Table 2 Relevant outcome measures and results of reviewed studies

Boisvert et al. (2013) (Canada) [38]

EDI- 1 item from each of DT, B, BD

Religiosity- ‘I would describe myself as religious’ (King et al., 2001)

8 IV

Correlation- Spirituality and eating disorder symptomatology (r = −.19, p < .001).

Spirituality- SWBS- 2 items from EWB subscale

ANCOVA- Spirituality and eating disorder symptomatology (F(1, 497) = 13.74, p < .001).

Boisvert et al. (2012) (Canada) [37]

EDI- 1 item from each of DT, B, BD

Religiosity- ‘I would describe myself as religious’. (King et al., 2001)

8 IV

Correlation- Eating disorder symptomatology not significantly related to spirituality (r = −.07) or religion (r = .08).

Spirituality- SWBS- 2 items from EWB subscale

Path analysis- 10 % of the variance in EDI was explained by BMI (b = .28, p < .001) and body shame (b = .15, p < .001). 6 % of the variance in body shame was explained by age (older men had higher EDI) (b = .15, p < .001), being Asian (b = .11, p < .01) and lower spirituality (b = −.15, p < .001). Spirituality had a sig indirect effect on EDI mediated by body shame (b = −.02).

Boyatzis et al. (2006) (USA) [35]

EDI – DT, B

Quest Scale

10 IV

Correlation- College freshmen + sophomores- Bulimia: Total Quest (r = .33, p < .01), Existential Questioning (r = .29, p < .05), Doubting as positive (r = .23, p < .10).

College juniors + seniors- Openness to Change: Bulimia (r = .28, p < .10), Drive for Thinness (r = .27, p < .10).

Buser (2013) (USA) [47]

BULIT-R

5 RCOPE subscales

13 IV

Correlation- Punishing God/Higher power reappraisal andbulimic symptoms (r = .25, p < .001). Other subscales not sig.

Punishing God/Higher power reappraisal, passive religious deferral, active religious surrender, benevolent religious reappraisal/spiritual support and religious focus.

Hierarchical Regression- 1st step: BMI + religious affiliation 7.9 % (F(2, 589) = 25.23. p < .001), 2^nd step: 5 coping scales 5.5 % (F(5, 584) = 7.469, p < .001) Cohen’s f² = .06. Punishing God/Higher power reappraisal- (β = .242, p < .001).

Mediation Analysis- Punishing God/Higher power reappraisal partially mediated relationship between stress and bulimic symptoms. Sobel Test (z = 2.975, p < .01).

Feinson et al. (2012) (Israel) [41]

DEB-SQ

Self-reported religious observance category

9 IV

ANOVA- No Sig Differences.

Forthun et al. (2003) (USA) [29]

FAD- GFF subscale

I/E-R

9 IV

Correlation- Intrinsic religiousness: Bulimia (r = −.13, p < .01), Drive for Thinness (r = −.07, p < .05).

FARS- items selected

Hierarchical regression + Simple Slope Analyses

EDI – DT, B

Family dysfunction- B: Family x Intrinsic (B = −0.53, p < .01), Family x Extrinsic (B = 0.45, p < .10).

DT: Family x Extrinsic (B = 0.69, p < .10).

Intrinsic buffered association between family and B. Extrinsic made people more vulnerable to B/DT.

Parental History- B: History x Extrinsic (B = 2.29, p < .01).

DT: History x Intrinsic (B = −2.45, p < .05), History x Extrinsic (B = 2.98, p < .05).

Intrinsic buffered association between history and DT. Extrinsic made people more vulnerable to B/DT.

Gates et al. (2009) (USA) [36]

EDI- B, DT

MQOS- Religious angst subscale

8 IV

ANOVA- Religious Angst: Higher levels of Drive for Thinness (F =1.86, p < 0.01), Bulimia (F =1.89, p < 0.01).

Gluck et al. (2002) (USA) [39]

EDE-Q

Orthodox Traditions Scale

9 IV

ANOVA- secular Jews had higher EDE-Q scores then Orthodox Jews (F[1,125] = 8.5; P = .004).

RIQ

Chi-square- Secular women were more likely to use laxatives to control their shape and weight (χ ² [1,123] = 5.8, p = .02).

MLR- SES + religious grouping: 11 % total EDE-Q (R² = .11, p = .001).

Jacobs-Pilipski et al. (2005) (USA) [33]

EDE

13-item self-report measure of spiritual and religious beliefs and practices.

9 IV

ANOVA- No significant relationships.

EDI-2- DT, B

Latzer et al. (2014) (USA) [43]

EDI

Brief R-COPE

9 IV

Correlation- Negative religious coping: total EDI (r = .28, p < .01), Drive for Thinness (r = .23 p < .05), Ineffectiveness (r = .24, p < .05), Maturity (r = .25, p < .05), EAT bulimia (r = .22, p < .05).

EAT-26

Regression- Negative religious coping (control BMI): total EDI (R² = .19, p < .01), Ineffectiveness (R² = .07, p < .05), Maturity (R² = .06, p < .05), EAT Bulimia (R² = .07, p < .05).

Latzer et al. (2007) (Israel) [40]

EDI-2

Level of Religiosity Questionnaire

10 IV

ANOVA- Religious observance: (low > Intermediate = high) total EDI (p < 0.008), Interpersonal Distrust (p < 0.03), Interceptive Awareness (p < 0.003), Asceticism (p < 0.01), Impulse Regulation (p < 0.02) and Social Insecurity (p < 0.03)

Correlation- Religiosity Level: EDI-2 total (r = −0.15 p < 0.01), Body dissatisfaction (r = −0.14 p < 0.01), Ineffectiveness (r = −0.14 p < 0.01), Interpersonal distrust (r = −0.11 p < 0.01), Impulse regulation (r = −0.17 p < 0.01), Social insecurity (r = −0.16 p < 0.01)

Regression- Religiosity and total EDI (p = .043)

Mahoney et al. (2005) (USA) [31]

Dieting Practices Inventory- Unhealthy Dieting Practices subscale

Manifestation of God in the Body scale.

9 IV

Partial Correlation (BMI + gender controlled): Unhealthy dieting practises: Manifestation of God (r = −.12,p < .05), Sacred Qualities (r = −.18, p < .001) and Global Religiousness (r = −.13, p < .05). Binge eating: Sacred Qualities (r = −.14, p < .01).

Binge Eating Scale

Sacred Qualities of the Body scale.

Global Religiousness Score

Pinhas et al. (2008) (Canada) [42]

EAT 26

Self Reported level of observance

9 IV

No sig differences.

Watkins et al. (2006) (USA) [32]

QEWP-R

SWBS- RWB + EWB

11 IV

ANOVA- SWBS: None binge- 96.35, Objective binge- 91.78, Binge trait- 86.11 (p ≤ 0.000) (BT < NB, BT < OB) RWB: None binge- 46.34, Objective binge- 43.63, Binge trait- 40.93 (p ≤ 0.000) (BT < NB) EWB: None binge- 50.09, Objective binge- 48.15, Binge trait- 45.18 (p ≤ 0.000) (BT < NB, BT < OB).

Weinberger-Litman et al. (2008) (USA) [44]

EAT-26

Self-reported level of religious observance

10 IV

ANOVA- Religious orientation: EAT total (F = 5.48, p = .001), Dieting (F = 3.78, p = .01), Bulimia (F = 4.49, p = .004), Oral Control (F = 4.45, p = .004). Generally lowest symptom severity to highest: Intrinsic, Anti-Religious, Pro-Religious and Extrinsic.

ROS

ANCOVA (control age, BMI, religious observance, anxiety and depression)- all remain sig except oral control.

SWB

Boisvert et al. (2013) (Canada) [38]

OBC- 2 items from Body Shame subscale

Religiosity- ‘I would describe myself as religious’ (King et al., 2001)

8 IV

Correlation- Spirituality and Body Shame (r = −.21, p < .01).

Spirituality- SWB- 2 items from EWB subscale

ANCOVA- Spirituality and Body Shame (F(1, 502) = 21.36, p < .001).

Boisvert et al. (2012) (Canada) [37]

OBC- 2 items from Body Shame subscale

Religiosity- ‘I would describe myself as religious’. (King et al., 2001)

8 IV

Correlation- Spirituality and Body shame (r = −.16, p < .01).

Spirituality- SWB- 2 items from EWB subscale

Path analysis- 10 % of the variance in EDI was explained by BMI (b = .28, p < .001) and Body Shame (b = .15, p < .001). 6 % of the variance in Body Shame was explained by age (older men had higher EDI) (b = .15, p < .001), being Asian (b = .11, p < .01) and lower spirituality (b = −.15, p < .001). Spirituality had a sig indirect effect on EDI mediated by Body Shame (b = −.02).

Boyatzis et al. (2007) (USA) [25]

BE Scale

Demographic Survey- Self reported importance of religion

10 IIb

ANOVA- BE-Appearance: (F(2, 122) = 5.42, p = 0.006). Religious group (M = 1.26) increased more than Control group (M = −0.98) (p = 0.005).

BE- Appearance and BE- weight subscales

BE- Weight: no sig differences.

ANCOVA- BMI and pre-test scores as covariates. BE- Appearance: F(2,114) = 4.20, p < 0.02 (pre-test scores), F(2, 114) = 5.53, p < 0.01 (BMI).

BE- weight: no sig differences.

Boyatzis et al. (2006) (USA) [35]

EDI – BD

Quest Scale

10 IV

Correlation-College freshmen + sophomores- Body Dissatisfaction: Total Quest scores (r = .34, p < .01.), Existential Questioning (r = .31, p < .05), Doubting as Positive (r = .28, p < .05).

BE scale

College juniors + seniors- Body Esteem and Openness to Change (r = −.31, p < .05)

College graduates- Body dissatisfaction and Existential Questioning (r = −.25, p < .10).

Gates et al. (2009) (USA) [36]

EDI- BD

MQOS- Religious angst subscale

8 IV

ANOVA- Religious Angst: Higher levels of body dissatisfaction (F (33, 234) = 1.38, p = 0.09).

Gluck et al. (2002) (USA) [39]

BSQ

Orthodox Traditions Scale

9 IV

ANOVA- Secular women had higher body dissatisfaction then orthodox women (F[1, 125] = 8.0, p = .005).

FRS

RIQ

Logistic Regression- Secular students 2x likely to have a fear of becoming fat (odds ratio [Exp (B)] = 2.3, p = .05) + 4x likely to be greatly influenced by their shape and weight (odds ratio [Exp (B)] = 3.8, p = .001).

MLR- Religious grouping: 6 % total BSQ (R² = .06; p < .03).

Hayman et al. (2007) (USA) [46]

OBC

FMS- 20 items

10 IV

Correlation- Whole sample: No sig. Women: No sig, Men: spirituality and body surveillance (r = −.41, p = .001).

VAS

CDRS

Homan et al. (2013) (USA) [45]

BAOS

AGI

10 IV

Correlation- AG- Anxiety: Body Appreciation (r = −.28, p < .01), Body Acceptance by others (r = −.22, p < .05), Body Surveillance (r = −.33, p < .01).

BAS

1 item ‘How religious are you’

AG- Avoidance: Body Surveillance (r = −.21, p < .05).

OBC- Body Surveillance

1 item frequency of worship attendance

Regression- AG- Anxiety: predicted Body Appreciation (B = −.22, p = .038) and Body Surveillance (B = −.29, p = .008).

AG-Avoidance: no sig

Homan et al. (2010) (USA) [28]

Perceived Sociocultural Pressure Scale (Stice, n.d.)

AGI

9 IIb

Correlation- AGI (Anxiety): Pressure to be thin (r = .22, p < .01), Thin-ideal Internalisation (r = .30, p < .001), Body Dissatisfaction (r = .25, p < .001), Dieting (r = .26, p < .001). AGI (Avoidance): Body Dissatisfaction (r = .18, p < .001).

SATAQ- Internalisation subscale

Regression- Thin-ideal Internalisation predicted body dissatisfaction in secure and anxious relationship groups (t(222) = 2.31, p < .02, R² = .08 vs. R² = .27). Secure relationship buffers this relationship.

MBSRQ- Body Areas Satisfaction subscale

Hierarchical regression- Anxious group- Body Dissatisfaction predicted by Pressure to be Thin (R² = .03, p < .05), Thin Ideal Internalisation (R² = .03, p < .05). Secure group- not sig.

DRES

Inman et al. (2014) (USA) [26]

BE

Demographic Survey- Self reported importance of religion

10 IIb

ANOVA- Feelings of love and acceptance between the four statement groups (F(3, 106) = 8.22, p < .001). Religious (Mnonbody = 7.59, Mbody = 7.41) greater then Non-Religious (Mcontrol = 5.22, Mpositive = 6.73) (t(106) = 4.08, p < .001).

EDI- DT, BD

GCS

2 way ANCOVA- Appearance Esteem between the different statement groups not sig (F(3, 101) = .29, p = .83). Affirmation group x religious commitment interaction (F(3, 101) = 2.99, p < .04, η2 = .08).

RDS

RCS

Weight esteem between the different statement groups not sig (F(3, 101) = .52, p = .60), Affirmation group x religious commitment interaction (F(3, 101) = 3.42, p < .03, η2 = .09).

Emotions

Correlation- Religious Commitment: Appearance Esteem (r = .20, p < .05), Weight Esteem (r = .19, p < .05), Body Dissatisfaction (r = −.21, p < .05).

Partial Correlation (control for general commitment)- Religious commitment: Body Dissatisfaction (r = −.20, p < .05).

Inman (2014) (USA) [27]

BE

RCS

10 IIb

ANOVA- no sig differences between groups.

Emotions

2 way ANOVA- religiously committed men had higher Appearance Esteem at time 1 then less religious men (F(1, 48) = 17.76, p = .001, η2 = .27.). No sig interactions.

Correlation- Religious commitment: higher Appearance (r = .34, p < .05) and Weight Esteem (r = .35, p < .01).

Hierarchical regression and simple slope analysis: Religious commitment moderated the effect of harmful media on Weight Esteem in heavy men.

Jacobs-Pilipski et al. (2005) (USA) [33]

EDI-2- BD

Brief COPE

9 IV

ANOVA- No significant relationships between S/R beliefs/practices and Body Dissatisfaction.

13-item self-report measure of spiritual and religious beliefs and practices.

T- Test- Women with strong S/R beliefs/practices more likely to read religious works (T = 7.3, p = .000), pray (T = 9.3, p = .000) and meditate (T = 3.8 p = .007) than women without strong S/R beliefs/practices.

Women with strong S/R beliefs/practices less likely to use distraction to cope with body image distress (T = 2.7, p = .007) and reported prayer to be an effective strategy for dealing with dissatisfaction with weight and shape (T = 3.0, p = .004).

Kim (2006) (USA) [20]

EDI- BD

Compilation of measures- religious practice, application, commitment, identity, coping, social support

11 IV

Regression (controlling BMI + demographics).

Men- Body satisfaction: positive spiritual coping (beta = 0.18, p < 0.05), negative spiritual coping (beta = −0.58, p < 0.01). Closer relationship with God (beta = 0.14, p < 0.05). All religious variables (R² = 0.174, p < 0.05).

Women- Body satisfaction: positive spiritual coping (beta = 0.19, p < 0.01), negative spiritual coping (beta = −0.47, p < 0.01), spending more hours on religious and spiritual activities (beta = 0.01, p < 0.01), prayer (beta = 0.23, p < 0.05). All religious variables (R² = 0.106, p < 0.01).

Mahoney et al. (2005) (USA) [31]

MBSRQ- Body Areas Satisfaction and Appearance Orientation subscales

Manifestation of God in the Body scale.

9 IV

Partial Correlation (BMI + gender controlled)- Body area satisfaction: Manifestation of God (r = .13,p < .05), Sacred Qualities (r = .25, p < .001) and Global Religiousness (r = .13, p < .05).

Sacred Qualities of the Body scale.

Hierarchical regression- Body area satisfaction- Step 1: Race, gender, global religiousness (R² Change = .05, p < .01), Step 2: Manifestation of God, Sacred Qualities (R² Change = .04, p < .01).

Global Religiousness Score

Weinberger-Litman et al. (2008) (USA) [44]

BSQ

Self-reported level of religious observance

10 IV

ANOVA- Religious Orientation: Body Satisfaction (F = 6.15, p < .001). Intrinsic > extrinsic (p = .01), pro-religious (p < .001). ANCOVA- (control age, BMI, religious observance, anxiety and depression)- all remain sig

ROS

ANOVA- Spiritual Well-Being: High > Moderate (F = 4.57, p = .03, d = .27) Body Satisfaction. Existential Well-Being: High > Moderate (F = 18.23, p < .001, d = .51) Body Satisfaction. ANCOVA- (control age, BMI, religious observance)- all remain sig, (control anxiety and depression)- loose sig.

SWB