Education Level | 0.139 | |||
| < High School | 15.91 | 12.18 | ||
| HS Graduate | 17.49 | 19.45 | ||
| Some College | 35.81 | 33.64 | ||
| Bachelor’s (+) | 30.79 | 34.72 | ||
| Employment | <0.001 | |||
| Employed | 34.63 | 59.82 | ||
| Unemployed | 4.23 | 7.39 | ||
| Homemaker | 7.95 | 6.15 | ||
| Student | 0.71 | 6.95 | ||
| Retired | 44.77 | 13.98 | ||
| Disabled | 7.46 | 5.29 | ||
| Other | 0.26 | 0.42 | ||
| Exercise Categoryb | <0.001 | |||
| Inactive (0c) | 33.47 | 25.09 | ||
| Insufficiently Active (>0 & <150c) | 29.80 | 28.92 | ||
| Active (>=150c) | 36.74 | 45.98 | ||
| Gender | 0.002 | |||
| Female | 58.65 | 50.85 | ||
| Male | 41.35 | 49.15 | ||
| Health Insurance | <0.001 | |||
| No | 5.97 | 16.07 | ||
| Yes | 94.03 | 83.93 | ||
| Health Status | <0.001 | |||
| Excellent | 6.25 | 12.77 | ||
| Very Good | 32.70 | 36.46 | ||
| Good | 39.32 | 36.72 | ||
| Fair | 16.62 | 11.70 | ||
| Poor | 5.12 | 2.35 | ||
| Income | <0.001 | |||
| $0–$9,999 | 5.17 | 9.90 | ||
| $10,000–$14,999 | 5.59 | 5.81 | ||
| $15,000–$19,999 | 7.77 | 5.25 | ||
| $20,000–$34,999 | 15.12 | 13.96 | ||
| $35,000–$49,999 | 14.64 | 15.00 | ||
| $50,000–$74,999 | 16.62 | 16.73 | ||
| $75,000–$99,999 | 14.73 | 13.13 | ||
| $100,000–$199,999 | 15.20 | 15.49 | ||
| $200,000+ | 5.18 | 4.72 | ||
| Marital Status | <0.001 | |||
| Married or living with partner | 66.51 | 55.98 | ||
| Not married or living with partner | 33.50 | 44.03 | ||
| Race/Ethnicity | <0.001 | |||
| Hispanic | 10.07 | 15.52 | ||
| Asian | 1.40 | 5.03 | ||
| Black/AA | 6.38 | 11.47 | ||
| White | 80.40 | 65.67 | ||
| Other | 1.75 | 2.31 | ||
| Diagnosed with: | ||||
| Arthritis or rheumatism | 41.54 | 20.28 | <0.001 | |
| Diabetes or high blood sugar | 25.06 | 13.09 | <0.001 | |
| High blood pressure or hypertension | 55.42 | 32.33 | <0.001 | |
| Time Since Cancer Diagnosis | ||||
| Less than 1 year | 12.20 | n/a | n/a | |
| 2–5 years | 24.19 | n/a | n/a | |
| 6–10 years | 19.22 | n/a | n/a | |
| 11+ years | 44.38 | n/a | n/a |
Open in a separate window
aWald design-based chi-square tests were performed to assess independence between sample characteristic and cancer survivorship status. Cancer survivor status information was missing for 60 individuals
bCut points derived from national recommendations4, 5
cReported minutes of moderate-to-vigorous exercise in a typical week
Motivations for physical activity and physical activity levels
Generalized linear regression models indicated that those who were more motivated by “getting enjoyment from exercise” reported 26.4% more MVPA (+49.8 min/week; n = 5555, B = .23, t(78) = 2.28, p = .025) than those who were less motivated by this factor, adjusting for potentially confounding variables. Conversely, those who were more motivated by “concern over the way you look” reported 22.1% less MVPA (−55.5 min/week, n = 5525, B = −.25, t(78) = −3.22, p = .002) than those who were less motivated by this factor, adjusting for potentially confounding variables.
Effect modification by cancer survivor status
Results differed for cancer survivors for one type of motivation: the relationship between “feeling guilty when you skip exercising” and MVPA (Figure 1; n = 5468, B = −.38, t(76) = −2.16, p = .034). Cancer survivors who reported being more motivated by “feeling guilty when you skip exercising” reported 36.2% less MVPA (−71.75 min/week) than those who were less motivated by this factor. There was not a statistically reliable difference in those with no history of cancer. Analyses did not produce evidence for interactions by arthritis/rheumatism, diabetes/high blood sugar, or hypertension/high blood pressure statuses.
Figure 1.
Model-predicted values of weekly moderate-to-vigorous physical activity (MVPA) by type of motivation and cancer survivor statusa
aAdjusted for age, gender, race/ethnicity, income, health status, arthritis/rheumatism status, diabetes/high blood sugar status, hypertension/high blood pressure status, and survey wave
*significant interaction between motivation, cancer survivor status, and physical activity (p=.034)
Discussion
Our finding that those who attributed more physical activity-related motivation to “getting enjoyment from exercise” tended to be more active than those who indicated less such motivation is concordant with SDT and our hypothesis. It is somewhat surprising, however, that the positive endorsem*nt of any particular type of motivation would be associated with less physical activity; we found this to be the case regarding “concern over the way you look”. SDT literature may provide some insight. It theorizes that extrinsic motivations from external sources may produce only short-lived behavior change efforts, and in some cases may even undermine behavior change efforts.6 From a public health standpoint, it is alarming that in this nationally representative sample the most frequently endorsed motivation for exercise was “concern over the way you look” (59.8%), yet this was associated with markedly lower physical activity levels. These data may be used to encourage individuals to re-assess their approach to behavior change in favor of developing more intrinsic motivation for long term success. Findings also have implications for communication campaigns aiming to increase physical activity, as they corroborate other research indicating that messaging should avoid weight- or appearance-related motivational content.16
We found evidence of an interaction on the relationship between “feeling guilty when you skip exercising” and MVPA by cancer survivor status (Figure 1). A study conducted by Castonguay and colleagues investigated the relationship between body-related guilt (which included aspects of diet and weight), SDT motivations, and physical activity in breast cancer survivors.17 The authors found body-related guilt to be positively associated with MVPA. This study investigated constructs that are distinct from those assessed in the current study. As such, our results are not necessarily discordant, but this apparent discrepancy does seem to point to a more complex relationship for which more research is needed. It is also important to note that the dynamic between guilt and physical activity may differ in individuals with different types of cancer. Our study results may reflect dissonance that cancer survivors experience as a result of initiating health-protective lifestyle changes in the time following diagnosis,18 but subsequently experiencing pronounced challenges to sustained behavior change. Additionally, cancer survivors are more likely to be retired than those without a history of cancer, and this difference may also have bearing on this dynamic.
A limitation of the current study is the cross-sectional survey design, which precludes establishing temporal relationships between variables. Accordingly, our results may simply reflect different stages of habit formation. The Maintain IT model posits that the cognitive processes involved in behavioral initiation and maintenance are distinct. Consistent with this model, it is quite possible that extrinsic motivators do have an important role to play in the more effortful processes involved in physical activity initiation, while enjoyment is associated with less the less effortful, automatic cognitive processes that underlie successful physical activity maintenance. Future studies should investigate this with superior study designs.19 Another limitation is the lack of formally validated measures, which may leave this study vulnerable to measurement error. Objective measurement of physical activity and the Behavioural Regulation in Exercise Questionnaire20 to assess exercise motivation are recommended for future research. We were also unable to investigate strength training, as it was not included in both HINTS waves. Strength training should be investigated in future studies, as motivation for it may differ from that of aerobic activity. Despite these limitations, these nationally representative data are useful for surveillance purposes and for generating hypotheses for future research.
This study corroborates previous literature by highlighting the primacy of enjoyment for physical activity adherence. Physical activity promotion efforts, both in the general population and in its subset of cancer survivors, should aim to develop high quality motivation. Public health officials should also be aware of the inverse relationship between appearance-based physical activity motivation and actual physical activity levels. Finally, this study provides preliminary evidence suggesting that physical activity-specific motivation attributed to guilt may operate differently in cancer survivors than it does in those without a history of cancer. Public health efforts should be sensitive to this, and future studies are needed to elucidate the nature of this phenomenon.
Funding Source
Center for Energy Balance in Cancer Prevention and Survivorship, Duncan Family Institute; the University of Texas MD Anderson Cancer Center Grant, P30CA016672. The Janice Davis Gordon Memorial Postdoctoral Fellowship in Colorectal Cancer Prevention.
Footnotes
Conflicts of Interest
None
REFERENCES
1. Lee IM. Physical activity and cancer prevention--data from epidemiologic studies. Med Sci Sports Exerc. 2003;35(11):1823–1827. doi: 10.1249/01.MSS.0000093620.27893.23 [doi]. [PubMed] [CrossRef] [Google Scholar]
2. Troiano RP, Berrigan D, Dodd KW, Mâsse LC, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc. 2008; 2007;40(1):181. doi: 10.1249/mss.0b013e31815a51b3. [PubMed] [CrossRef] [Google Scholar]
3. Ottenbacher A, Yu M, Moser RP, Phillips SM, Alfano C, Perna FM. Population estimates of meeting strength training and aerobic guidelines, by gender and cancer survivorship status: Findings from the Health Information National Trends Survey (HINTS). Journal of Physical Activity and Health. 2015;12(5):675–679. [PubMed] [Google Scholar]
4. Physical Activity Guidelines Advisory Committee. Physical activity guidelines for Americans. Washington, DC: US Department of Health and Human Services. 2008:15–34. [Google Scholar]
5. Schmitz KH, Courneya KS, Matthews C, et al. American College of Sports Medicine roundtable on exercise guidelines for cancer survivors. Med Sci Sports Exerc. 2010;42(7):1409–1426. doi: 10.1249/MSS.0b013e3181e0c112. [PubMed] [CrossRef] [Google Scholar]
6. Ryan RM, Deci EL. Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. Am Psychol. 2000;55(1):68. [PubMed] [Google Scholar]
7. Teixeira PJ, Carraça EV, Markland D, Silva MN, Ryan RM. Exercise, physical activity, and self-determination theory: a systematic review. International Journal of Behavioral Nutrition and Physical Activity. 2012;9(1):1. [PMC free article] [PubMed] [Google Scholar]
8. Pinto BM, Ciccolo JT. Physical activity motivation and cancer survivorship. Recent results in cancer research.Fortschritte der Krebsforschung.Progrès dans les recherches sur le cancer2011;186:367. [PubMed] [Google Scholar]
9. Blaney J, Lowe-Strong A, Rankin-Watt J, Campbell A, Gracey J. Cancer survivors’ exercise barriers, facilitators and preferences in the context of fatigue, quality of life and physical activity participation: a questionnaire–survey. Psycho-Oncology. 2013;22(1):186–194. [PubMed] [Google Scholar]
10. U.S. Department of Health and Human Services, National Institutes of Health, National Cancer Institute (2012). Health Information National Trends Survey, Administration 4, Cycle 2, [Data set]. Retrieved from http://hints.cancer.gov.
11. U.S. Department of Health and Human Services, National Institutes of Health, National Cancer Institute (2014). Health Information National Trends Survey, Administration 4, Cycles 4, [Data set]. Retrieved from http://hints.cancer.gov.
12. IPAQ Research Committee. Guidelines for data processing and analysis of the International Physical Activity Questionnaire (IPAQ)–short and long forms. RetrievedSeptember. 2005;17:2008.
13. Moser RP, Naveed S, Cantor D, et al. Integrative analytic methods using population-level cross-sectional data. Hints.cancer2013. [Google Scholar]
14. Ver Hoef JM, Boveng PL. Quasi-Poisson vs. negative binomial regression: How should we model overdispersed count data?. Ecology. 2007;88(11):2766–2772. [PubMed] [Google Scholar]
15. R Core Team. R: A language and environment for statistical computing. 2017.
16. Homan KJ, Tylka TL. Appearance-based exercise motivation moderates the relationship between exercise frequency and positive body image. Body Image. 2014;11(2):101–108. [PubMed] [Google Scholar]
17. Castonguay AL, Wrosch C, Pila E, Sabiston CM. Body-related shame and guilt predict physical activity in breast cancer survivors over time. 2017;44(4):465–475. [PubMed] [Google Scholar]
18. Demark-Wahnefried W, Aziz NM, Rowland JH, Pinto BM. Riding the crest of the teachable moment: promoting long-term health after the diagnosis of cancer. J Clin Oncol. 2005;23(24):5814–5830. doi: JCO.2005.01.230[pii]. [PMC free article] [PubMed] [Google Scholar]
19. Caldwell AE, Masters KS, Peters JC, et al. Harnessing centred identity transformation to reduce executive function burden for maintenance of health behaviour change: the Maintain IT model. Health psychology review. 2018:1–23. [PMC free article] [PubMed] [Google Scholar]
20. Markland D, Tobin V. A modification to the behavioural regulation in exercise questionnaire to include an assessment of amotivation. Journal of Sport and Exercise Psychology. 2004;26(2):191–196. [Google Scholar]
