- Email: [email protected]
The association between discrepant weight perceptions and objectively measured physical activity
Preventive Medicine 87 (2016) 47–50
Contents lists available at ScienceDirect
Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed
Brief Original Report
The association between discrepant weight perceptions and objectively measured physical activity Lois Coleman, Paul D. Loprinzi ⁎ Department of Health, Exercise Science and Recreation Management, The University of Mississippi, USA
a r t i c l e
i n f o
Article history: Received 14 October 2015 Received in revised form 30 January 2016 Accepted 1 February 2016 Available online 11 February 2016 Keywords: Accelerometry Epidemiology Weight perception
a b s t r a c t Background. Some individuals perceive themselves as being normal weight, despite having an excess body fat percentage (e.g., underestimate weight). Conversely, other individuals perceive themselves as being overweight, despite having a normal body fat percentage (e.g., overestimate weight). When perceived and actual weight statuses are incongruent, individuals possess a discrepant weight perception. The association between discrepant weight perceptions and engagement in moderate-to-vigorous physical activity (MVPA) has not been thoroughly investigated, which was this study's purpose. Methods. For this cross-sectional study, data from the 2003–2006 National Health and Nutrition Examination Survey were utilized (N = 5462 adults). MVPA was assessed via accelerometry. Based on measured body mass index and whether participants considered themselves as overweight, underweight, or about the right weight, we classified individuals as accurate perception, overestimate weight (discrepant), or underestimate weight (discrepant). A negative binomial logistical regression was used to assess the association between discrepant weight perception (independent variable) and engagement in MVPA (outcome variable). Results. Females who said that they are normal weight, but were in fact overweight based on body mass index, engaged in 13% less MVPA (rate ratio = .87, 95% confidence interval: .769–.999, P = .05). Also, older adults (N 60 yrs) who said that they are normal weight, but were overweight based on body mass index, engaged in 23% less MVPA (rate ratio = .77, 95% confidence interval: .616–.965, P = .025). Conclusion. Discrepant weight perceptions were associated with less objectively measured MVPA. Interventions should take weight perceptions into consideration when designing and evaluating intervention impact. © 2016 Elsevier Inc. All rights reserved.
1. Introduction Weight perception has been defined as how an individual perceives his or her body weight, and is affected by their “ideal” and “normative” body images (Park, 2011). Some individuals perceive themselves as being normal weight, despite having an excess body fat percentage (e.g., underestimate weight). Conversely, other individuals perceive themselves as being overweight, despite having a normal body fat percentage (e.g., overestimate weight) (Duncan et al., 2011; Harring et al., 2010). Furthermore, males are more likely to underestimate their weight, in contrast to females who generally overestimate their weight (Kuchler and Variyam, 2003; Yaemsiri et al., 2011; Park, 2011). Discrepant weight perceptions may be problematic in terms of both health behaviors and resulting health outcomes (Loprinzi, 2015; Loprinzi et al., 2015). This is in alignment with various theoretical models, including the Health Belief Model, which posits that beliefs ⁎ Corresponding author at: Center for Health Behavior Research, Department of Health, Exercise Science and Recreation Management, The University of Mississippi, 229 Turner Center, University, MS 38677, USA. E-mail address: [email protected] (P.D. Loprinzi).
http://dx.doi.org/10.1016/j.ypmed.2016.02.004 0091-7435/© 2016 Elsevier Inc. All rights reserved.
and attitudes play an important role in behavioral initiation and maintenance (Rosenstock, 1974). Studies suggest that the discrepancy not only contributes to the current obesity epidemic (Park, 2011), but individuals who possess a discrepant weight perception are likely to engage in negative health behaviors including: substance use, alcohol abuse, disordered eating, and smoking (Antin and Paschall, 2011; Eichen et al., 2012). They are also at an increased risk of suffering from depression (Harring et al., 2010). This is concerning as these potentially negative consequences of discrepant weight perceptions may influence health-enhancing behaviors, such as physical activity. However, no recent studies to our knowledge, have examined the association between weight perception and engagement in accelerometerassessed moderate-to-vigorous physical activity (MVPA) using a nationally representative sample. While even light physical activity is considered beneficial (Loprinzi, 2016a, 2016b), exploring the link between weight perception and MVPA is important given that MVPA is associated with numerous positive health outcomes including: a reduced risk for certain cancers, type 2 diabetes, coronary heart disease, stroke, and metabolic syndrome; improved cognitive function, sleep quality, and bone density; and lower levels of depression (U.S. Department of Health and Human
48
L. Coleman, P.D. Loprinzi / Preventive Medicine 87 (2016) 47–50
Services, 2008). Also, national physical activity objectives for Healthy People 2020 include increasing the proportion of adults who engage in physical activity of at least moderate intensity; weight perceptions are a potential barrier to reaching those objectives (U.S. Department of Health and Human Services, 2016). Consequently, there is a need to investigate whether there is an association between weight perception and engagement in MVPA. Therefore, the primary purpose of this study was to examine the association between weight perception and engagement in MVPA using a nationally representative sample. A secondary objective was to determine if individuals' demographic profiles (race, gender, age, and socioeconomic status) moderate the potential association between weight perception and engagement in MVPA. This is plausible because previous research has demonstrated that weight perception and engagement in MVPA are both associated with the aforementioned demographic variables (Ebben and Brudzynski, 2010; Fountoulakis and Grogan, 2014; Markland and Ingledew, 2007; Martin et al., 2009). We hypothesized that individuals who are overweight, but perceive themselves as normal weight (i.e., underestimate weight) will engage in less MVPA than those who possess an accurate weight perception. Furthermore, we hypothesized that the potential association between weight perceptions and MVPA would vary by race, gender, age, and socioeconomic status. 2. Methods 2.1. Design and participants The data for this study came from the 2003–2006 National Health and Nutrition Education Survey (NHANES), which is conducted by the National Center for Health Statistics (NCHS) of the Centers for Disease Control and Prevention; these cycles were evaluated because these are the only current publically available cycles with accelerometry data. NHANES is a crosssectional, nationally representative survey of the noninstitutionalized US civilian population. The NHANES data are publicly released for 2-year survey periods to protect confidentiality and increase statistical reliability (Zipf et al., 2013). The present study included 5462 adults 20 and older; all participants provided written informed consent before data collection. The National Center for Health Statistics ethics committee approved the study. 2.2. Discrepant weight perception To assess discrepant weight perception, height and weight were objectively measured and used to calculate BMI. The participants were also asked, “Do you consider yourself now to be overweight, underweight, or about the right weight” (Kuchler and Variyam, 2003). We then created a discrepant weight variable by categorizing the participants into three subgroups: Those who have an accurate weight perception (i.e., measured BMI was between 18.5 kg/m2 and 24.9 kg/m2) and they self-reported being the “right weight.” We labeled this as “accurate weight perception”. Those who say that they are normal weight, but are overweight (i.e., BMI was N25 kg/m2), which we labeled as “underestimate weight.” Those who say that they are overweight, but are normal weight (BMI was between 18.5 kg/m2 and 24.9 kg/m2), which we labeled as “overestimate weight.” Notably, due to few participants considering themselves as underweight, these individuals were not evaluated herein.
2.3. Physical activity At the MEC, the participants who were able to walk were asked to wear an ActiGraph 7164 accelerometer on their right hip for 7 days. Accelerometers were fastened to an elastic belt worn around the participant's waist near the iliac crest. The participants were asked to wear the accelerometer during all activities, except during water-based activities and while sleeping. The
accelerometer measured the frequency, intensity, and duration of physical activity by generating an activity count proportional to the measured acceleration (Loprinzi et al., 2014). Additional details about the mechanics of accelerometry can be found elsewhere (Chen and Bassett, 2005). Estimates for physical activity were summarized in 1-minute epochs. Activity counts N 2020 per minute were classified as MVPA (Troiano et al., 2008). To determine the amount of time the monitor was worn, non-wear was defined by a period of a minimum of 60 consecutive minutes of zero activity counts, with the allowance of 1 to 2 min of activity counts between 0 and 100. The participants were given $40.00 upon the return of accelerometers.
2.4. Covariates Unless stated otherwise, all of the following covariates were included in each model: age, gender, race–ethnicity, and socioeconomic status. Age, gender, and race–ethnicity were self-reported. Socioeconomic status was assessed using the poverty income ratio (PIR). Ranging from 0 to 5, PIR was defined as the ratio of the family income to the federal poverty threshold. For example, a PIR of .5 suggests that the family income is 50% below the poverty threshold.
2.5. Data analysis A negative binomial logistical regression was used to assess the association between discrepant weight perception (independent variable) and engagement in MVPA (outcome variable). A negative binomial logistic regression was used because MVPA was highly skewed and in the form of count data. To examine whether any effect moderation was present, separate models were computed for gender, race–ethnicity, age group, and poverty level. These person-based (sub-group) tests of moderation were used to examine for whom (e.g., which individuals) the potential relationship was observable.
3. Results Table 1 displays the weighted sample characteristics. Table 2 displays the weighted regression associations between weight perception and MVPA, with results stratified by age, gender, race–ethnicity, and poverty level. When comparing the entire sample, and after adjustments, weight discrepancy was not associated with engagement in MVPA; for underestimate vs. accurate: RR = .98, 95% CI: .89–1.07, P = .62; for overestimate vs. accurate: RR = 1.08, 95% CI: .96–1.22, P = .21. However, there was some evidence of effect moderation by gender and age. Females who said that they are normal weight, but were in fact overweight, engaged in 13% less MVPA (RR = .87, 95% CI: .769–.999, P = .05). Also, older adults (N60) who said that they are normal weight, but were overweight, engaged in 23% less MVPA (RR = .77, 95% CI: .616–.965, P = .025). Race–ethnicity and poverty did not moderate the relationship between weight perception and MVPA engagement. Additional analyses included education status as a covariate, and with its inclusion, the results were unchanged (data not shown). Additionally, there was no evidence of multicollinearity in the models, as the mean variance inflation factor was 1.2, the highest individual variance inflation factor was 1.85, and the lowest tolerance statistic was .53.
Table 1 Weighted sample means (standard error) for selected characteristics of the NHANES 2003–2006 sample (N = 5462).
Age/Years % Female % NHWd PIR a b c d
Accuratea
Underb
Overc
46.5 (.5) 54.4 (.9) 73.5 (2.1) 3.16 (.6)
47.6 (.9) 27.2 (1.7) 63.6 (3.0) 2.93 (.8)
44.1 (1.0) 85.6 (2.3) 80.3 (3.1) 3.52 (.1)
Those who have an accurate weight perception. Those who say that they are normal weight, but are overweight. Those who say that they are overweight, but are normal weight. Non-Hispanic White.
L. Coleman, P.D. Loprinzi / Preventive Medicine 87 (2016) 47–50 Table 2 Weight perception and MVPA stratified by age, gender, race–ethnicity and poverty level, NHANES 2003–2006 (N = 5462). Model Entire sample Underestimate vs. accurate Overestimate vs. accurate Age b40 Underestimate vs. accurate Overestimate vs. accurate 40–60 Underestimate vs. accurate Overestimate vs. accurate N60 Underestimate vs. accurate Overestimate vs. accurate Gender Male Underestimate vs. accurate Overestimate vs. accurate Female Underestimate vs. accurate Overestimate vs. accurate Race–ethnicity Mexican American Underestimate vs. accurate Overestimate vs. accurate Other Hispanic Underestimate vs. accurate Overestimate vs. accurate Non-Hispanic White Underestimate vs. accurate Overestimate vs. accurate Non-Hispanic Black Underestimate vs. accurate Overestimate vs. accurate Other Hispanic Underestimate vs. accurate Overestimate vs. accurate Poverty level b1 Underestimate vs. accurate Overestimate vs. accurate ≥1 Underestimate vs. accurate Overestimate vs. accurate a b
RRa
95% CIb
.98 1.08
.89–1.07 .96–1.22
1.06 1.08
.92–1.23 .91–1.28
1.05 1.10
.92–1.20 .94–1.28
.77 1.09
.62–.97 .76–1.57
1.00 .91
.90–1.12 .74–1.14
.87 1.09
.77–.99 .96–1.24
1.03 .99
.90–1.18 .81–1.00
.87 1.33
.60–1.26 .62–2.82
.97 1.09
.86–1.08 .96–1.24
.93 .94
.79–1.10 .66–1.36
1.10 .76
.73–1.66 .49–1.21
1.07 1.33
.83–1.37 .98–1.80
.95 1.10
.87–1.04 .98–1.24
Rate ratio. Confidence interval.
4. Discussion The primary purpose of this study was to examine the association between weight perception and engagement in MVPA. We also examined the association of individuals' demographic profiles on engagement in MVPA. Specifically, the study sought to determine whether those who believe that they are normal weight, when they are in fact overweight, engage in less MVPA. The present study did not find that weight discrepancy was associated with engagement in MVPA, when evaluating the entire sample. However, in partial support of our hypothesis, the findings did suggest that overweight women, who underestimate their weight, are likely to engage in less MVPA. Furthermore, older adults (N60) who are overweight, and underestimated their weight, are likely to engage in less MVPA. Considering the paucity of research in this area, the role of weight perception deserves more attention when discussing engagement in MVPA among adults. The present findings, as well as previous research using nationally representative samples, have demonstrated that discrepant weight perceptions are prevalent among adults (Kuchler and Variyam, 2003; Ver Ploeg et al., 2008) and that some adults who possess a discrepant weight perception and underestimate weight are less likely to engage in MVPA (Duncan et al., 2011). Thus, our findings extend previously observed effects of weight perception on physical activity.
49
Study strengths include use of a large nationally representative sample and objectively measured weight, height, and MVPA. However, there are limitations that must be acknowledged. First, inferences based on these results are limited to associations due to the crosssectional design of our study. Well-designed prospective studies are required to determine whether weight perception is in fact causally related to MVPA. If so, there is an additional need to examine whether weight perception reduces the likelihood of engagement in MVPA among those adults who are overweight, but believe that they are a normal weight. Also, given that the 2003–2006 NHANES data are approximately a decade old, replication of this study using more recent data is warranted. Although further research is needed to support the present findings, the suggested association between weight perception and engagement in MVPA among women and older adults has several implications. First, additional studies are needed to establish why individuals develop discrepant weight perceptions, including the mechanisms that influence their attitudes and ideals regarding weight and body size. Gaining further insight into discrepant weight perceptions can aid in elucidating the prevalence of overweight and obesity among the identified populations, as well as barriers to weight reduction and motivation to engage in positive health behaviors. Having a discrepant weight perception could possibly reduce the efficacy of programs aimed at weight loss or increasing MVPA; those individuals who possess a discrepant weight perception may lack the motivation necessary to elicit a desire to adopt healthy behaviors. Moreover, developing efficacious programs will require intervention-based focus on converging individuals' perceived and actual weight statuses. That convergence can also help individuals overcome barriers which constrain the adoption of positive health behaviors, such as engagement in MVPA. Targeted public health and education messages to women and older adults might be needed to help prevent discrepancies between perceived and actual weight. Since excess weight is considered problematic because it is a risk factor for chronic diseases, health care professionals should utilize unused opportunities to correct discrepant weight perceptions (Yaemsiri et al., 2011). Many individuals who possess discrepant weight perceptions may be unaware that their weight is problematic. Therefore, simply communicating risks or attempting to dictate change could impede efforts to correct weight perceptions and increase engagement in physical activity. A viable approach to consider is motivational interviewing, which allows health care professionals to promote change that is derived from individuals' own volition (Pollak et al., 2007). This might, in fact, reduce resistance to recommendations. Overall, exploring and addressing this phenomenon will likely improve personal health and wellness, thereby reducing negative health outcomes.
Conflict of interest No funding was used to prepare this manuscript and no conflicts are disclosed.
References Antin, T.M.J., Paschall, M.J., 2011. Weight perception, weight change intentions, and alcohol use among young adults. Body Image 8, 149–156. Chen, K.Y., Bassett Jr., D.R., 2005. The technology of accelerometry-based activity monitors: current and future. Med. Sci. Sports Exerc. 37, S490–S500. Duncan, D.T., Wolin, K.Y., Scharoun-Lee, M., Ding, E.L., Warner, E.T., Bennett, G.G., 2011. Does perception equal reality? Weight misperception in relation to weight-related attitudes and behaviors among overweight and obese US adults. Int. J. Behav. Nutr. Phys. Act. 8 (20). Ebben, W.P., Brudzynski, L., 2010. Body image as a motivator and barrier to exercise participation. Int. J. Exerc. Sci. 3 (1), 14–24. Eichen, D.M., Conner, B.T., Daly, B.P., Fauber, R.L., 2012. Weight perception, substance use, and disordered eating behaviors: comparing normal weight and overweight high school students. J. Youth Adolesc. 41, 1–13. Fountoulakis, C., Grogan, S., 2014. An investigation of the links between body image and exercise participation. Sport Exerc. Rev. 10 (3), 19–30.
50
L. Coleman, P.D. Loprinzi / Preventive Medicine 87 (2016) 47–50
Harring, H.A., Montgomery, K., Hardin, J., 2010. Perceptions of body weight, weight management strategies, and depressive symptoms among US college students. J. Am. Coll. Heal. 59 (1), 43–50. Kuchler, F., Variyam, J.N., 2003. Mistakes were made: misperception as a barrier to reducing overweight. Int. J. Obes. 27, 856–861. Loprinzi, P.D., 2015. Factors influencing the disconnect between self-reported health status and actual health profile: implications for improving self-awareness of health status. Prev. Med. 73, 37–39. Loprinzi, P.D., 2016a. Light-intensity physical activity and all-cause mortality. Am. J. Health Promot. (epub-ahead-print). Loprinzi, P.D., 2016b. Effects of light-intensity physical activity on red blood cell distribution width: implications for a novel mechanism through which light-intensity physical activity may influence cardiovascular disease. Int. J. Cardiol. 203, 724–725. Loprinzi, P.D., Darnell, T., Hager, K., Vidrine, J.I., 2015. Physical activity-related beliefs and discrepancies between beliefs and physical activity behavior for various chronic diseases. Physiol. Behav. 151, 577–582. Loprinzi, P.D., Walker, J.F., LeeH, 2014. Association between physical activity and inflammatory markers among U.S. adults with chronic obstructive pulmonary disease. Am. J. Health Promot. 29 (2), 81–88. Markland, D., Ingledew, D.K., 2007. The relationship between body mass and body image and relative autonomy for exercise among adolescent males and females. Psychol. Sport Exerc. 8, 836–853. Martin, M.A., Frisco, M.L., May, A.L., 2009. Gender and race/ethnic differences in inaccurate weight perceptions among U.S. adolescents. Women's Health Issues 19, 292–299.
Park, E., 2011. Overestimation and underestimation: adolescents' weight perception in comparison to BMI-based weight status and how it varies across sociodemographic factors. J. Sch. Health 81 (2), 57–64. Pollak, K.I., Ostbye, T., Alexander, S.C., et al., 2007. Empathy goes a long way in weight loss discussions. J. Fam. Pract. 56 (12), 1031–1036. Rosenstock, I.M., 1974. Historical origins of the health belief model. Health Educ. Behav. 2 (4), 328–335. Troiano, R.P., Berrigan, D., Dodd, K.W., Masse, L.C., Tilert, T., McDowell, M., 2008. Physical activity in the United States measured by accelerometer. Med. Sci. Sports Exerc. 40 (1), 181–188. U.S. Department of Health and Human Services, 2008. 2008 Physical Activity Guidelines for Americans. Author, Washington, DC. U.S. Department of Health and Human Services, 2016. Healthy People 2020. Author, Washington, DC. Ver Ploeg, M.L., Chang, H.H., Lin, B.H., 2008. Obesity 16 (9), 2120–2125, Over, under, or about right: misperceptions of body weight among food stamp participants. Yaemsiri, S., Slining, M.M., Agarwal, S.K., 2011. Int. J. Obes. 35, 1063–1070, Perceived weight status, overweight diagnosis, and weight control among US adults. Zipf, G., Chiappa, M., Porter, K., Ostchega, Y., Lewis, B., Dostal, J., 2013. The National Health and Nutrition Examination Survey: plan and operations, 1999–2010. Vital Health Stat. 1 (56), 10–22.
Contents lists available at ScienceDirect
Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed
Brief Original Report
The association between discrepant weight perceptions and objectively measured physical activity Lois Coleman, Paul D. Loprinzi ⁎ Department of Health, Exercise Science and Recreation Management, The University of Mississippi, USA
a r t i c l e
i n f o
Article history: Received 14 October 2015 Received in revised form 30 January 2016 Accepted 1 February 2016 Available online 11 February 2016 Keywords: Accelerometry Epidemiology Weight perception
a b s t r a c t Background. Some individuals perceive themselves as being normal weight, despite having an excess body fat percentage (e.g., underestimate weight). Conversely, other individuals perceive themselves as being overweight, despite having a normal body fat percentage (e.g., overestimate weight). When perceived and actual weight statuses are incongruent, individuals possess a discrepant weight perception. The association between discrepant weight perceptions and engagement in moderate-to-vigorous physical activity (MVPA) has not been thoroughly investigated, which was this study's purpose. Methods. For this cross-sectional study, data from the 2003–2006 National Health and Nutrition Examination Survey were utilized (N = 5462 adults). MVPA was assessed via accelerometry. Based on measured body mass index and whether participants considered themselves as overweight, underweight, or about the right weight, we classified individuals as accurate perception, overestimate weight (discrepant), or underestimate weight (discrepant). A negative binomial logistical regression was used to assess the association between discrepant weight perception (independent variable) and engagement in MVPA (outcome variable). Results. Females who said that they are normal weight, but were in fact overweight based on body mass index, engaged in 13% less MVPA (rate ratio = .87, 95% confidence interval: .769–.999, P = .05). Also, older adults (N 60 yrs) who said that they are normal weight, but were overweight based on body mass index, engaged in 23% less MVPA (rate ratio = .77, 95% confidence interval: .616–.965, P = .025). Conclusion. Discrepant weight perceptions were associated with less objectively measured MVPA. Interventions should take weight perceptions into consideration when designing and evaluating intervention impact. © 2016 Elsevier Inc. All rights reserved.
1. Introduction Weight perception has been defined as how an individual perceives his or her body weight, and is affected by their “ideal” and “normative” body images (Park, 2011). Some individuals perceive themselves as being normal weight, despite having an excess body fat percentage (e.g., underestimate weight). Conversely, other individuals perceive themselves as being overweight, despite having a normal body fat percentage (e.g., overestimate weight) (Duncan et al., 2011; Harring et al., 2010). Furthermore, males are more likely to underestimate their weight, in contrast to females who generally overestimate their weight (Kuchler and Variyam, 2003; Yaemsiri et al., 2011; Park, 2011). Discrepant weight perceptions may be problematic in terms of both health behaviors and resulting health outcomes (Loprinzi, 2015; Loprinzi et al., 2015). This is in alignment with various theoretical models, including the Health Belief Model, which posits that beliefs ⁎ Corresponding author at: Center for Health Behavior Research, Department of Health, Exercise Science and Recreation Management, The University of Mississippi, 229 Turner Center, University, MS 38677, USA. E-mail address: [email protected] (P.D. Loprinzi).
http://dx.doi.org/10.1016/j.ypmed.2016.02.004 0091-7435/© 2016 Elsevier Inc. All rights reserved.
and attitudes play an important role in behavioral initiation and maintenance (Rosenstock, 1974). Studies suggest that the discrepancy not only contributes to the current obesity epidemic (Park, 2011), but individuals who possess a discrepant weight perception are likely to engage in negative health behaviors including: substance use, alcohol abuse, disordered eating, and smoking (Antin and Paschall, 2011; Eichen et al., 2012). They are also at an increased risk of suffering from depression (Harring et al., 2010). This is concerning as these potentially negative consequences of discrepant weight perceptions may influence health-enhancing behaviors, such as physical activity. However, no recent studies to our knowledge, have examined the association between weight perception and engagement in accelerometerassessed moderate-to-vigorous physical activity (MVPA) using a nationally representative sample. While even light physical activity is considered beneficial (Loprinzi, 2016a, 2016b), exploring the link between weight perception and MVPA is important given that MVPA is associated with numerous positive health outcomes including: a reduced risk for certain cancers, type 2 diabetes, coronary heart disease, stroke, and metabolic syndrome; improved cognitive function, sleep quality, and bone density; and lower levels of depression (U.S. Department of Health and Human
48
L. Coleman, P.D. Loprinzi / Preventive Medicine 87 (2016) 47–50
Services, 2008). Also, national physical activity objectives for Healthy People 2020 include increasing the proportion of adults who engage in physical activity of at least moderate intensity; weight perceptions are a potential barrier to reaching those objectives (U.S. Department of Health and Human Services, 2016). Consequently, there is a need to investigate whether there is an association between weight perception and engagement in MVPA. Therefore, the primary purpose of this study was to examine the association between weight perception and engagement in MVPA using a nationally representative sample. A secondary objective was to determine if individuals' demographic profiles (race, gender, age, and socioeconomic status) moderate the potential association between weight perception and engagement in MVPA. This is plausible because previous research has demonstrated that weight perception and engagement in MVPA are both associated with the aforementioned demographic variables (Ebben and Brudzynski, 2010; Fountoulakis and Grogan, 2014; Markland and Ingledew, 2007; Martin et al., 2009). We hypothesized that individuals who are overweight, but perceive themselves as normal weight (i.e., underestimate weight) will engage in less MVPA than those who possess an accurate weight perception. Furthermore, we hypothesized that the potential association between weight perceptions and MVPA would vary by race, gender, age, and socioeconomic status. 2. Methods 2.1. Design and participants The data for this study came from the 2003–2006 National Health and Nutrition Education Survey (NHANES), which is conducted by the National Center for Health Statistics (NCHS) of the Centers for Disease Control and Prevention; these cycles were evaluated because these are the only current publically available cycles with accelerometry data. NHANES is a crosssectional, nationally representative survey of the noninstitutionalized US civilian population. The NHANES data are publicly released for 2-year survey periods to protect confidentiality and increase statistical reliability (Zipf et al., 2013). The present study included 5462 adults 20 and older; all participants provided written informed consent before data collection. The National Center for Health Statistics ethics committee approved the study. 2.2. Discrepant weight perception To assess discrepant weight perception, height and weight were objectively measured and used to calculate BMI. The participants were also asked, “Do you consider yourself now to be overweight, underweight, or about the right weight” (Kuchler and Variyam, 2003). We then created a discrepant weight variable by categorizing the participants into three subgroups: Those who have an accurate weight perception (i.e., measured BMI was between 18.5 kg/m2 and 24.9 kg/m2) and they self-reported being the “right weight.” We labeled this as “accurate weight perception”. Those who say that they are normal weight, but are overweight (i.e., BMI was N25 kg/m2), which we labeled as “underestimate weight.” Those who say that they are overweight, but are normal weight (BMI was between 18.5 kg/m2 and 24.9 kg/m2), which we labeled as “overestimate weight.” Notably, due to few participants considering themselves as underweight, these individuals were not evaluated herein.
2.3. Physical activity At the MEC, the participants who were able to walk were asked to wear an ActiGraph 7164 accelerometer on their right hip for 7 days. Accelerometers were fastened to an elastic belt worn around the participant's waist near the iliac crest. The participants were asked to wear the accelerometer during all activities, except during water-based activities and while sleeping. The
accelerometer measured the frequency, intensity, and duration of physical activity by generating an activity count proportional to the measured acceleration (Loprinzi et al., 2014). Additional details about the mechanics of accelerometry can be found elsewhere (Chen and Bassett, 2005). Estimates for physical activity were summarized in 1-minute epochs. Activity counts N 2020 per minute were classified as MVPA (Troiano et al., 2008). To determine the amount of time the monitor was worn, non-wear was defined by a period of a minimum of 60 consecutive minutes of zero activity counts, with the allowance of 1 to 2 min of activity counts between 0 and 100. The participants were given $40.00 upon the return of accelerometers.
2.4. Covariates Unless stated otherwise, all of the following covariates were included in each model: age, gender, race–ethnicity, and socioeconomic status. Age, gender, and race–ethnicity were self-reported. Socioeconomic status was assessed using the poverty income ratio (PIR). Ranging from 0 to 5, PIR was defined as the ratio of the family income to the federal poverty threshold. For example, a PIR of .5 suggests that the family income is 50% below the poverty threshold.
2.5. Data analysis A negative binomial logistical regression was used to assess the association between discrepant weight perception (independent variable) and engagement in MVPA (outcome variable). A negative binomial logistic regression was used because MVPA was highly skewed and in the form of count data. To examine whether any effect moderation was present, separate models were computed for gender, race–ethnicity, age group, and poverty level. These person-based (sub-group) tests of moderation were used to examine for whom (e.g., which individuals) the potential relationship was observable.
3. Results Table 1 displays the weighted sample characteristics. Table 2 displays the weighted regression associations between weight perception and MVPA, with results stratified by age, gender, race–ethnicity, and poverty level. When comparing the entire sample, and after adjustments, weight discrepancy was not associated with engagement in MVPA; for underestimate vs. accurate: RR = .98, 95% CI: .89–1.07, P = .62; for overestimate vs. accurate: RR = 1.08, 95% CI: .96–1.22, P = .21. However, there was some evidence of effect moderation by gender and age. Females who said that they are normal weight, but were in fact overweight, engaged in 13% less MVPA (RR = .87, 95% CI: .769–.999, P = .05). Also, older adults (N60) who said that they are normal weight, but were overweight, engaged in 23% less MVPA (RR = .77, 95% CI: .616–.965, P = .025). Race–ethnicity and poverty did not moderate the relationship between weight perception and MVPA engagement. Additional analyses included education status as a covariate, and with its inclusion, the results were unchanged (data not shown). Additionally, there was no evidence of multicollinearity in the models, as the mean variance inflation factor was 1.2, the highest individual variance inflation factor was 1.85, and the lowest tolerance statistic was .53.
Table 1 Weighted sample means (standard error) for selected characteristics of the NHANES 2003–2006 sample (N = 5462).
Age/Years % Female % NHWd PIR a b c d
Accuratea
Underb
Overc
46.5 (.5) 54.4 (.9) 73.5 (2.1) 3.16 (.6)
47.6 (.9) 27.2 (1.7) 63.6 (3.0) 2.93 (.8)
44.1 (1.0) 85.6 (2.3) 80.3 (3.1) 3.52 (.1)
Those who have an accurate weight perception. Those who say that they are normal weight, but are overweight. Those who say that they are overweight, but are normal weight. Non-Hispanic White.
L. Coleman, P.D. Loprinzi / Preventive Medicine 87 (2016) 47–50 Table 2 Weight perception and MVPA stratified by age, gender, race–ethnicity and poverty level, NHANES 2003–2006 (N = 5462). Model Entire sample Underestimate vs. accurate Overestimate vs. accurate Age b40 Underestimate vs. accurate Overestimate vs. accurate 40–60 Underestimate vs. accurate Overestimate vs. accurate N60 Underestimate vs. accurate Overestimate vs. accurate Gender Male Underestimate vs. accurate Overestimate vs. accurate Female Underestimate vs. accurate Overestimate vs. accurate Race–ethnicity Mexican American Underestimate vs. accurate Overestimate vs. accurate Other Hispanic Underestimate vs. accurate Overestimate vs. accurate Non-Hispanic White Underestimate vs. accurate Overestimate vs. accurate Non-Hispanic Black Underestimate vs. accurate Overestimate vs. accurate Other Hispanic Underestimate vs. accurate Overestimate vs. accurate Poverty level b1 Underestimate vs. accurate Overestimate vs. accurate ≥1 Underestimate vs. accurate Overestimate vs. accurate a b
RRa
95% CIb
.98 1.08
.89–1.07 .96–1.22
1.06 1.08
.92–1.23 .91–1.28
1.05 1.10
.92–1.20 .94–1.28
.77 1.09
.62–.97 .76–1.57
1.00 .91
.90–1.12 .74–1.14
.87 1.09
.77–.99 .96–1.24
1.03 .99
.90–1.18 .81–1.00
.87 1.33
.60–1.26 .62–2.82
.97 1.09
.86–1.08 .96–1.24
.93 .94
.79–1.10 .66–1.36
1.10 .76
.73–1.66 .49–1.21
1.07 1.33
.83–1.37 .98–1.80
.95 1.10
.87–1.04 .98–1.24
Rate ratio. Confidence interval.
4. Discussion The primary purpose of this study was to examine the association between weight perception and engagement in MVPA. We also examined the association of individuals' demographic profiles on engagement in MVPA. Specifically, the study sought to determine whether those who believe that they are normal weight, when they are in fact overweight, engage in less MVPA. The present study did not find that weight discrepancy was associated with engagement in MVPA, when evaluating the entire sample. However, in partial support of our hypothesis, the findings did suggest that overweight women, who underestimate their weight, are likely to engage in less MVPA. Furthermore, older adults (N60) who are overweight, and underestimated their weight, are likely to engage in less MVPA. Considering the paucity of research in this area, the role of weight perception deserves more attention when discussing engagement in MVPA among adults. The present findings, as well as previous research using nationally representative samples, have demonstrated that discrepant weight perceptions are prevalent among adults (Kuchler and Variyam, 2003; Ver Ploeg et al., 2008) and that some adults who possess a discrepant weight perception and underestimate weight are less likely to engage in MVPA (Duncan et al., 2011). Thus, our findings extend previously observed effects of weight perception on physical activity.
49
Study strengths include use of a large nationally representative sample and objectively measured weight, height, and MVPA. However, there are limitations that must be acknowledged. First, inferences based on these results are limited to associations due to the crosssectional design of our study. Well-designed prospective studies are required to determine whether weight perception is in fact causally related to MVPA. If so, there is an additional need to examine whether weight perception reduces the likelihood of engagement in MVPA among those adults who are overweight, but believe that they are a normal weight. Also, given that the 2003–2006 NHANES data are approximately a decade old, replication of this study using more recent data is warranted. Although further research is needed to support the present findings, the suggested association between weight perception and engagement in MVPA among women and older adults has several implications. First, additional studies are needed to establish why individuals develop discrepant weight perceptions, including the mechanisms that influence their attitudes and ideals regarding weight and body size. Gaining further insight into discrepant weight perceptions can aid in elucidating the prevalence of overweight and obesity among the identified populations, as well as barriers to weight reduction and motivation to engage in positive health behaviors. Having a discrepant weight perception could possibly reduce the efficacy of programs aimed at weight loss or increasing MVPA; those individuals who possess a discrepant weight perception may lack the motivation necessary to elicit a desire to adopt healthy behaviors. Moreover, developing efficacious programs will require intervention-based focus on converging individuals' perceived and actual weight statuses. That convergence can also help individuals overcome barriers which constrain the adoption of positive health behaviors, such as engagement in MVPA. Targeted public health and education messages to women and older adults might be needed to help prevent discrepancies between perceived and actual weight. Since excess weight is considered problematic because it is a risk factor for chronic diseases, health care professionals should utilize unused opportunities to correct discrepant weight perceptions (Yaemsiri et al., 2011). Many individuals who possess discrepant weight perceptions may be unaware that their weight is problematic. Therefore, simply communicating risks or attempting to dictate change could impede efforts to correct weight perceptions and increase engagement in physical activity. A viable approach to consider is motivational interviewing, which allows health care professionals to promote change that is derived from individuals' own volition (Pollak et al., 2007). This might, in fact, reduce resistance to recommendations. Overall, exploring and addressing this phenomenon will likely improve personal health and wellness, thereby reducing negative health outcomes.
Conflict of interest No funding was used to prepare this manuscript and no conflicts are disclosed.
References Antin, T.M.J., Paschall, M.J., 2011. Weight perception, weight change intentions, and alcohol use among young adults. Body Image 8, 149–156. Chen, K.Y., Bassett Jr., D.R., 2005. The technology of accelerometry-based activity monitors: current and future. Med. Sci. Sports Exerc. 37, S490–S500. Duncan, D.T., Wolin, K.Y., Scharoun-Lee, M., Ding, E.L., Warner, E.T., Bennett, G.G., 2011. Does perception equal reality? Weight misperception in relation to weight-related attitudes and behaviors among overweight and obese US adults. Int. J. Behav. Nutr. Phys. Act. 8 (20). Ebben, W.P., Brudzynski, L., 2010. Body image as a motivator and barrier to exercise participation. Int. J. Exerc. Sci. 3 (1), 14–24. Eichen, D.M., Conner, B.T., Daly, B.P., Fauber, R.L., 2012. Weight perception, substance use, and disordered eating behaviors: comparing normal weight and overweight high school students. J. Youth Adolesc. 41, 1–13. Fountoulakis, C., Grogan, S., 2014. An investigation of the links between body image and exercise participation. Sport Exerc. Rev. 10 (3), 19–30.
50
L. Coleman, P.D. Loprinzi / Preventive Medicine 87 (2016) 47–50
Harring, H.A., Montgomery, K., Hardin, J., 2010. Perceptions of body weight, weight management strategies, and depressive symptoms among US college students. J. Am. Coll. Heal. 59 (1), 43–50. Kuchler, F., Variyam, J.N., 2003. Mistakes were made: misperception as a barrier to reducing overweight. Int. J. Obes. 27, 856–861. Loprinzi, P.D., 2015. Factors influencing the disconnect between self-reported health status and actual health profile: implications for improving self-awareness of health status. Prev. Med. 73, 37–39. Loprinzi, P.D., 2016a. Light-intensity physical activity and all-cause mortality. Am. J. Health Promot. (epub-ahead-print). Loprinzi, P.D., 2016b. Effects of light-intensity physical activity on red blood cell distribution width: implications for a novel mechanism through which light-intensity physical activity may influence cardiovascular disease. Int. J. Cardiol. 203, 724–725. Loprinzi, P.D., Darnell, T., Hager, K., Vidrine, J.I., 2015. Physical activity-related beliefs and discrepancies between beliefs and physical activity behavior for various chronic diseases. Physiol. Behav. 151, 577–582. Loprinzi, P.D., Walker, J.F., LeeH, 2014. Association between physical activity and inflammatory markers among U.S. adults with chronic obstructive pulmonary disease. Am. J. Health Promot. 29 (2), 81–88. Markland, D., Ingledew, D.K., 2007. The relationship between body mass and body image and relative autonomy for exercise among adolescent males and females. Psychol. Sport Exerc. 8, 836–853. Martin, M.A., Frisco, M.L., May, A.L., 2009. Gender and race/ethnic differences in inaccurate weight perceptions among U.S. adolescents. Women's Health Issues 19, 292–299.
Park, E., 2011. Overestimation and underestimation: adolescents' weight perception in comparison to BMI-based weight status and how it varies across sociodemographic factors. J. Sch. Health 81 (2), 57–64. Pollak, K.I., Ostbye, T., Alexander, S.C., et al., 2007. Empathy goes a long way in weight loss discussions. J. Fam. Pract. 56 (12), 1031–1036. Rosenstock, I.M., 1974. Historical origins of the health belief model. Health Educ. Behav. 2 (4), 328–335. Troiano, R.P., Berrigan, D., Dodd, K.W., Masse, L.C., Tilert, T., McDowell, M., 2008. Physical activity in the United States measured by accelerometer. Med. Sci. Sports Exerc. 40 (1), 181–188. U.S. Department of Health and Human Services, 2008. 2008 Physical Activity Guidelines for Americans. Author, Washington, DC. U.S. Department of Health and Human Services, 2016. Healthy People 2020. Author, Washington, DC. Ver Ploeg, M.L., Chang, H.H., Lin, B.H., 2008. Obesity 16 (9), 2120–2125, Over, under, or about right: misperceptions of body weight among food stamp participants. Yaemsiri, S., Slining, M.M., Agarwal, S.K., 2011. Int. J. Obes. 35, 1063–1070, Perceived weight status, overweight diagnosis, and weight control among US adults. Zipf, G., Chiappa, M., Porter, K., Ostchega, Y., Lewis, B., Dostal, J., 2013. The National Health and Nutrition Examination Survey: plan and operations, 1999–2010. Vital Health Stat. 1 (56), 10–22.