Table 5 Summary table of the relationship between cognitions related to weight gain and excess gestational weight gain in systematic review of psychological antecedents of excess gestational weight gain

DiPietro 2003 [40]

Pregnancy and Weight Gain Attitude Scale, validated

Excess GWG

Proportions within GWG categories (p-value):

NA

Pre-pregnancy BMI

Only 1 item and two sub-scales were significant on univariate analyses ➔

Individual items:

-Embarrassed about weight

28%, 8% ( p <0.05)

-Worried will get fat

43%, 37% (p-value NS)

Feel unattractive

28%, 14% (p-value NS)

-Embarrassed when nurse weight me

21%, 21% (p-value NS)

-Cannot wear what is in style

18%, 27% (p-value NS)

Subscales:

Negative pregnancy body image r = 0.28 (p < 0.001)

Pregnancy experience scale r = 0.20 (p < 0.001)

McDonald 2013 [61]

Pregnancy and Weight Gain Attitude Scale, validated (Attitude towards weight gain scale)

Excess GWG

Mean (SD) in those gaining above 17.4 (3.4) vs within 17.9 (2.8); OR 0.95 (0.86 to 1.05)

NA

NA

NS on univariate therefore not included in multivariate

Olson 2003 [49]

Pregnancy and Weight Gain Attitude Scale, validated

Excess GWG ^$$

Effect estimate not reported; (p-value NS)

NA

NA

NS on univariate analysis

(^$$$modified 1990 Institute of Medicine guidelines)

Variable not entered in the multivariate model

Sangi-haghpeykar 2013 [51]

Pregnancy and Weight Gain Attitude Scale, validated

Excess GWG ^$

Proportions within GWG categories (p-value):

OR (95% CI )

Pre-pregnancy BMI, USA born, unmarried

Only a few items were significant on univariate or multivariate analyses ➔

Individual items

-Worried will get fat: 28%, 15% ( p <0.05)

-Embarrassed when nurse weighed me: 4.61 (1.18 to 29.80)

-Embarrassed when nurse weighed me: 14%, 3% ( p <0.05)

-Don’t care how much I gain: 3.80 (1.47 to 11.36)

-Don’t care how much I gain: 23%, 9% (p <0.05)

Stevens-Simon 1993 [53]

Pregnancy and Weight Gain Attitude Scale, validated

Rate of weight gain categorised into slow (<0.23 kg/wk), average (0.23 – 0.4 kg/wk), rapid (>0.4 kg/wk)

Correlation co-efficient (p-value):

NA

NA

Only a few items were significant on univariate analyses ➔

Total scale score

Multivariate analysis was not done

r = 0.12 (p <0.14)

Mean (± SD) attitude score among three outcome categories

3.4(±0.6), 3.5(±0.5), 3.5(±0.6) (p >0.05)

Individual items (Correlation co-efficient not reported):

-Liked wearing maternity clothes: (p <0.05)

-Felt unattractive: (p <0.05)

-Embarrassed when nurse weighed me: (p <0.05)

-Cannot wear what is in style: (p <0.05)

Strychar 2000 [54]

Investigator developed, Not validated

Excess GWG

NR

Sub-scale – less favourable attitude towards weight gain led to excess weight gain

Pre-pregnancy BMI, age, marital status, education, smoking, and alcohol

Only a sub-scale was significant on multivariate analysis ➔

Effect estimate not reported (p <0.05

Strychar 2000 [54]

Investigator developed, Not validated

Excess GWG

NR

Sub-scale: Perceived concern about their weight – more concerned leads to excess weight gain Effect estimate not reported; (p <0.05)

Pre-pregnancy BMI, age, marital status, education, smoking, and alcohol

Only a sub-scale, namely, ‘perceived concern’ was significant on multivariate analysis ➔

Strychar 2000 [54]

Investigator developed, Not validated

Excess GWG

NR

Sub-scale: Importance of not gaining an excess amount of weight– Less knowledge leads to excess weight gain

Pre-pregnancy BMI, age, marital status, education, smoking, and alcohol

Only a sub-scale, namely, ‘ importance of not gaining an excess amount of weight’ was significant on multivariate analysis ➔

Effect estimate not reported; (p <0.05)

Cogswell, 1999 [37]

Investigator developed single item; Not validated

Excess GWG

NR

OR (95% CI )

Pre-pregnancy BMI, maternal height, age, race, education, marital status, parity, prenatal care, WIC participants,, income

Significant on multivariate analysis

Target weight gain categories

➔ (> recommended)

<Recommended 0.4 (0.2 to 0.6)

(< recommended)

Recommended 1.0 (Reference)

>Recommended 6.1 (4.1 to 8.9)

McDonald 2013 [61]

Investigator developed single item; Investigator developed, not validated

Excess GWG

OR (95% CI )

OR (95% CI )

Pre-pregnancy BMI group, first birth, planned

Planned gain above the guidelines Significant on both univariate and multivariate analysis

Planned gain above the guidelines 9.31 (3.86 to 22.42), planned gain below 0.78 (0.33 to 1.84)

Planned gain above the guidelines 11.18 (4.45 to 28.06); planned gain below 0.69 (0.26 to 1.80)

weight gain, daily soda or juice consumption, watching television before bedtime, locus of control to Eysenck’s neurotic scale of emotional instability, and satisfaction with pre-pregnancy weight

➔ (> recommended) planned gain below NS on univariate or multivariate multivariate

Herring 2008 [41]

Previously published single item adopted National Health and Nutrition Examination Survey, No reference to validation

Excess GWG

Proportion of Excess GWG within each exposure category:

OR (95% CI ):

Pre-pregnancy BMI, age, education, marital status, income, employment, ethnicity, parity, smoking, gestational length

Significant on univariate or multivariate analyses ➔

Normal weight, accurate assessor 47%

Normal weight, accurate assessor 1.0 (reference);

Normal weight, over-assessor 57%

Normal weight, over-assessor 2.0 (1.3 to 3.0);

Overweight, accurate assessor 62%

Overweight accurate assessor 2.9 (2.2 to 3.9);

Overweight under-assessor 81% (p <0.05)

Overweight under-assessor 7.6 (3.4 to 17.0)

Mehta-Lee 2013 [63]

Single item, Perceived weight status was defined as “accurate” or “inaccurate” based upon the level of concordance between BMI (derived from actual weight) and self reported overweight or obesity (no reference to validation)

Excess GWG

OR (95% CI ): Inaccurate reporters 1.2 (0.8, 1.8);

OR (95% CI ): Inaccurate reporters 1.1 (0.7, 1.7);

Stratified by BMI; adjusted for: WIC status, employment status, race, native born, smoking, parity and either pre-gestational or gestational diabetes

NS on univariate and on multivariate analyses

Bagheri 2013 [35]

Body Image Assessment for Obesity (BIA-O), Validated

Excess (cases) vs. Adequate (controls) GWG

OR (95% CI ):

OR (95% CI ):

Pre-pregnancy BMI, age, parity, social class, energy intake

Significant on univariate or multivariate analyses ➔

Heavier body size preference 0.54 (0.27 to 1.04)

Heavier body size preference 0.44 (0.18 to 1.10)

Thinner Body Size Preference 2.17 (1.17 – 4.02)

Thinner body size preference 3.12 (1.97 to 4.95)

Hill 2013 [42]

Body Attitude Questionnaire (BAQ), Validated, modified

Excess GWG ^$

NR

Effect estimates were not reported; p-value NS

Pre-pregnancy BMI, age, parity, education level

NS on multivariate analysis

Mehta 2011 [46]

Body Image Assessment for Obesity (BIA-O), Validated

Excess GWG

RR (95% CI ):

RR (95% CI ):

Pre-pregnancy BMI

Significant on multivariate analysis ➔

Heavier body size preference 1.79 (0.52-9.58)

Thinner body size preference

Thinner body size preference 0.88 (0.82 to 0.94)

<16 years of education 1.11 (1.00 to 1.22)

≥16 years of education 0.92 (0.83 to 1.01)

McDonald 2013 [61]

Satisfaction with pre-pregnancy weight , not stated if validated or not

Excess GWG

OR (95% CI ):

NA

NA

Significant on univariate analysis

Not or not at all satisfied vs. satisfied or very satisfied 0.25 (0.10 to 0.60)

NS on multivariate analysis

McDonald 2013 [61]

Locus of control score, validated

Excess GWG

OR (95% CI ) 1.12 (1 to 1.26)

NA

NA

NS on univariate analysis; Variable not entered in the multivariate model

Olson 2003 [49]

Weight Locus of Control (WLOC), Validated

Excess GWG ^$$

Effect estimate not reported; p-value NS

NA

NA

NS on univariate analysis; Variable not entered in the multivariate model

Wright 2013 [59]

Single item from Attitude towards weigh gain scale by Palmer, Validated, modified

Excess GWG;

Effect estimate not reported for Adequacy ratio

Effect estimate not reported for Excess GWG

Pre-pregnancy BMI, age, race

Results were reported to be similar to secondary outcome , hence considered significant on univariate or multivariate analysis

GWG (continuous)^$$

β (95% CI ) for secondary outcome:-11.6 (−21.4 to −1.9)

β (95% CI ) for secondary outcome: −16.1 (−28.7 to −3.4)