Exploring parental secretive eating of sugary foods and drinks, and its associations with food consumption in families

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ Exploring parental secretive eating of sugary foods and drinks, and its associations with food consumption in families © 2024 the Authors Published version Sarvanne, Tuuli; Kokko, Sami; Abdollahi, Anna M.; Serasinghe, Nithya; Kinnunen, Satu; Lehto, Reetta; Vepsäläinen, Henna Sarvanne, T., Kokko, S., Abdollahi, A. M., Serasinghe, N., Kinnunen, S., Lehto, R., & Vepsäläinen, H. (2024). Exploring parental secretive eating of sugary foods and drinks, and its associations with food consumption in families. Appetite, 200, Article 107578. https://doi.org/10.1016/j.appet.2024.107578 2024 Appetite 200 (2024) 107578 Available online 21 June 2024 0195-6663/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Exploring parental secretive eating of sugary foods and drinks, and its associations with food consumption in families Tuuli Sarvanne a , Sami Kokko a , Anna M. Abdollahi b , Nithya Serasinghe c , Satu Kinnunen b , Reetta Lehto c , Henna Veps¨ al¨ ainen b , * a Faculty of Sport and Health Sciences, University of Jyv¨ askyl¨ a, PO Box 35, 40014, Finland b Department of Food and Nutrition, University of Helsinki, P.O. Box 66, 00014, Finland c Folkh¨ alsan Research Center, Topeliuksenkatu 20, Helsinki, 00250, Finland ARTICLE INFO Keywords: Food parenting practices Covert control Role modeling Sugar-sweetened beverages Snack food Mixed methods ABSTRACT As a way of modeling healthier eating habits for their children, parents may intentionally avoid consuming sugary foods and drinks (SFDs) in their presence but consume these on other occasions (later referred to as parental secretive eating). This study aimed to 1) explore the prevalence of parental secretive eating, 2) investigate the associations between parental secretive eating and SFD consumption in parents and children, and 3) qualitatively explore the reasons for parental secretive eating. Participants were Finnish mothers (n =362), fathers (n =123), and their 3–6-year-old children (n =403); this data was collected in 2017 as part of the baseline assessment of the DAGIS intervention. Parents reported how often they avoided eating SFDs in the presence of their child, completed food frequency questionnaires for themselves and their child, and responded to an open-ended question of explaining reasons for secretive eating. The overall prevalence of parental secretive eating was 68%. It was more common among mothers than fathers (p <0.001) and most prevalent in chocolate (61%) and sweets (59%). Parental secretive eating was positively associated with SFD consumption both among mothers (β =0.274, p <0.001) and fathers (β =0.210, p =0.028) in linear regression models adjusted for parents’ and child’s age, child’s gender, parental education level, and number of household members. Mothers’ or fathers’ secretive eating and child’s SFD consumption were not associated (β =0.031, p =0.562; β =-0.143; p =0.167). Three themes describing reasons for parental secretive eating were found: family food rules, avoiding child’s requests, and aspiration for healthy modeling. In conclusion, parental secretive eating may play an important role in determining SFD consumption in families with preschoolers. Additional research is needed to determine whether parents can prevent their own eating habits from influencing their child through secretive eating. 1. Introduction Parents have an important role in the development of their child’s eating habits (Cooke, 2007; Haines et al., 2019; Scaglioni et al., 2018). When it comes to sugary foods and drinks (SFDs), the parental role may be even more crucial, as preschool children in Finland consume most of their SFDs outside of preschool hours (Korkalo et al., 2019a). Thus, understanding children’s consumption of SFDs requires insight into the home food environment and parental behavior. Parents are responsible for the physical home food environment including the availability and accessibility of foods (Bekelman et al., 2017; Rosenkranz & Dzewaltowski, 2008). Unsurprisingly, the availability of unhealthy foods (e.g., SFDs) is positively associated with children’s consumption of these foods (Bassul et al., 2020; Blaine et al., 2017; Boles et al., 2019; Paasio et al., 2022; Veps¨ al¨ ainen et al., 2018a; Yee et al., 2017). Moreover, parents affect the social home food environment with their food perceptions, eating habits, parenting style, and specific food parenting practices (e.g., rules, and modeling) (Rosenkranz & Dzewaltowski, 2008). Parental modeling has a clear role in shaping children’s eating habits as there is an association between healthy modeling and children’s healthier eating habits (Mazarello Paes et al., 2015; Yee et al., 2017) and unhealthy modeling and children’s less healthy eating habits (Brown & Ogden, 2004; Dickens & Ogden, 2014; Yee et al., 2017). Parents are * Corresponding author. University of Helsinki, Agnes Sj¨ obergin katu 2, P.O. Box 66, 00014, Finland E-mail address: [email protected] (H. Veps¨ al¨ ainen). Contents lists available at ScienceDirect Appetite journal homepage: www.elsevier.com/locate/appet https://doi.org/10.1016/j.appet.2024.107578 Received 8 March 2024; Received in revised form 17 June 2024; Accepted 20 June 2024 Appetite 200 (2024) 107578 2 aware of their role as models (Cook et al., 2021; Ray et al., 2016; Versele et al., 2021) and try to avoid unhealthy modeling of SFD consumption for their child (Cook et al., 2021; Lindsay et al., 2018; Versele et al., 2021). Intentional avoidance of unhealthy modeling, indicated by statements “If I would like to eat/drink sweets/soft drinks, I would restrain myself because of the presence of my child”, was negatively associated with child’s soft drink consumption in a Belgian cross-sectional study (Vereecken et al., 2004). However, in a longitudinal study, intentional avoidance of unhealthy modeling was not associated with children’s consumption of sweets, soft drinks and crisps in multivariate regression models (Vereecken et al., 2010), although the results may have been different if this study focused on SFDs rather than combining salty foods and SFDs. Studies with preschool children found that mothers may unintentionally model intake of less healthy foods even if they try to be healthy models (Palfreyman et al., 2013, 2014). Unintentional modeling referred to instances where the mother had not used verbal or direct behavioral modeling. Even if parents intentionally avoid modeling SFD consumption, they remain continuous models for their child (Rhee, 2008). A qualitative study found that Danish parents consumed SFDs as a reward for themselves after a stressful day when their child was asleep (Moura & Aschemann Witzel, 2020). The perception of SFDs as a reward may be transmitted from parents to children through unintentional modeling, fostering a temptation toward these foods in children. Though it may be common that there is a contradiction between parents’ eating habits and their aim to be healthy models, little research has studied the intentional avoidance of modeling SFD consumption for children and the reasons behind it. Hence, it remains unclear whether intentional avoidance of unhealthy modeling is beneficial for children’s eating habits or not. In the present study, we refer to the behavior in which parents intentionally avoid SFD consumption in the presence of their child but consume these foods on other occasions, as parental secretive eating. Parental secretive eating differs from intentional avoidance of unhealthy modeling in that parental secretive eating requires parental SFD consumption, whereas intentional avoidance of unhealthy modeling does not. This study had three aims. First, we explored the prevalence of parental secretive eating among parents of 3-6-years-olds in Finland. Mothers and fathers were considered separately to gain knowledge about possible differences in parental secretive eating, as differences between mothers’ and fathers’ food parenting practices (Khandpur et al., 2014) and Finnish women’s and men’s SFD consumption have been detected (Valsta et al., 2018). We were also interested in examining if secretive eating in mothers and fathers have different impacts on children’s food consumption. Research on fathers’ food parenting is still relatively limited (Jansen et al., 2018), even though father’s caretaking responsibilities (Jones & Mosher, 2013) and involvement in child feeding have raised (Rahil et al., 2020). Next, we investigated the associations between parental secretive eating and SFD consumption frequency in parents and children to understand if parent’s or child’s SFD related eating habits differed by parental secretive eating. Finally, we performed a qualitative exploration of underlying reasons for parental secretive eating to gain insight into the motivations driving this behavior. Understanding the motivations involved enables more accurate explanations of the behavior and supports theory development in future studies. As this was an explorative study there was no hypothesis for any of these three aims. To the best of our knowledge, no previous studies have specifically examined parental secretive eating; however, avoiding unhealthy modeling has been investigated as a component of food parenting practices in a limited number of studies (Gevers et al., 2015; Vereecken et al., 2004, 2010). Increasing our understanding of parental secretive eating enables the broadening of knowledge about SFD-related eating habits in families with preschool-aged children. 2. Methods 2.1. Study design and participants This study reports secondary analyses from the baseline data of the DAGIS (Increased Health and Wellbeing in Preschools) intervention, a randomized controlled trial conducted in 2017–2018 (Ray et al., 2020). The intervention studied children’s energy balance-related behaviors and was reviewed by the Research Ethics Committee in the Humanities and Social and Behavioral Sciences of the University of Helsinki (22/2017; May 16, 2017) and found ethically acceptable. The prospective trial registration number is ISRCTN57165350 (January 2015). The recruitment process has been described in detail elsewhere (Ray et al., 2020). To summarize, participants were recruited through preschools (n =32) from two municipalities in Southern Finland: Salo (n = 29) and Riihim¨ aki (n =3). Consent to participate was received from 714 parents of 802 (47% of invitees) children from 3–6-year age groups (87 families had two or three participating children and 82 families had both the mother and father participating in the study). Baseline data was collected in autumn of 2017. The analytic sample of this study includes parents and children who fulfilled the following criteria: parent had data from at least one SFD item in secretive eating questions (mothers, n = 362; fathers, n =123) and a child that could be connected to at least one of the parents (n =403). 2.2. Measures Parents filled in a 55-item parental food frequency questionnaire (FFQ), a 51-item FFQ for their children, and answered questions about secretive eating and background information of the family. If available, both parents were requested to complete the parental FFQ and answer secretive eating questions. Questionnaires were sent to research participants by email; in cases where participants could not fill the electronic questionnaire, hardcopies were subsequently sent by post. 2.2.1. Consumption frequency of SFDs Table 1 presents the formation of all variables. In the parental FFQ, parents were asked to report foods and drinks consumed during the past week. When answering for children, parents were asked to report only those foods consumed outside preschool hours because parents would not have been able to report the foods eaten in preschool. In Finland, preschool meals are provided by municipalities, so most of the variation in diets of children is from outside preschool. The parental and child FFQs were identical, except that the parental FFQ included additional questions about coffee and alcohol, totaling four extra items. FFQs were based on a 47-item FFQ designed for the DAGIS survey, which has shown acceptable reproducibility (M¨ a¨ att¨ a et al., 2018) and validity compared to 3-day food records in children (Korkalo et al., 2019b). Both FFQs included three options for each food item: not at all, times per week, and times per day. Parents were instructed to either tick the “not at all” box or to write a number in one of the other columns. The consumption frequency variables for all seven SFD items for mothers, fathers, and children were formed by summing up the consumption frequencies (times/week) of all SFDs (Table 1). To ensure normally distributed residuals in regression analyses, outliers (>3SD, mothers n =3, fathers n =1) were winsorized with new values (=mean +3SD). 2.2.2. Parental secretive eating In the questionnaire, parents were asked if they avoided consuming certain foods (13 items) in the presence of their children. As the aim of this study was to examine SFDs, other foods were excluded from analyses (6 items: chips, popcorns, and other salty snacks, pizza, hamburgers, and other fast foods, restaurant meals, 100 % juices, mild alcohol drinks, and strong alcohol drinks). All seven SFDs were included in the analysis (Table 1). Answer options were: never, rarely, sometimes, T. Sarvanne et al. Appetite 200 (2024) 107578 3 usually, or I don’t eat this food at all. The categorial variable “Parental secretive eating” (eats in secret, does not eat in secret) and continuous variable “Secretive eating frequency” were formed, both separately for mothers and fathers (Table 1). 2.2.3. Confounding variables Analyses were adjusted with parent’s and child’s age, child’s gender, maternal/paternal educational level, and the number of adults and children in the household (Table 1), as these factors are found to be associated with family’s food consumption (Elfhag & Rasmussen, 2008; Fismen et al., 2020; Skaffari et al., 2023), food parenting practices, healthy modeling (Paasio et al., 2022), and the similarity between parent’s and child’s food consumption (Veps¨ al¨ ainen et al., 2018b). Parent’s age was categorized in thirds for cross-tabulations due to different age distribution in mothers and fathers. Child’s age was categorized for cross-tabulations in two groups (3–4; 5–6 years old). Continuous age variables were used in further analyses. Participants originally categorized with the parental status of “other” (stepdad, n = 1, foster mother, n =1, alternative caregiver, n =1) were subsequently categorized as either mothers or fathers based on their gender. Maternal/paternal educational level was categorized into three categories: low (=high school or vocational school graduate or lower), middle (=bachelor’s degree or equivalent), and high (=master’s degree or higher). Maternal/paternal educational level was used as a proxy for family’s socioeconomic status. Families with more than two parents (n =3) were combined to the category with two parents because of their small prevalence. 2.2.4. Reasons for parental secretive eating After the question about parental secretive eating in the questionnaire, parents had the possibility to further explain their responses in an open-ended format. The question was “If you avoid eating the following foods in the presence of your child, please provide more details here (e.g., why or when?)”. Altogether 167 parents (mothers n =150, fathers n = 17) provided answers to the open-ended question inquiring about parental secretive eating. Answers where an SFD and a reason for secretive eating were mentioned (n =63, 38% of all answers, mothers n =59, fathers n =4) were included in the analysis. Parents answered in Finnish. The citations presented in the present study were translated to English by a native English speaker to retain the original tone of the answer. 2.3. Explorative procedures 2.3.1. Statistical analyses Statistical analyses were carried out with the IBM SPSS Statistics 28.0 software package. A significance level of p <0.05 was applied for all Table 1 Variables used in the analyses. Variable name Survey question Original answer options Variable used in analyses Type of the variable Dependent variables Parental consumption frequency of all SFDs How many times have you eaten the following foods during the past week? Numeric answer separately for each food item: times/ day or times/week. Food items: 1) chocolate, 2) sweets, 3) ice cream, 4) cookies, snack bars and cereal bars, 5) cakes, cupcakes, buns, and other sweet pastries, 6) sugary juices, 7) soft drinks with added sugar. Sum variable: mothers, range 0–27; fathers, range 0–30 times/week. Continuous Children’s consumption frequency of all SFDs How many times has your child eaten the following foods during the past week? Only foods eaten outside preschool hours were considered As above. Sum variable: range 0–29 times/week Continuous Independent variables Parental secretive eating Do you avoid eating the following foods in the presence of your child/ children? Never; rarely; sometimes; usually; I don’t eat this food at all. Food items: 1) chocolate, 2) sweets, 3) ice cream, 4) cookies, snack bars and cereal bars, 5) cakes, cupcakes, buns, and other sweet pastries, 6) sugary juices, 7) soft drinks and energy drinks with added sugar. For each food item separately and all combined: Snacks in secret (sometimes; usually); Doesn’t snack in secret (never; rarely; I don’t eat this food at all); Categorial Secretive eating frequency As above. Each food item recategorized: 0 =never; I don’t eat this food at all; 1 =rarely; 2 =sometimes; 3 =usually. Then summed up. Range 0–21. Continuous Confounders Parent’s age Your/your partner’s age as years? Numeric answer Unmodified Continuous Categorized in thirds. Mothers: 24–33; 34–38; 39–49. Fathers: 23–35; 36–40; 41–59. Categorial Number of adults in the household How many people are part of your household currently including yourself Numeric answer Number of adults in two classes: one adult; two adults or more Categorial Maternal/paternal educational level What is the highest degree of education you have achieved? Comprehensive school; vocational school; secondary school; bachelor’s degree or equivalent; master’s degree; licentiate/doctoral degree; other Low (high school or vocational school graduate or lower); middle (bachelor’s degree or equivalent); high (master’s degree or higher) Categorial Child’s age What is your child’s date of birth? Date of birth. Age, in years and months, calculated from the date of birth. Continuous Categorized 3–4; 5-6 Categorial Child’s gender Gender of your child: Girl; boy; I don’t want to answer Gender of child in two classes: girl; boy Categorial Number of children in the household How many people are part of your household currently including yourself Numeric answer in age groups: 0–2; 3–6; 7–17 living at home Age groups summed up. Number of children in three classes: one child; two children; three or more Categorial SFD=Sugary foods and drinks. T. Sarvanne et al. Appetite 200 (2024) 107578 4 statistical tests. Descriptive statistical analyses were used to assess the prevalence of parental secretive eating. Differences in parental secretive eating between mothers and fathers, as well as between background variables among mothers and fathers separately, were explored with cross-tabulations using a Chi-square test. Results are presented as χ 2 (chi-square statistic value), degrees of freedom, and exact p-values. As the sample included 82 mother-father pairs from the same families, the differences between mothers and fathers in the prevalence of secretive eating was also explored within a sub-sample of non-dependent observations (n =403). In this sensitivity analysis, to keep the number of included fathers as high as possible, we excluded all mothers from the families from which a father was also a participant (n =82). A Mann Whitney U test was performed to evaluate whether consumption frequency of all SFD items (dependent variable) differed by parental secretive eating of any SFDs. This was further examined with food item specific consumption frequency (dependent variable) and food item specific parental secretive eating. Analyses were carried out separately for mothers, fathers, and children. For children, the analyses had two stages: the first grouping variable was the mother’s secretive eating and the second was the father’s secretive eating. Results are reported as z-values, mean ranks, means, Mann Whitney U test value and p-value (Monte Carlo p-values were used in analyses with consumption frequency of all SFDs, since the dataset was too large for the exact algorithms; exact p-values were used in food item specific analyses). Means were used instead of medians due to their better ability to describe small variations between these groups. Associations between secretive eating frequency (independent variable) and consumption frequency of all SFDs (dependent variable) were explored with linear regression separately in mothers, fathers, and children. In model 1, we included all independent variables one at a time. Independent variables were mother’s/father’s secretive eating frequency, mother’s/father’s age, and child’s age (continuous), and mother’s/father’s highest education, number of adults and children in the household, and child’s gender (categorial). Model 2 was a full model with all independent variables included at the same time. For children, analyses had two stages: in the first stage, the independent variables were mother’s secretive eating frequency, mother’s and child’s age, mother’s highest education, number of adults and children in the household, and child’s gender. In the second stage, the independent variables were father’s secretive eating frequency, father’s and child’s age, father’s highest education, number of adults and children in the household, and child’s gender. 2.3.2. Content analysis of the open-ended question Reasons for parental secretive eating were explored with content analysis by the first author. Analysis included four stages: 1) identifying responses that answered the research question, 2) coding relevant phrases, 3) classifying into groups, and 4) formation of themes. All answers were read thoroughly several times and process cycled interactively back and forth between data, codes, groups, and themes. Relevant data was identified with the following criteria: an answer had a mention of SFDs (e.g., “chocolate”, “sweets day”, “sugar”), and provided any kind of reason for parental secretive eating. Answers where parent did not specify if word like “treats” meant sugary treats or other treats (e.g., salty treats like chips) were excluded from the analysis. Units for coding were words and phrases that were used to describe reasons for parental secretive eating. There were no previously defined categories for coding as the research was explorative. Three categories were identified during the process: 1) parent related reasons, 2) child related reasons, and 3) home environment related reasons. Coded answers were categorized into seven groups that described the reason for parental secretive eating in more detail. Groups for parent related reasons were: 1) avoids modeling of SFD consumption because of own hardships with SFDs, 2) avoids modeling of breaking family’s rules for SFDs, 3) avoids always modeling SFD consumption. Groups for child related reasons were: 4) avoiding child’s disappointment through denial, 5) avoiding child’s requests, and 6) to support child’s healthier eating habits. For home environment related reasons there was only one group: 7) different food rules for adults and children in the family. Themes were created by identifying common groups and overlapping reasons for parental secretive eating between groups. Reasons related to modeling behavior and aims for child’s healthy eating habits (1, 2, 3, and 6) were combined into one theme, groups related to child’s behavior and emotions (4 and 5) into second theme and group for family’s food rules (7) into third theme. 3. Results 3.1. Participants Table 2 presents the characteristics of the participants. The statistical sample included more mothers (n =362) than fathers (n =123). The average ages of mothers and fathers were 35.8 years (SD 5.03) and 37.9 years (SD 5.96), respectively. Almost half of the mothers and nearly half of the fathers had middle level education (bachelor’s degree or equivalent). The number of girls and boys among the children participating in the study was similar; their mean age was 5.2 years (SD 1.02). Families with two children (50%) and at least two adults (87%) were most common. Participants in the content analysis were mostly mothers (94%). Otherwise, the sample was quite similar to the statistical sample. 3.2. Prevalence of parental secretive eating Fig. 1 shows the prevalence of parental secretive eating. Over two thirds of the parents (68%) reported parental secretive eating at least sometimes. The prevalence of parental secretive eating varied by food item and parental status. Chocolate (61%) and sweets (59%) were consumed in secret most commonly. About one third of the parents Table 2 Characteristics of the participants. Background variable Participants in statistical analyses Participants in content analysis n % n % Parental status Mother 362 75 59 94 Father 123 25 4 6 Total 485 100 63 100 Maternal educational level High 64 18 15 25 Middle 175 48 26 44 Low 108 30 16 27 Missing 15 4 2 3 Paternal educational level High 24 20 2 50 Middle 50 41 1 25 Low 41 33 1 25 Missing 8 7 0 0 Child’s gender Girl 193 48 33 52 Boy 210 52 30 48 Total 403 100 63 100 Number of children in the household One child 67 17 7 11 Two children 203 50 39 62 Three or more 118 29 14 22 Missing 15 4 3 5 Number of adults in the household One adult 38 9 2 3 Two or more 352 87 58 92 Missing 13 3 3 5 Maternal/paternal educational level: low (high school or vocational school graduate or lower), middle (bachelor’s degree or equivalent), high (master’s degree or higher) SFD=sugary foods and drinks T. Sarvanne et al. Appetite 200 (2024) 107578 5 Fig. 1. Prevalence of parental secretive eating. T. Sarvanne et al. Appetite 200 (2024) 107578 6 consumed cookies, snack bars, and cereal bars in secret (30%), and about one fourth consumed ice cream (28%) or cakes, cupcakes, buns, and other sweet pastries (27%). Parental secretive eating of soft drinks and energy drinks with added sugar (14%) and sugary juices (7%) was comparatively uncommon. Almost three quarters of mothers (73%) and over half of fathers (55%) reported parental secretive eating of any SFDs at least sometimes. Parental secretive eating was significantly more common in mothers than fathers (Fig. 1, χ 2 =14.43; df 1; p <0.001). Similar results were found for chocolate ( χ 2 =19.99; df 1; p <0.001), sweets ( χ 2 =17.65; df 1; p <0.001) and cookies, snack bars, and cereal bars ( χ 2 =6.83; df 1; p =0.009). For all but one food groups, the results of the sensitivity analysis were similar to the results derived from the original sample (data not shown). A statistically significant difference between mothers and fathers was observed in secretive eating of cakes, cupcakes, buns, and other sweet pastries (mothers 31% vs. fathers 20%, χ 2 =4.92; df 1; p =0.027). Significant differences in parental secretive eating of any SFDs by other confounding variables were found only in fathers. Fathers with younger children reported parental secretive eating more often than fathers with older children ( χ 2 =6.16; df 1; p =0.017). Parental secretive eating among fathers also differed depending on the number of children in the household; fathers with two children reported parental secretive eating more often than fathers with one child or three or more children ( χ 2 =6.09; df 2; p =0.047). 3.3. Difference in SFD consumption by parental secretive eating Table 3 presents the results from the Mann Whitney U tests. The results indicated that both mothers (z = − 3.22; p =0.002) and fathers (z = − 1.99; p =0.046) who reported parental secretive eating had a significantly higher SFD consumption frequency than those who did not. Mothers who reported secretive eating of chocolate (z = − 4.27; p < 0.001), sweets (z = − 5,12; p <0.001), cookies, snack bars, and cereal bars (z = − 4.49; p <0.001), cakes, cupcakes, buns and other sweet Table 3 Difference in mother’s, father’s, and children’s SFD consumption by parental secretive eating. Parents’ SFD consumption frequency Children’s SFD consumption frequency Group n Mean Mean rank Mann- Whitney U p-value Group n Mean Mean rank Mann- Whitney U pvalue Mother Chocolate Eats in secret 243 2.26 196.40 10351.500 <0.001 Eats in secret 239 0.93 179.15 13826.00 0.856 Doesn’t eat in secret 117 1.45 147.47 Doesn’t eat in secret 117 0.97 177.17 Sweets Eats in secret 232 1.77 198.20 9814.50 <0.001 Eats in secret 228 1.37 183.45 13464.50 0.193 Doesn’t eat in secret 124 0.89 141.65 Doesn’t eat in secret 128 1.19 169.69 Ice cream Eats in secret 104 0.80 192.84 11612.50 0.056 Eats in secret 106 1.03 193.20 11691.50 0.060 Doesn’t eat in secret 252 0.61 172.58 Doesn’t eat in secret 250 0.89 172.27 Cookies, snack bars, cereal bars Eats in secret 121 2.24 210.00 10285.00 <0.001 Eats in secret 121 1.69 193.71 12255.50 0.032 Doesn’t eat in secret 234 1.14 161.45 Doesn’t eat in secret 234 1.34 169.87 Cakes, cupcakes, buns, and other sweet pastries Eats in secret 105 2.06 206.94 10611.00 0.001 Eats in secret 106 1.27 188.64 12175.50 0.207 Doesn’t eat in secret 255 1.50 169.61 Doesn’t eat in secret 250 1.21 174.20 Sugary juices Eats in secret 28 1.32 236.88 3097.50 <0.001 Eats in secret 27 2.37 207.00 3626.00 0.109 Doesn’t eat in secret 333 0.64 176.30 Doesn’t eat in secret 328 1.98 175.55 Soft drinks and energy drinks with added sugar Eats in secret 51 1.18 237.92 5002.00 <0.001 Eats in secret 50 0.60 203.93 6328.50 0.015 Doesn’t eat in secret 310 0.52 171.64 Doesn’t eat in secret 305 0.42 173.75 Any SFD Eats in secret 252 9.11 182.90 8971.00 0.002 Eats in secret 256 8.32 178.03 12151.50 0.747 Doesn’t eat in secret 92 6.89 144.01 Doesn’t eat in secret 97 8.16 174.27 Father Chocolate Eats in secret 55 1.84 66.94 1598.50 0.153 Eats in secret 55 0.80 56.96 1593.00 0.272 Doesn’t eat in secret 68 1.31 58.01 Doesn’t eat in secret 65 1.03 63.49 Sweets Eats in secret 53 1.87 70.16 1422.50 0.022 Eats in secret 53 1.08 54.35 1449.50 0.064 Doesn’t eat in secret 70 1.24 55.82 Doesn’t eat in secret 67 1.46 65.37 Ice Cream Eats in secret 30 0.83 71.07 1093.00 0.059 Eats in secret 31 0.68 60.82 1369.50 0.949 Doesn’t eat in secret 92 0.65 58.38 Doesn’t eat in secret 89 0.87 60.39 Cookies, snack bars, cereal bars Eats in secret 26 1.81 70.42 1042.00 0.152 Eats in secret 26 1.19 53.06 1028.50 0.262 Doesn’t eat in secret 97 1.57 59.74 Doesn’t eat in secret 92 1.50 61.32 Cakes, cupcakes, buns, and other sweet pastries Eats in secret 25 1.96 72.08 973.00 0.102 Eats in secret 25 1.32 59.32 1158.00 0.845 Doesn’t eat in secret 98 1.34 59.43 Doesn’t eat in secret 95 1.40 60.81 Sugary juices Eats in secret 7 0.86 70.79 337.50 0.432 Eats in secret 6 2.33 58.75 331.50 0.905 Doesn’t eat in secret 115 1.51 60.93 Doesn’t eat in secret 114 2.27 60.59 Soft drinks and energy drinks with added sugar Eats in secret 16 2.13 81.78 491.50 0.003 Eats in secret 17 0.53 64.18 813.00 0.546 Doesn’t eat in secret 104 0.99 57.23 Doesn’t eat in secret 103 0.36 59.89 Any SFD Eats in secret 64 10.31 65.24 1360.50 0.046 Eats in secret 66 8.20 56.76 1535.00 0.325 Doesn’t eat in secret 54 8.31 52.69 Doesn’t eat in secret 52 8.67 62.98 SFD =sugary foods and drinks T. Sarvanne et al. Appetite 200 (2024) 107578 7 pastries (z = − 3.18; p =0.001), sugary juices (z = − 3.79; p <0.001), or soft drinks and energy drinks with added sugar (z = − 5.21; p <0.001) consumed these food items significantly more frequently than those who did not report secretive eating of these food items. Fathers who reported secretive eating of sweets (z = − 2.29; p =0.022) or soft drinks and energy drinks with added sugar (z = − 2.86; p =0.003) consumed these food items significantly more frequently than those who did not eat these in secret. Children’s consumption frequency of any SFDs did not differ significantly by mothers’ or fathers’ secretive eating. However, children whose mothers reported secretive eating of cookies, snack bars, and cereal bars (z = − 2.15; p =0.032) or soft drinks and energy drinks with added sugar (z = − 2.42; p =0.015) consumed these food items significantly more frequently than children whose mothers did not eat these in secret. No significant differences were observed between children’s SFD consumption frequencies and fathers’ secretive eating. 3.4. Associations between parental secretive eating and SFD consumption 3.4.1. Mothers Table 4 presents the results from the linear regressions. In model 1, mothers’ (β =0.277, p <0.001) secretive eating was positively associated with their SFD consumption frequency. These associations remained statistically significant in the full model (model 2: β =0.274; p <0.001). Of the independent variables included in the mother’s full model, parental secretive eating was most strongly associated with their SFD consumption frequency. In addition, mothers’ age (β =0.194; p = 0.001) was positively associated with their SFD consumption frequency. 3.4.2. Fathers In model 1, fathers’ (β =0.230, p =0.012) secretive eating was positively associated with their SFD consumption frequency. These associations remained statistically significant in the full model (model 2: β =0.210; p =0.028). Of the independent variables included in the father’s full model, parental secretive eating was only variable associated with their SFD consumption frequency. 3.4.3. Children No association with parental secretive eating and children’s SFD consumption frequency was detected in model 1 (mothers, β =0.053, p =0.323; fathers, β =-0.119, p =0.198) or 2 (mothers, β =0.031, p = 0.562; fathers, β =-0.143, p =0.167). There was a significant association between mother’s age (β =0.194; p <0.001), father’s age (β = 0.113; p =0.014), and the children’s SFD consumption frequency in model 1. Mother’s age (β =0.213; p <0.001) and number of adults in the household (β = − 0.118; p =0.037) were significantly associated with children’s consumption of SFDs in the mothers’ model. Children with older mothers consumed SFDs more frequently, and children with only one adult in the household consumed SFDs less frequently. In the full model with data from fathers, no significant associations were found. 3.5. Reasons for parental secretive eating We found three themes describing reasons for parental secretive eating. These themes were: 1) family food rules, 2) avoiding child’s requests, 3) aspiration for healthy modeling. 3.5.1. Family food rules Several parents, including both mothers and fathers, described that they had different SFD rules for the children and adults at home, indicating a common theme within the data. Many of these answers, simply mentioned secretive eating due to differing food rules without delving into further explanation or motivation for this behavior. All these answers included a mention of “sweets day” meaning that children were allowed to eat “sweets day” -treats only once or twice a week, but their parents ate these more often. However, some parents provided a more detailed answer. Some of them felt that their child should not eat SFDs as frequently as adults. Perceptions of suitable frequency varied between these parents. Most of these parents explained that their child was allowed to eat sweets at weekends, while few mothers explained that their children were only allowed SFDs on festive occasions. There was also one mother, who reported that sweets in particular were completely prohibited for her child, but she did not follow the same rule herself. There were also mothers and a father who explained that they tried to follow same SFD rules that they had for their child, but sometimes failed. “Children have a “sweets day” once a week, when I also eat sweets. Sometimes I crave chocolate/sweets in the middle of the week, and then I eat them in secret from my child.“(Mother, 33 years, 4 years old child) 3.5.2. Avoiding child’s requests Several mothers and a father explained their secretive eating by reporting that they aimed to avoid having their child request or desire SFDs. Some of them explained further that they wanted to prevent their child’s disappointment through denial, and few admitted that it just was easier to eat in secret. Some of these mothers described that their child would also want SFDs, if they saw their parent eating them. So, when they ate SFDs and did not want to give them to the child or could not because of allergies, they did it secretly. “I usually eat sweets in the evenings to prevent my child from feeling upset when they see adults eating treats. At parties, everyone can freely eat without worry and my child can also have sweets there.” (Mother, 36 years, 4 years old child) 3.5.3. Aspiration for healthy modeling Many participants, both mothers and a father, wanted to be healthy models for their child and avoid setting a “bad example” through their SFD consumption. Almost every one of these parents believed that SFD consumption was a “bad example” only in certain situations (e.g., before a meal, during weekdays), while few of the mothers held the perception that modeling consumption of sweets and chocolate in particular was consistently a negative influence on a child. Some mothers hoped that by avoiding eating too much SFDs in the presence of their child, children would learn healthier eating habits. These mothers were worried that their “bad habits” would be passed on to their child and described their sweets and chocolate consumption as a “vice”. “I don’t want my child to crave sweets and other sweet foods like I do. Sometimes I go days without and sometimes I buy a chocolate bar every day.” (Mother, 36 years, 5 years old child.) 4. Discussion This study investigated parental secretive eating and SFD consumption in Finnish families with 3–6-year-olds. The findings indicate that parental secretive eating is common among parents of preschool children. Parental secretive eating was more frequent among mothers than fathers. Among the various types of SFDs, sweets and chocolate were the most commonly eaten in secret. We found a positive association between parental secretive eating and SFD consumption frequency, both among mothers and fathers, whereas parental secretive eating was not associated with children’s SFD consumption. Three types of reasons for parental secretive eating were identified: family food rules, avoiding child’s requests, and aspiration for healthy modeling. 4.1. Prevalence of parental secretive eating A possible explanation for the finding that parental secretive eating was more common among mothers than fathers may be the fact that T. Sarvanne et al. Appetite 200 (2024) 107578 8 Table 4 The association between parental secretive eating and SFD consumption frequency in mothers, fathers, and children. Association with parent’s SFD consumption frequency Association with children’s SFD consumption frequency Model 1 Model 2 Model 1 Model 2 n β CI 95% n B β CI 95% n β CI 95% n B β CI 95% Mother Secretive eating frequency 344 0.277 0.227–0.493*** 328 0.355 0.274 0.221–0.488*** 353 0.053 −0.056–0.171 337 0.033 0.031 −0.080–0.147 Mother’s age 330 0.221 0.131–0.372*** 328 0.222 0.194 0.090–0.355** 339 0.221 0.107–0.299*** 337 0.198 0.213 0.087–0.308*** Maternal educational level High ref. ref. ref. ref. Middle 330 0.084 −0.275–2.191 328 −0.169 −0.015 −1.813–1.474 339 0.035 −0.673–1.325 337 0.466 0.050 −0.916–1.849 Low 330 ¡0.162 ¡3.319– -0.676** 328 −1.216 −0.099 −3.126–0.694 339 −0.083 −1.911–0.238 337 0.498 0.050 −1.104–2.101 Number of adults in the household Two or more ref. ref. ref. ref. One adult 334 −0.082 −3.664–0.503 328 −1.782 −0.090 −3.921–0.357 343 −0.102 −3.284–0.069 337 ¡1.901 ¡0.118 ¡3.683- -0.120* Number of children in the household One child ref. ref. ref. ref. Two children 332 0.063 −0.510–1.953 328 0.236 0.021 −1.516–1.988 341 0.053 −0.503–1.493 337 0.539 0.058 −0.907–1.986 Three or more 332 −0.039 −1.814–0.850 328 −0.434 −0.035 −2.342–1.474 341 0.006 −1.026–1.154 337 0.102 0.010 −1.482–1.687 Child’s gender Boy ref. ref. ref. ref. Girl 344 0.046 −0.697–1.750 328 0.002 0.000 −1.188–1.193 353 0.000 −1.003–1.000 337 −0.173 −0.019 −1.169–0.822 Child’s age 344 0.006 −0.591–0.659 328 −0.019 −0.003 −0.651–0.612 353 0.104 −0.004–1.016 337 0.422 0.089 −0.108–0.953 Constant −1.262 −7.025–4.501 −1.803 −6.622–3.017 Father Secretive eating frequency 118 0.230 0.068–0.545* 106 0.284 0.210 0.031–0.538* 118 −0.119 −0.285–0.060 107 −0.129 −0.143 −0.314–0.055 Father’s age 109 ¡0.217 ¡0.440– -0.033* 106 −0.076 −0.069 −0.290–0.139 110 0.045 −0.112–0.182 107 0.048 0.065 −0.111–0.206 Paternal educational level High ref. ref. ref. ref. Middle 110 −0.068 −3.395–1.608 106 1.032 0.078 −2.130–4.194 111 −0.001 −1.819–1.799 107 0.232 0.026 −2.082–2.545 Low 110 0.222 0.489–5.600* 106 3.001 0.215 −0.361–6.364 111 0.050 −1.385–2.365 107 0.750 0.080 −1.713–3.213 Number of adults in the household Two or more ref. ref. ref. ref. One adult 113 ¡0.196 ¡9.749– -0.296* 106 −2.595 −0.065 −10.151–4.962 113 −0.076 −5.530–−2.337 107 −5.252 −0.519 −11.926–1.421 Number of children in the household One child ref. ref. ref. ref. Two children 112 0.261 1.043–5.878** 106 2.765 0.207 −0.821–6.352 112 −0.057 −2.229–1.197 107 −0.805 −0.089 −3.429–1.818 Three children 112 −0.146 −4.868–0.593 106 0.631 0.043 −3.321–4.574 112 0.093 −0.955–2.805 107 0.147 0.015 −2.708–3.002 Child’s gender Boy ref. ref. ref. ref. Girl 118 0.128 −0.737–4.183 106 1.507 0.114 −0.891–3.906 118 −0.028 −2.042–1.500 107 −0.884 −0.099 −2.662–0.894 Child’s age 118 −0.173 −2.298–0.055 106 −0.999 −0.159 −2.167–0.169 118 −0.003 −0.865–0.840 107 −0.359 −0.084 −1.231–0.531 Constant 12.015 0.715–23.315 9.765 1.373–18.157 SFD =sugary foods and drinks. Maternal/paternal educational level: low (high school or vocational school graduate or lower), middle (bachelor’s degree or equivalent), high (master’s degree or higher). Mothers: Model 1: one independent variable at a time; Model 2: full model; Durbin Watson 0.762; F =6.06; p <0.001; R 2 =0.122. Fathers: Model 1: one independent variable at a time; Model 2: full model; Durbin Watson 0.447; F =3.09; p =0.003; R 2 =0.152. Children (mother): Model 1: one independent variable at a time; Model 2: full model; Durbin Watson 1.745; F =2.30; p =0.002; R 2 =0.051. Children (father): Model 1: one independent variable at a time; Model 2: full model; Durbin Watson 1.834; F =0.79; p =0.623; R 2= −0.018. B=Unstandardized coefficients, β =Standardized coefficients, CI =confidence interval ***p <0.001, **p <0.01, *p <0.05. T. Sarvanne et al.