The relationship between caregivers and daughters ’ food and body shape scripts: A dyadic analysis

Research indicates a key role for parenting in a daughter ’ s relationship with food and body shape. One possible mechanism for this translational process is through scripts. The present study used a dyadic design to investigate the relationship between caregivers and daughters ’ scripts regarding food and body shape. Caregivers (n = 40) and their daughters (n = 40) rated word lists of their current and childhood scripts and caregivers completed an additional measure of their parenting scripts. Non-dyadic analysis showed consistent correlations between daughters ’ current scripts and their recollections of childhood scripts, and several (but less consistent) correlations between caregivers ’ current scripts, their own childhood scripts and the scripts they had used as a parent. Dyadic analysis showed consistent correlations between the daughter ’ s current scripts relating to negative eating and both positive and negative body scripts and their caregiver ’ s parenting scripts, and between the daughter ’ s current scripts relating to positive and negative eating and body looking positive and their caregiver ’ s current scripts. No associations were found between the daughter ’ s current scripts and the caregiver ’ s childhood scripts. The results indicate that whilst daughters ’ current scripts relating to food and body shape are often concordant to those reported by their caregivers, they are more closely linked to what they remember from their childhoods. Further, they suggest that those scripts which do transfer between generations may be more related to body size and negative eating than food per se. Finally, the results suggest that scripts do not inevitably pass across the generations possibly due to parents choosing not to repeat the errors of their own parents or due to the role of factors other than just parenting in creating the scripts we hold.


Introduction
Over the past few decades research has consistently shown an increase in problems relating to body shape and eating behaviour including dieting, eating disorders and obesity which are apparent in both adults and children (eg.Bucchianeri, Arikian, Hannan, Eisenberg, & Neumark-Sztainer, 2013;Wardle & Johnson, 2002;Ogden, 2010).Core to many of these problems is weight concern and underlying issues relating to body image and food (Ogden, 2010(Ogden, , 2018;;Posavac & Posavac, 2002;Sweeting & West, 2002;Tiggemann, 2005;Wardle & Johnson, 2002).Whilst the causes of these problems are clearly multifactorial involving a combination of biological, social and psychological factors, one key influence is the family, particularly parents which has been explored in terms of parental modelling, food parenting, general parenting and the language parents use.Some research has explored the degree of concordance between parents and their children with a focus on the role of modelling for both food and body related issues.For example, in terms of food, concordance between parents and children has been shown for attitudes towards fruit (Guidetti et al., 2012), fat intake (Feunekes, Stafleu, De Graaf, & Van Staveren, 1997), consumption of energy dense foods (Prichard, Hodder, Hutchinson, & Wilson, 2012), snack food intake (Brown et al., 2008), and eating breakfast (Pearson, Biddle, & Gorely, 2009) and Arcan et al. (2007) reported concordance between parent and child vegetable and dairy intake both whilst the child was still living at home and after 5 years at follow up.In terms of body issues, research also shows concordance between mothers and daughters for emotional eating (Snoek, Engels, Janssens, & van Strien, 2007), body dissatisfaction (Brown & Ogden, 2004;Hall & Brown, 1982), restrained eating (Steiger, Stotland, Trottier, & Ghadirian, 1996;Hill, Weaver, and Blundell, 1991) and fasting behaviour (Hill & Franklin, 1998).Further, Dickens and Ogden (2014) reported concordance between a parent and their child's intake of unhealthy snacks and emotional eating one year after the child had left home.Such research therefore indicates a key role for parents and highlights how parental behaviour and attitudes may be modelled by their child.Some research has also emphasised a key role for aspects of food parenting, particularly the use of exposure, reward and aspects of parental control (Ogden, 2014).For example, simply having food available at home can increase a child's liking of that food (Birch & Marlin, 1982;Cullen et al., 2003;Mennella & Beauchamp, 2002;Wardle, Herrera, Cooke, & Gibson, 2003) and parental reward of eating behaviour can change food preferences and intake (Savage, Fisher, & Birch, 2007;Hendy et al., 2007).Research exploring parental control, however, is more complex.For example, some research indicates that parental restriction of food may lead to weight gain, preoccupation for the restricted food, dieting and overeating (Birch & Fisher, 1998, 1998, 1998;Carper, Fisher, & Birch, 2000;Fisher & Birch, 1999;Ogden et al., 2013).In contrast, other studies show some benefits to parental control.For example, Brown and Ogden (2004) showed an association between greater parental control and a higher intake of healthy snacks by the child and Ogden and colleagues concluded that whereas overt control was associated with an increased intake of healthy snacks, covert control was linked to a decrease in unhealthy snacks (Brown, Ogden, Vogele, & Gibson, 2008;Jarman et al., 2015;Ogden, Reynolds, & Smith, 2006).
Parents, however, not only manage their children's diets but also parent in a more general way which can also influence a child's relationship with food and their body shape.For example, authoritarian parenting may be associated with a more controlling approach to food, permissive and 'laissez faire' parenting has been linked with 'nutritional neglect', and authoritative parenting has been associated with positive feeding behaviours, healthy eating, active lifestyles and a lower BMI (Hughes, Power, Fisher, Mueller, & Nicklas, 2005;Johnson, Welk, Saint-Maurice, & Ihmels, 2012;Sleddens, Gerards, Thijs, de Vries, & Kremers, 2011;Vollmer & Mobley, 2013).Further, Ogden and Steward (2000) explored the dynamic between mothers and daughters using a dyadic analysis and concluded that higher restrained eating and body dissatisfaction by daughters was related to mothers reporting a lower belief in their own autonomy and both mothers and daughters rating projection as important in their relationship.
Finally, some research has specifically focused on the role of language and the ways in which parents speak about issues relating to food and body shape.Some research has used observational data to provide a degree of objectivity to the analysis of language.For example, Bova and Arcidiacono (2014) gathered 30 video recordings of meal-times of 10 families and explored the strategies used by parents in food-related discussions.The results showed that parents focused on the quality and quantity of food.Quality strategies were either positive, describing food as nutritious and tasty or negative such as too salty and used 43% of the time, whereas quantity-based strategies (too much or too little food) were used 35%.Likewise, Roach et al. ( 2017) used a recording device worn by the child from the time they arrived home from school to bedtime, for 3 days to explore how food was talked about outside of mealtimes.The results showed that food talk made up between 0% and 66% of total speech during this period and that there was a positive correlation between maternal and child food talk.For example, when mothers talked about their enjoyment of and desire for food their children were more likely to do the same.
Other research has used retrospective designs with children being asked about the language used by their parents.For example, Gross and Nelson (2000) reported that women with higher eating disturbances were more likely to reflect that their mothers had used negative messages such as 'you need to lose weight' rather than positive messages such as 'your health is what is important, not your weight'.Likewise, Neumark-Sztainer et al. (2010) reported correlations between adolescent girls' current disordered eating including unhealthy and extreme weight control behaviours and binge eating and their perceptions of parent weight talk, family weight teasing and mother dieting.
In contrast, some research has used collected data independently from both parties.For example, Bauer, Bucchianeri, and Neumark-Sztainer (2013) used a dyadic analysis of mothers and daughters and reported that more weight related comments from mothers to their daughters was associated with daughters having higher depressive symptomatology, greater weight control behaviours and greater prevalence of binge eating.Furthermore, higher levels of mother's weight talk about their own weight or shape was also related to their daughter's lower self-worth and higher depression.Likewise, Cooley, Toray, Wang, and Valdez (2008) reported that the mother's own reported negative feedback about their daughter's body shape and eating behaviour was related to their daughter's own reported weight concern.However, the best predictors were the daughter's perception of these factors rather the mother's own recollection.Similarly, Berge et al. (2013) used matched data between mothers, fathers and their adolescent son or daughter to explore the impact of weight-related versus healthful eating conversations.The results showed that a higher level of weight-related conversations predicted adolescents who were more likely to diet, use unhealthy weight-control behaviours, and engage in binge eating.Such research indicates the negative impact of a parent's comments on their child.In contrast, some research also shows a more positive relationship between parents and their children.For example, Berge et al. (2013) also reported that those adolescents who were either overweight or obese, and whose mothers engaged more in healthful eating conversations, were less likely to diet and use unhealthy weight-control behaviours.Furthermore, this beneficial effect was increased when both parents engaged in more healthful rather than weight related conversations.In a similar vein, Handford, Rapee, and Fardouly ( 2018) carried out an experimental study in which mothers and daughters viewed thin-ideal magazine advertisements and mothers made either negative comments about their own weight, shape or diet or made no appearance or diet related-comments.The results showed that the daughters who had heard their mothers make self-critical comments reported lower body esteem and body satisfaction, and greater problematic attitudes and restrained eating habits at the end of the study.
Previous research has therefore highlighted the role of parental influence on their children's relationship with food and body shape and has identified a role for modelling, both food and general parenting and the language they use.Further, research has used both retrospective and cross-sectional designs and collected both self-reported and dyadic data.In general, research in this area illustrates a strong association between parents and their children for both food and body shape related issues.To date, however, less is known about how the transmission from parent to child occurs and how consistent this transmission process is through the generations.One possible explanation lies in the notion of scripts.Early research in cognitive and developmental psychology identified the role of cognitive representations or schema to reflect internal dialogues resulting from childhood interactions which have also been called scripts, life scripts or family scripts (Bergen & Fromberg, 2009;Byng-Hall, 1985, 1998;Fiske & Taylor, 1991;Markus, 1977;Nelson, 1981;Steiner, 1990).These scripts reflect an underlying cognitive structure which underpin how people make decisions as they navigate their way through the world and can form the focus of psychological interventions such as transactional analysis, cognitive behavioural therapy or family therapy in the form of rescripting (Byng-Hall, 1985;Steiner, 1990;Wheatley et al., 2007).The notion of scripts has also been used in the context of eating behaviour to explore how rural women construct food choices (Blake & Bisogni, 2003), how people hold individualised food systems (Connors, Bisogni, Sobal, & Devine, 2001), how food is allocated in families and managed within couples (Charles & Kerr, 1988;Bove & Sobal, 2006;Blake et al., 2008) and how individuals construct their evening meal (Blake, Bisogni, Sobal, Jastran, & Devine, 2008).It also finds reflection in research which has examined the extent and impact of fat and body talk in the media and within families and the ways in which the words we hear can be internalised as part of our cognitive schema (Herbozo, Tantleff-Dunn, Gokee-Larose, & Thompson, 2004;Sharpe, Naumann, Treasure, & Schmidt, 2013;Nichter & Vuckovic, 1994, pp.109-131; Salk & Engel-Maddox, 2011; Vanderkruik, Conte, & Dimidjian, 2020).Such scripts may therefore offer a mechanism to explain how parents transmit their own issues surrounding food and body shape to their children.In line with this, the present study used non-dyadic and dyadic data to explore food and body shape scripts for both daughters and their key caregivers.First, the study aimed to explore the degree of concordance within daughters for their current food and body shape scripts compared to those they remember being exposed to by their parents as children.Second, the study aimed to explore the degree of concordance within caregivers for their current scripts, those they heard from their parents as a child and the scripts they used when parenting their own child.Third, using a dyadic design the study aimed to explore the degree of concordance between daughters and their caregivers' scripts.This approach enabled a comparison between scripts, an exploration of the relative impact of remembered scripts versus actual scripts and an assessment of how scripts may transfer across the generations.

Design
The study used a dyadic design to assess food and body shape scripts within pairs of daughters and their caregivers.Daughters rated their own current scripts and childhood scripts and caregivers rated their own current scripts, childhood scripts and parenting scripts.Associations between scripts were assessed within the daughter (current and childhood), within the caregiver (current, childhood and parenting), and within the caregiver/daughter dyad (current and current; current and parenting; current and childhood).Favourable ethical approval was given by the University Ethics Committee (FHMS 21-22 215).

Participants
The sample consisted of 40 pairs of caregivers and daughters (40 daughters, 39 mothers and 1 father).The daughters had a mean age of 22.9 years (SD = 3.4, ranging from 18 to 39 age), and parental caregivers had a mean age of 52.8 years (SD = 5.7, ranging from 40 to 63 years).Participants were recruited through social media platforms (Instagram and Facebook) and spread via word of mouth.Participants were excluded from the study if they were below the age of 18 or did not speak English.Based upon previous research a sample size of 40 dyads was deemed reasonable to explore within subjects correlations (Ogden & Taylor, 2000;Ogden, Cheung & Hudson, 2022).

Demographics
Participants described their age, gender, ethnicity, weight, height and whether they were a daughter or a caregiver.

Food and body scripts
2.3.2.1.Developing the food and body shape scripts.A pilot study was used to create the word list to assess the food and body shape scripts.Based upon previous research nine categories of script were generated to reflect types of food and body shape scripts: food positive, food negative; food treat; eating positive; eating negative; body looking positive; body looking negative; body doing positive; body doing negative.Nine participants were then given the 9 categories and asked to generate 5 words suited to each category which were then collated by the researchers.Next, the same 9 participants were asked to rate the words that were the best fit for each category.These ratings were then reviewed by the researchers and the final word list was formed.This final list consisted of 9 categories with 5 words each, totalling 45 words.Daughters rated their current scripts and childhood scripts and caregivers rated their current scripts, childhood scripts and parenting scripts in response to the questions: 'In your head, how often do you use these words now' (current script); 'Imagine back to when you were a child.How often did your parents use these words?' (childhood script); and 'How often did you use these words towards your daughter when she was growing up?' (parenting script).All words were rated on a 5point Likert scale (1 = Never, 5 = Always).

Procedure
The study took place online via the Qualtrics platform and targeted daughters for their initial response.The aims of the study were framed as 'to investigate the effect of language used by parents towards their daughters as they were growing up, and how this has affected the language they use towards themselves as adults'.After completing the consent form, daughters described their demographics and completed the word rating task.They were then asked to generate an ID code, to give this to their 'main caregiver whilst growing up' along with the link for the caregiver's questionnaire.Once completed matched dyadic data sets were generated via the code.Participants received no renumeration.
Participants were asked to complete their own measures independently from each other but due to the online nature of the study this cannot be guaranteed.

Data analysis
Mean scores were computed for daughters' and caregivers' 9 categories of food and body scripts.The majority of variables met parametric assumptions.Seven variables did not.Pearson's correlational analysis was therefore used for analysis when both variables were parametric and Spearman's correlational analysis was used when at least one variable did not meet parametric assumptions.Analyses assessed the following: i) Within daughters: current scripts and childhood scripts; ii) Within caregivers: current scripts, childhood scripts and parenting scripts; iii) Within dyads: daughters' current scripts with caregivers' current, parenting and childhood scripts.All analysis was conducted using Jamovi (df = 38 throughout).

Participant demographics
Daughters' and caregivers' demographic characteristics are shown in Table 1.
All caregivers were female apart from 1 male, and the majority of both daughters and caregivers were of white ethnicity, followed by Asian, followed by Middle Eastern.The mean BMI for daughters was in the healthy weight range (<25), whilst the mean for caregivers was just in the overweight range (>25).

Within daughter's scripts
The correlation analysis between daughters' current scripts and their childhood scripts are shown in Table 2.
The results showed significant positive correlations between all scripts for both food and body related items indicating a degree of concordance between the daughters' current food and body scripts and how they remember their parents using these words in their childhood.

Within caregivers' scripts
The correlation analysis between caregivers' current scripts, childhood scripts and parenting scripts are shown in Table 3.
The results showed significant correlations between all current and childhood scripts except for body looking negative and body doing positive.The results also showed significant correlations between all current and parenting scripts.Further, the results showed significant correlations between all childhood scripts and parenting scripts except for food negative, eating negative, body looking negative and body doing negative.This suggests a degree of concordance between food and body scripts at these different stages of life.It also suggests that whilst most scripts may traject from childhood to the current day and to the scripts used when parenting, not all do, and that those relating to body negative may develop due to factors other than what was heard in childhood.

Dyadic analysis: caregiver-daughter scripts
The correlation analysis within daughter/caregiver dyads is shown in Table 4.
The results showed significant positive correlations between the daughters' current scripts and the caregivers' parenting scripts for eating negative, body looking positive, body looking negative, body doing positive and body doing negative but not for food positive, food negative, food treat and eating positive.The results also showed significant positive correlations between the daughters' current scripts and the caregivers' current scripts for eating positive, eating negative and body looking positive but not for food positive, food negative, treat, body looking negative, body doing positive and body doing negative.No correlations were found between the daughters' current scripts and the caregivers' childhood scripts.This indicates a degree of concordance between daughters and their caregivers' body and food scripts but that this is stronger in the current day and gets weaker as it trajects back through the generations.

Discussion
The present study aimed to assess the degree of concordance within daughter's scripts and within caregiver's scripts and to use a dyadic design the study aimed to assess the degree of concordance between daughters and their caregivers.
In terms of daughters' own scripts, the results showed consistent significant correlations between daughters' current scripts and their childhood scripts for all food and body related words.Similarly, consistent corelations were also found between most of the caregiver's current scripts and their own recollection of the scripts they were exposed to as a child.The exception for the caregivers, however, was for scripts relating to the body looking negative and the body doing positive.For these aspects of body scripts the correlation was very small.There are several possible reasons for these gaps.First, it could be due to the additional time between current and childhood scripts for this older sample reflecting poorer recollection.Second, it could indicate that over time factors other than parenting have a part to play in the development of body scripts such as friends, peers, partners or the media.Third, this could indicate a cohort effect and that this generation of people were parented differently in the past with conversations about bodies being either less frequent in general or less about how the body looked or what it could do.
In terms of other aspects of the caregivers' scripts, the results showed consistent concordance between all current and parenting scripts indicating that caregivers parent in line with the way they currently think about food and body shape.Further, there were also many correlations between caregivers' recollection of how they were parented and their own parenting scripts specifically for the positive aspects of both food and body size.This was not the case, however, for negative food and body size scripts.This suggests that the positive food and body size scripts that we hear as a child may traject forwards into the ways in which we parent our own children.This may not be the case, however, for the more negative scripts.This may reflect a social desirability bias with caregiver's wishing to present their own parenting in a more positive way than the parenting they received.It may also reflect either an unconscious or conscious decision to parent more positively that they themselves had been parented.Finally, it may indicate that factors other than their own more negative parenting experiences generate the ways in which they parent their own children.From this non dyadic analysis many correlations can therefore be seen between current food and body scripts and recollections of those exposed to when a child.This was less consistent however, for caregivers, particularly for negative scripts, suggesting either a social desirability bias or the impact of time and that as we get older factors, other than what we were exposed to as a child, influence the scripts we have.
To further assess the relationship between caregiver's and daughters scripts the results were also analysed using dyadic data.The results from this analysis indicate no concordance between the daughter's current scripts and the caregiver's childhood scripts, some concordance between the daughter's current scripts and the caregiver's own current scripts and more (but not complete) concordance between the daughter's current scripts and the caregiver's reported parenting scripts.In particular, whilst only eating positive and negative and body looking positive scripts were concordant between the daughter's current and the caregiver's current scripts, all body related scripts were concordant between the daughter's current with the caregiver's parenting scripts.This pattern of findings provides several insights into the impact of parenting on a child's subsequent scripts.First, it indicates an impact of time and that scripts may not be consistent across the generations.Therefore, what a caregiver hears as a child may not inevitably form the basis of what their own child hears from them in future years.Second, it indicates a different impact for what a caregiver thinks and how they parent.In particular, whilst their own scripts may influence their child, it is how these scripts are expressed in parenting that has a greater impact.Finally, it also illustrates the different impact of different scripts.
In particular, from this dyadic analysis it would seem that both positive and negative body scripts together with negative eating scripts are more likely to be recreated as daughter scripts than any scripts relating to food or eating positively.
Previous research indicates a role for parenting on children's relationship with food and body shape and highlights a role for factors such as modelling, food and general parenting and language (eg.Arcan et al., 2007;Dickens and Ogden, 2014;Ogden et al., 2013;Jarman et al., 2015;Vollmer & Mobley, 2013;Neumark-Sztainer et al., 2010).To date, less is known about the mechanism behind this translational process from parent to child.The results from the present study indicate that scripts relating to both food and body shape may offer an explanation of the ways in which how a parent thinks about food and body shape influences their child's own thoughts.In particular, the results suggest that a key caregiver may hold certain scripts which underpin their own behaviour which in turn are encoded by their child in the form of their own scripts.The results from the present study, however, indicate greater consistency between the daughter's current scripts and those she remembers from her childhood rather than the more objective version of events presented by her caregiver.This reflects previous research suggesting a stronger role for remembered parenting rather than the parent's own view of what was said (Cooley et al., 2008).Therefore, whilst they may be a direct impact of parental script on child script, this is sometimes a perceptual process influenced by what the child both hears and retains.Furthermore, the results suggest that the translation of scripts across time and between generations is not inevitable.In particular, the results showed more consistency between scripts in the present day (eg.daughter current and caregiver parenting) than between those scripts from the past (eg caregiver parenting and caregiver's own childhood).This suggests that either memory is problematic as time passes, or that parents can make active choices to parent differently to their own experiences as a child.Furthermore, it suggests that factors other than our parents' scripts make up our own scripts and that as we traject through adulthood we may well develop scripts in response to our peers, partners and the media as well as our parents.Finally, the results also indicate that some forms of script may be more robust across generations than others and that those relating to body shape and the negative aspects of eating may be more resistant to time and memory than those relating to food.
There are some problems with the current study that need to be addressed.First, the recruitment process highlighted the focus on language.This may have resulted in a degree of response bias as those taking part may have been more interested in the role of language than those who did not.Second, due to the online dyadic design it is not possible to know whether parents and their daughters colluded whilst completing the measures which could have resulted in responder bias.In addition, whilst the sample was quite ethnically diverse, the sample size within each cell was too small to explore any differences in dyadic relationships by ethnicity.Further, the study relies on the recall of language used rather than having any measure of in the moment language.This is clearly open to aspects of measurement bias.Future research could try more observational or laboratory-based methods as a means to overcome some of these problems.In addition, research could target dyads which differ by a range of demographics (ethnicity, family structure, country of origin, language spoken etc) as a means to explore the use of scripts within these different types of communication dyads and whether the translation process varies.
In summary, parents are clearly key to the development of a child's relationship with food and body shape.The results from the present study highlight a key role for scripts as the mechanism behind this translational process.In particular, parental scripts may underpin how they behave around their child which in turn generate matched scripts in the minds of their child.The results from the present study, however, indicate that this is not an inevitable process and may be disrupted by a number of factors such as time, memory or an active choice to parent differently.

Funding
None.

Declaration of competing interest
None.
2.3.2.2.Final scripts.The final food and body scripts were as follows:Food scripts: i) Food positive: Food is … healthy fruit and vegetables; good for you; nutritious; balanced; fresh; ii) Food negative: Food is … too many calories; unhealthy; greasy; processed; fattening; iii) Treat: A treat food is … a reward; a special treat; a snack; tasty; sweet; iv) Eating positive: You are … healthy; careful; vigilant; health conscious; you make good decisions v) Eating negative: You are … picky; a bottomless pit; a bit of a pig; greedy; an overeater.