Exploring Physical Self-Perceptions Among Survivors of Adolescent and Young Adult Cancer

Abstract

This study explored the factor structure of four subscales from the Physical Self-Description Questionnaire-Short Form (PSDQ-S). Associations between subscales and personal and medical factors were also examined. The analytic sample consisted of 89 survivors of adolescent and young adult cancer (M_{age at time of study} = 32.96 ± 4.37 years; M_{age at diagnosis} = 31.16 ± 4.84 years; 75.3% female). Confirmatory factor analysis suggested a reasonable fit to the data, indicating that the PSDQ-S subscales examined could be used in future investigations with this population [χ²(38) = 46.9, p = 0.15; Root Mean Square Error of Approximation = 0.05, 90% confidence interval = 0–0.10; Comparative Fit Index = 0.97; Standardized Root Mean Square of the Residuals = 0.07]. Multiple linear regressions showed that personal and medical factors accounted for a significant amount of variance in the body fat subscale, with female sex and higher body mass index being significantly associated with lower positive perceptions about the amount of one's body fat. More research examining the factor structure of the PSDQ-S subscales is warranted, and future investigations exploring personal, medical, and modifiable factors associated with physical self-perceptions should be conducted.

Background

Physical self-perceptions reflect how a person feels about his/her physical appearance and abilities.¹ Theoretically, physical self-perceptions are related to physical and global self-worth,¹ which are positively related to psychological well-being.² Furthermore, physical self-perceptions have been associated with a range of health-enhancing behaviors such as physical activity and nutrition.^3,4 As a result, physical self-perceptions are often considered important for both physical and psychological health—not least for young people for whom perceptions of physical capabilities are paramount.⁵ To facilitate investigations of physical self-perceptions among young people, Marsh et al.⁶ developed the Physical Self-Description Questionnaire (PSDQ). The PSDQ is widely acknowledged as a valuable, valid, and popular multidimensional instrument to assess physical self-perceptions during adolescence and young adulthood⁷—a period of life characterized by significant challenges and changes across physical, psychological, and social domains of functioning.⁸ To increase the use of the PSDQ, a short form (i.e., PSDQ-S) was also created.⁹

The reliability and various types of validity of PSDQ-S scores have been supported in the general population.¹⁰ However, researchers have seldom explored physical self-perceptions among young people living with chronic conditions nor have the psychometric properties of the PSDQ-S been tested in these populations. This is problematic, as individuals living with chronic conditions likely have impaired physical self-perceptions related to their disease and the treatments they had to endure. Survivors of adolescent and young adult (AYA) cancer are one vulnerable young population who experience numerous debilitating symptoms and side effects as a function of their disease and its treatments. Indeed, survivors of AYA cancer typically report changed physical fitness (e.g., increased weight status),¹¹ decreased confidence,¹² and a range of acute symptoms and side effects (e.g., fatigue, cardiovascular complications).^13,14 Thus, it is perhaps unsurprising that many survivors of AYA cancer report reduced perceptions of physical abilities and declines in perceptions of appearance.^15,16 To address calls to explore psychological aspects of health and well-being among survivors of AYA cancer, specifically as they relate to changed appearance and abilities,¹⁷ it is necessary to ensure physical self-perception questionnaires that can be used in future research with this population.

In a previous study, we conducted cognitive interviews, a technique to evaluate how participants interpret and respond to questionnaires,¹⁸ with seven survivors of AYA cancer.¹⁹ Items comprising the PSDQ strength, endurance, body fat, and appearance subscales were found to be clear, appropriate, and relevant; however, length was cited as problematic and participants recommended that the PSDQ-S be used.¹⁹ Notwithstanding the contributions of our previous study, cognitive interviews are not meant to confirm score validity or reliability.¹⁸ Thus, the first objective of this brief research report was to provide evidence for the factor structure of the PSDQ-S strength, endurance, body fat, and appearance subscales using confirmatory factor analysis (CFA) with survivors of AYA cancer. These subscales were selected because the items were deemed likely to capture survivors' perceptions of their physical appearance and abilities following their cancer treatment.^11–16 The second objective was to explore personal and medical factors related to physical self-perceptions as a way to provide insight into which subgroups may require interventions to optimize their psychological well-being. Drawing on theoretical and empirical evidence,^20,21 personal (i.e., age, sex, body composition, education, and employment) and medical factors (i.e., cancer type, time since treatment, and treatment received) were explored.

Methods

Participants

Survivors of AYA cancer were recruited through websites, organizations, and snowball sampling to participate in a larger study assessing the cross-sectional relationships between physical activity and a range of psychological outcomes. Individuals were eligible if they: (1) were diagnosed with cancer between the ages of 15–39 years; (2) had completed cancer treatment within the past 5 years; (3) showed no evidence of progressive or recurrent disease or of secondary or second cancers; and, (4) were able to read, understand, and provide informed consent in English. Survivors of AYA cancer were not eligible if they: (1) had physical impairments precluding participation in physical activity; and/or, (2) were unwilling or unable to provide informed consent and passive parental consent (the latter if they were <18 years).

Data collection

Interested and potentially eligible survivors of AYA cancer received an online study link and were directed to study objectives, inclusion criteria, and an informed consent page. After self-screening, confirming eligibility, and providing informed consent (and passive parental consent when indicated), participants could access the secure online survey hosted on, Qualtrics. At the end of the survey, all participants had the option to enter their email address to receive a $10.00 iTunes gift card.

Personal and medical factors

Participants completed questions about their age, sex, height and weight (used to calculate body mass index [BMI], a proxy measure of body composition), education, employment, cancer type, time since treatment, and treatment received. For the current study, education, employment, treatment received, and cancer type were dichotomized due to sparse endorsement of certain categories. Age, BMI, and time since treatment were treated as continuous.

Physical self-perceptions

Strength, endurance, body fat, and appearance subscales of the PSDQ-S were administered using a Likert-type scale ranging from 1 (false) to 6 (true).⁹ After reverse scoring three negatively worded items, subscale scores were calculated by averaging responses. Higher scores indicate more positive appraisals of physical self-perceptions.

Data analysis

Data were analyzed in three phases. First, data were examined for patterns of missing data, normality, multicollinearity, and potential univariate and multivariate outliers following recommended procedures²² using SPSS (version 25). Of the 116 who completed the questionnaires, 27 were excluded from the analyses because they were missing >5% of data (n = 7), had provided insufficient information to confirm eligibility (n = 10), or were ineligible based on reported age at diagnoses and/or time since treatment (n = 10). Participants who were missing <5% of data had their missing data replaced using their mean score (from remaining items) on the subscale.²³ No deviations from normality or issues of multicollinearity were observed. Two multivariate outliers were identified and retained after running the analyses with and without them and obtaining the same results. Next, descriptive statistics (i.e., means, standard deviations, and frequencies) were computed for all study variables.

Second, CFA was used to test the factor structure of the PSDQ-S strength, endurance, body fat, and appearance subscales (objective 1) using MPlus (version 7.31). A correlated four-factor CFA with Robust Maximum Likelihood estimation was tested.²⁴ The fit of the model was examined using chi-square (χ²) test, Root Mean Square Error of Approximation (RMSEA) with its corresponding 90% confidence interval (90% CI), Comparative Fit Index (CFI), and Standardized Root Mean Square of the Residuals (SRMR).²⁵ A nonsignificant p value for the χ² test, values of ≥0.90 for the CFI, values <0.08 for the RMSEA and SRMR, and the strength of the standardized factor loadings between each indicator and its corresponding latent variable were inspected to ascertain appropriate model fit.²⁵ Third, four linear regressions were estimated to explore the relationships between the PSDQ-S strength, endurance, body fat, and appearance subscales and personal and medical factors (objective 2) using SPSS (version 25).

Results

The analytic sample consisted of 89 AYAs (75.3% female; 88.6%^* self-identified as White) who were diagnosed with cancer between the ages of 17–39 years (M_{age at diagnosis} = 31.16 ± 4.84 years), who were aged 32.96 ± 4.37 years, and 2.08 ± 1.38 years post-treatment at the time of data collection. BMI ranged from 16.71–46.46 kg/m² (M_BMI = 25.38 ± 6.0 kg/m²). Most participants held a university degree (78.4%),* were working or transitioning to work (85.4%), and reported receiving one or two types of treatment (57.3%). Just under half of the sample had been diagnosed with breast cancer (49.4%). The remainder reported being diagnosed with leukemia (4.5%), lymphoma (11.2%), soft tissue sarcoma (1.1%), osteosarcoma (1.1%), melanoma (6.7%), other cancers (22.5%; e.g., thyroid, brain, colorectal, ovarian, bladder, and gynecological), and multiple cancers (3.4%). Table 1 depicts participants mean scores on the PSDQ-S subscales. There were no statistically significant differences between those participants diagnosed with breast cancer or other types of cancer on any of the PSDQ-S subscales. All subscale scores were comparable to those found in the original PSDQ-S validation study.⁹

Table 1.

Means, Standard Deviations, and Factor Loadings of PSDQ-S Strength, Endurance, Body Fat, and Appearance Items for Model 1a and Model 1b

			Model 1a ^b (AIC: 3146.95)		Model 1b ^b (AIC: 2854.63)
Item	Score range ^a	Mean ± SD ^a	Factor loadings	R ²	Factor loadings	R ²
Factor 1: Strength (ST)
ST1. I am a physically strong person	1.00–6.00	3.53 ± 1.29	0.44^*	0.19	—	—
ST2. I have a lot of power in my body	1.00–6.00	3.89 ± 1.06	0.71^**	0.50^*	0.66^**	0.44^*
ST3. I could do well in a test of strength	1.00–6.00	3.73 ± 1.16	0.74^**	0.55^*	0.86^**	0.74^**
Subscale total	1.33–6.00	3.72 ± 0.88
Modified model (Model 1b) subscale total^c	1.00–6.00	3.81 ± 0.98
Factor 2: Endurance (EN)
EN1. I can run a long way without stopping	1.00–6.00	2.99 ± 1.28	0.84^**	0.70^**	0.84^**	0.70^**
EN2. I can be physically active for a long period of time without getting tired	1.00–6.00	3.21 ± 1.42	0.83^**	0.69^**	0.83^**	0.69^**
EN3. I am good at endurance activities like distance running, aerobics, bicycling, swimming, or cross-country skiing	1.00–6.00	3.05 ± 1.21	0.72^**	0.52^**	0.72^**	0.53^**
Subscale total	1.00–6.00	3.08 ± 1.14
Factor 3: Body fat (BF)
BF1. My waist is too large^d	1.00–6.00	3.70 ± 1.62	0.79^**	0.63^**	0.79^**	0.63^**
BF2. I have too much fat on my body^d	1.00–6.00	3.37 ± 1.58	0.86^**	0.74^**	0.86^**	0.74^**
BF3. I am overweight^d	1.00–6.00	3.76 ± 1.84	0.94^**	0.89^**	0.94^**	0.89^**
Subscale total	1.00–6.00	3.61 ± 1.54
Factor 4: Appearance (APP)
APP1. I have a nice looking face	2.00–6.00	4.23 ± 0.97	0.60^**	0.36^*	0.59^**	0.35^*
APP2. I am better looking than most of my friends	1.00–6.00	3.32 ± 0.92	0.63^**	0.40^*	0.63^**	0.40^*
APP3. I am good looking	1.00–6.00	4.12 ± 0.89	0.77^**	0.59^*	0.77^**	0.60^*
Subscale total	2.00–5.67	3.89 ± 0.74

Derived from SPSS.

Derived from Mplus.

2-Item strength subscale (i.e., ST2 and ST3).

Denotes reverse scored item.

p < 0.05; ^**p < 0.001.

AIC, Akaike Information Criterion; PSDQ-S, Physical Self-Description Questionnaire-Short Form; SD, standard deviation.

Objective 1

Three of the four fit statistics [χ²(48) = 78.77, p < 0.05; RMSEA = 0.09, 90% CI = 0.05–0.12; SRMR = 0.10] suggested that the correlated four-factor model did not fit the data adequately. Therefore, factor loadings and the modification index were inspected. Factor loadings (Table 1) showed that ST1 was below recommended cutoff criteria²⁵ and that it cross-loaded. As well, the modification index suggested that removing ST1 would lead to the largest decrease in the model's χ² test. After removing ST1, all four fit statistics [χ²(38) = 46.9, p = 0.15; RMSEA = 0.05, 90% CI = 0–0.10; CFI = 0.97; SRMR = 0.07] and factor loadings suggested an adequate fit for a correlated four-factor CFA model (Model 1b). Moreover, the Akaike Information Criterion (AIC) value for the modified model was lower than the original model, suggesting an improved fit. Thus, Model 1b was retained.

Objective 2

Subscale scores for strength (excluding ST1; α = 0.72), endurance (α = 0.84), body fat (α = 0.90), and appearance (α = 0.70) were used in the subsequent regression analyses as dependent variables (Table 2). Personal and medical factors did not account for a significant amount of variance in perceptions of strength, endurance, and appearance (R² = 0.03–0.16), but did for perceptions of body fat (R² = 0.54), whereby sex (Std. βs = −0.26) and BMI (Std. βs = −0.64) were significant negative correlates of positive feelings about amount of one's body fat. Female sex and higher BMI were related to less positive perceptions of amount of body fat.

Table 2.

Linear Regression Results for PSDQ-S Subscales

	Strength ^a			Endurance			Body fat			Appearance
	F	Std. β	p	F	Std. β	p	F	Std. β	p	F	Std. β	p
	0.30		0.97	1.50		0.17	11.45		<0.001	1.91		0.07
Personal factors
Age (years)		0.09	0.48		−0.08	0.48		0.09	0.30		0.07	0.54
Sex (0 = male, 1 = female)		−0.10	0.50		−0.25	0.08		−0.26	0.01		0.19	0.18
Body mass index (kg/m²)		−0.04	0.72		−0.23	0.04		−0.64	<0.001		−0.20	0.08
Education (0 = some high school/university, 1 = completed university)		−0.04	0.72		0.02	0.86		−0.05	0.52		0.13	0.24
Employment (0 = working or transitioning to work, 1 = not working)		0.00	0.99		−0.05	0.64		−0.07	0.38		0.28	0.01
Medical factors
Time since treatment (years)		−0.07	0.57		−0.19	0.10		0.02	0.84		−0.03	0.78
Diagnosis (0 = breast cancer, 1 = other cancer)		−0.12	0.44		−0.08	0.60		−0.03	0.81		0.21	0.14
Treatments (0 = one or two treatment(s), 1 = three to five treatments)		0.03	0.80		0.06	0.57		−0.08	0.33		−0.02	0.86

Strength R² = 0.03; endurance R² = 0.13; body fat R² = 0.54; appearance R² = 0.16.

2-Item strength subscale (i.e., ST2 and ST3).

Discussion

Survivors of AYA cancer may feel negatively about their physical appearance and abilities as a result of their cancer and treatments. However, few researchers have examined physical self-perceptions among survivors of AYA cancer—a population for whom evaluations of appearance and competencies are central.⁵ Moreover, there has been a lack of research focused on identifying subgroups who may require interventions to optimize their physical self-perceptions and ultimately psychological well-being. To begin to address these gaps, we first sought to examine whether the factor structure of four of the PSDQ-S subscales could be confirmed in a sample of survivors of AYA cancer so that researchers can be confident when selecting a measure to assess physical self-perceptions in this population. After excluding one item, CFA results suggested an adequate fit of the modified correlated four-factor model. Furthermore, we found that personal and medical factors were not consistently correlated with the PSDQ-S subscales.

When using questionnaires developed with and for different populations, it is important to confirm whether the hypothesized relationships among observed and latent variables are theoretically reasonable and structurally sound. Our CFA led to a re-specification of the a priori model of the PSDQ-S subscales. Using a combination of criteria (i.e., factor loadings, modification index, and AIC values), a model excluding one item (i.e., ST1; I am a physically strong person) outperformed the original model, suggesting that the correct model in this population may be 11 items instead of 12. It would be useful to understand why ST1 did not load adequately on the factor representing strength. Among older adult breast cancer survivors, researchers have found that physical strength captures one's perception of their capacity to win a “fight” or “battle” over their disease.²⁶ Thus, it is possible that asking survivors of AYA cancer to reflect on whether they believe they are a physically strong person may not assess strength perceptions as intended by the PSDQ-S strength subscales.

In terms of exploring factors associated with the PSDQ-S subscales, we found that personal and medical factors were not consistently correlated with the PSDQ-S subscales. Although this is consistent with research among older patients and survivors of cancer,²⁷ the dearth of research concerning physical self-perceptions among survivors of AYA cancer makes it difficult to properly evaluate whether the lack of associations is specific to this sample or reflective of a lack of associations in general. Plausibly, the delays achieving developmental milestones and interrupted academic/career trajectories commonly reported among survivors of AYA cancer may impact how age, education, and employment relate to physical self-perceptions. Further work delineating these relationships with survivors of AYA cancer is required. With regards to medical factors, drawing on posttraumatic growth perspectives,²⁸ receiving a life threatening diagnosis, and enduring treatments may alter how some survivors of AYA cancer feel about their body such that they may: (1) develop techniques to manage their changed appearance and abilities; (2) feel less pressure to achieve body appearance and ability ideals; and/or, (3) have a greater appreciation for their body. For others, cancer and its treatment may have the opposite effect. These differing responses to the same traumatic event (i.e., cancer diagnosis and treatment) may cancel one another out in analyses. Further work exploring how survivors of AYA cancer feel about their body and capacities during and after treatments may elucidate if/when survivors of AYA cancer engage in regulatory posttraumatic growth processes and if/how these are related to physical self-perceptions.

At the same time, sex and BMI were negative correlates of body fat perceptions. Specifically, females were significantly more likely than males to have less positive perceptions of amount of one's body fat, supporting results from a study conducted by Campisi et al.²⁹ in which women were shown to have worse perceptions about their amount of body fat than men. Sex-specific intervention strategies that target women may be needed to help reduce negative perceptions and promote positive perceptions. In addition, those with greater BMI had less positive perceptions of amount of body fat. Given that BMI is modifiable, this finding is encouraging because reducing BMI may lead to more positive perceptions of one's body fat. Accordingly, applied researchers may be interested in designing interventions to reduce BMI and assess if changes in survivors of AYA cancer's body composition impact their perceptions of body fat.

Limitations and future directions

The findings presented should be considered in light of the study limitations. The correlated four-factor model was tested with a relatively small homogenous sample. Although models have been fit with data sets containing <100 observations,³⁰ replicating findings with data sets containing >100 observations will be important to ensure proper estimates. Moreover, after modifying a model, cross-validation is highly recommended to provide more support for the respecified structure of the model.³¹ It should also be noted that the data for this study were taken from a larger study assessing the cross-sectional relationships between physical activity and psychological outcomes. It is possible that the individuals who self-referred into the study may differ in physical self-perceptions from those who chose not to participate. To address this, additional research should examine the PSDQ-S subscales with survivors of AYA cancer who are/are not interested in physical activity and who may or may not be participating in physical activity. In the same vein, it would also be useful to explore the PSDQ-S subscales with different cancer types and over time to establish invariance and facilitate comparisons between diagnoses and across the cancer trajectory. Finally, numerous psychological and behavioral factors are likely associated with physical self-perceptions. Self-efficacy, self-esteem, and physical activity should be considered because of known associations with physical self-perceptions.^1,2

Conclusion

Researchers often rely on measures developed in other populations when doing research with survivors of AYA cancer. This study indicates that the PSDQ-S endurance, body fat, and appearance subscales and a truncated strength subscale perform well with survivors of AYA cancer. As personal and medical factors were not associated with strength, endurance, or appearance perceptions, it may be important to go beyond personal and medical factors when trying to identify which subgroups require interventions. It is hoped that this study provides impetus to continue exploring physical self-perceptions, its antecedents and correlates, and potential intervention strategies among survivors of AYA cancer. Given that physical self-perceptions are important for overall physical and mental health, this will be a worthwhile line of inquiry.

Footnotes

Acknowledgments

University Research Ethics Board approval was obtained before conducting this study. As well, informed consent was obtained from all individual participants included in the study. The authors thank the individuals who participated in this study. This work was conducted while the first author was supported by a Vanier Canada Graduate Scholarship.

Author Disclosure Statement

No competing financial interests exist.

*

Missing 1 case; denominator n = 88.

References

Shavelson

, Hubner

, Stanton

. Self-concept: validation of construct interpretations. Rev Educ Res. 1976; 46(3):407–41.

Fox

. Self-esteem, self-perceptions, and exercise. Int J Sport Psychol. 2000; 31:228–40.

Webb

, Benjamin

, Gammon

, et al. Physical activity, sedentary behaviour and physical self-perceptions in adolescent girls: a mediation analysis. Ment Health Phys Act. 2013; 6(1):24–9.

Christiansen

, Lund-Cramer

, Brondeel

, et al. Improving children's physical self-perception through a school-based physical activity intervention: the move for well-being in school study. Ment Health Phys Act. 2018; 14:31–8.

Bucchianeri

, Arikian

, Hannan

, et al. Body dissatisfaction from adolescence to young adulthood: findings from a 10-year longitudinal study. Body Image. 2013; 10(1):1–7.

Marsh

, Richards

, Johnson

, et al. Physical self-description questionnaire: psychometric properties and a multitrait-multimethod analysis of relations to existing instruments. J Sport Exerc Psychol. 1994; 16:270–305.

Marsh

, Cheng

JHS

. Physical self-concept. In: Tenenbaum G, Eklund R, Kamata A (Eds). Handbook of measurement in sport and exercise psychology. Illinois: Human Kinetics; 2012.

Lenz

. The transition from adolescence to young adulthood: a theoretical perspective. J Sch Nurs. 2001; 17(6):300–306.

Marsh

, Martin

, Jackson

. Introducing a short version of the physical self description questionnaire: new strategies, short-form evaluative criteria, and applications of factor analyses. J Sport Exerc Psychol. 2010; 32(4):438–82.

10.

Maiano

, Morin

, Mascret

. Psychometric properties of the short form of the Physical Self-Description Questionnaire in a French adolescent sample. Body Image. 2015; 12:89–97.

11.

Tai

, Buchanan

, Townsend

, et al. Health status of adolescent and young adult cancer survivors. Cancer. 2012; 118(19):4884–91.

12.

Perkins

, Baum

, Taylor

, Basen-Engquist

. Effects of treatment factors, comorbidities and health-related quality of life on self-efficacy for physical activity in cancer survivors. Psychooncology. 2009; 18(4):405–11.

13.

Spathis

, Booth

, Grove

, et al. Teenage and young adult cancer-related fatigue is prevalent, distressing, and neglected: it is time to intervene. A systematic literature review and narrative synthesis. J Adolesc Young Adult Oncol. 2015; 4(1):3–17.

14.

Chao

, Xu

, Bhatia

, et al. Cardiovascular disease risk profiles in survivors of adolescent and young adult (AYA) cancer: the Kaiser Permanente AYA Cancer Survivors Study. J Clin Oncol. 2016; 34(14):1626–33.

15.

Smith

, Bellizzi

, Keegan

, et al. Health-related quality of life of adolescent and young adult patients with cancer in the United States: the adolescent and young adult health outcomes and patient experience study. J Clin Oncol. 2013; 31(17):2136–45.

16.

Rosenberg

, Tamimi

, Gelber

, et al. Body image in recently diagnosed young women with early breast cancer. Psychooncology. 2013; 22(8):1849–55.

17.

Brunet

, Wurz

, Shallwani

. A scoping review of studies exploring physical activity among adolescents and young adults diagnosed with cancer. Psychooncology. 2018; 27(8):1875–88.

18.

Willis

(Eds). Cognitive interviewing a tool for improving questionnaire design. California: Sage Publications Ltd.; 2005.

19.

Wurz

, Brunet

. Evaluating questionnaires used to assess physical activity and psychosocial outcomes among survivors of adolescent and young adult cancer: a cognitive interview study. J Adolesc Young Adult Oncol. 2017; 6(3):482–88.

20.

Grogan

(Eds). Body image: understanding body dissatisfaction in men, women and children. New York: Routledge; 2016.

21.

Fingeret

, Teo

, Epner

. Managing body image difficulties of adult cancer patients: lessons from available research. Cancer. 2014; 120(5):633–41.

22.

Tabachnick

, Fidell

(Eds). Using multivariate statistic. New York: Pearson; 2007.

23.

Rowe

, Mahar

, Raedeke

, Lore

. Measuring physical activity in children with pedometers: reliability, reactivity, and replacement of missing data. Ped Exerc Sci. 2004; 16:343–54.

24.

Muthén

, Muthén

(Eds). Mplus user's guide. California: Muthén & Muthén; 1998–2012.

25.

Brown

. Confirmatory factor analysis for applied research. New York: The Guilford Press; 2014.

26.

Documet

, Trauth

, Key

, et al. Breast cancer survivors' perception of survivorship. Oncol Nurs Forum. 2012; 39(3):309–15.

27.

de Morais

, Freitas-Junior

, Rahal

, Gonzaga

. Sociodemographic and clinical factors affecting body image, sexual function and sexual satisfaction in women with breast cancer. J Clin Nurs. 2016; 25(11–12):1557–65.

28.

Hefferon

, Grealy

, Mutrie

. Transforming from cocoon to butterfly: the potential role of the body in the process of posttraumatic growth. J Humanist Psychol. 2010; 50(2):224–7.

29.

Campisi

, Finn

, Bravo

, et al. Sex and age-related differences in perceived, desired and measured percentage body fat among adults. J Hum Nutr Diet. 2015; 28(5):486–92.

30.

Lai

, Cella

, Peterman

, et al. Anorexia/cachexia-related quality of life for children with cancer. Cancer. 2005; 104(7):1531–9.

31.

MacCallum

, Roznowski

, Mar

, Reith

. Alternative strategies for cross-validation of covariance structure models. Multivariate Behav Res. 1994; 29(1):1–32.