Key Points
Question
Are family socioeconomic factors associated with eating disorder symptoms among adolescents?
Findings
This cohort study including 7824 participants found adolescents from more deprived backgrounds were at highest risk of eating disorder symptoms, especially those whose parents reported difficulties in affording cost of essential material goods. Children of parents who had only compulsory educations were also at highest risk of developing disordered eating compared with children of parents who were had university educations.
Meaning
These findings suggest that reducing socioeconomic inequalities could also help prevent eating disorders in the general population.
This cohort study uses data from the Avon Longitudinal Study of Parents and Children to examine the associations of childhood socioeconomic indicators and eating disorder symptoms across adolescence.
Abstract
Importance
Adolescents who experienced childhood socioeconomic deprivation report more eating disorder symptoms compared with their counterparts with higher socioeconomic status but may have more barriers in receiving diagnoses and accessing eating disorder services.
Objective
To investigate the associations of childhood socioeconomic indicators with eating disorder symptoms across adolescence.
Design, Setting, and Participants
This prospective cohort study used a population-based sample from the Avon Longitudinal Study of Parents and Children (ALSPAC). ALSPAC recruited pregnant women in the former region of Avon, United Kingdom, with expected delivery dates from April 1, 1991, to December 31, 1992. This study used follow-up data of the mother-offspring collected until 2010. The final analytical sample included children who were alive at 1 year of age and who had complete exposures, retaining 1 twin at random. Data were analyzed from October 1, 2022, to November 25, 2024.
Exposures
The main exposures were parental income, education, occupation, financial hardship (range, 0-15; higher score indicates more hardship), reported by mothers between 32 weeks’ gestation and 47 months postpartum, and area-level deprivation, derived from the Office for National Statistics indicators linked to the participant’s residential post code at 32 weeks’ gestation.
Main Outcomes and Measure
Primary outcomes were disordered eating, weight and shape concerns, and body dissatisfaction at ages 14, 16, and 18 years. Individual disordered eating behavior was a secondary outcome.
Results
The sample included 7824 participants (4003 [51.1%] male). A 1-point increase in financial hardship was associated with increased odds of disordered eating (odds ratio [OR], 1.06; 95% CI, 1.04-1.10), an increase in weight and shape concerns (coefficient, 0.02 (95% CI, 0.01-0.04), and an increase in body dissatisfaction (coefficient, 0.22 (95% CI, 0.06-0.37). Lower parental education was associated with higher odds of disordered eating (OR, 1.80; 95% CI, 1.46 to 2.23).
Conclusions and Relevance
This cohort study using ALSPAC data found that eating disorder symptoms were more common in individuals experiencing socioeconomic deprivation. Potential socioeconomic inequalities in eating disorder presentation and diagnosis in clinical settings require further investigation. Reducing population-level socioeconomic inequalities could also aid eating disorder prevention.
Introduction
Socioeconomic deprivation is a major determinant of poor mental and physical health in children.1,2 In the UK, 1 in 3 children live in poverty, with increasing proportions living in extreme poverty.3 Children from the most deprived households experience a higher prevalence of mental health problems, such as depression and anxiety, compared with those living in the least deprived households.1,2,3
In contrast, it is often believed that eating disorders are more common in people with families from higher socioeconomic positions,2,4,5 but evidence supporting this association is mixed. Most longitudinal register-based studies, where diagnoses are derived from clinical records, find a higher incidence of eating disorders in people whose parents had higher income and education and who lived in more affluent areas.2,4,5,6,7,8,9,10 Conversely, cross-sectional11,12,13,14 studies either find no evidence of any differences in the distribution of self-reported eating disorder symptoms by parental socioeconomic position12,15,16 or find increased risk of these symptoms in people from more deprived backgrounds.11,13,14,16,17,18,19,20,21,22 Longitudinal population studies found that individuals with lower personal socioeconomic position in adulthood or children from parents with lower educational attainment, greater financial hardship, or from lower overall socioeconomic position reported more eating disorder symptoms.15,16,17,18,19,20,21,22
Nevertheless, the literature has several limitations. Findings from register-based studies may be affected by selection bias if people from more deprived backgrounds experience barriers in accessing eating disorder services.11,23 On the other hand, investigating self-reported symptoms in general population samples reduces this risk, but it can still inform us on etiology of eating disorders. Longitudinal studies either used a long follow-up spanning into adulthood2,4,21,22 or, when focusing on young people, did not include the peak time of eating disorder symptom onset (age approximately 16 years),15,16,17,18,20,24 which could affect findings if early onset cases are underpinned by different etiological mechanisms.25 Most studies adjusted their analyses for factors that are potentially on the causal pathway between family socioeconomic position and offspring eating disorder, such as adverse life experiences22 or offspring’s body mass index (BMI),16,18,19,21,22 which can bias results. Finally, existing studies used either a single measure of socioeconomic position or composite indices, but exploring a wide range of socioeconomic indicators may be more helpful to develop future preventative strategies. In this study, we investigated the longitudinal association of parental income, occupation, education, financial hardship, and area-level deprivation in early childhood with adolescent eating disorder symptoms in a large UK general-population cohort.
Methods
Sample
This cohort study used data from the Avon Longitudinal Study of Parents and Children (ALSPAC). Ethical approval for the study was obtained from the ALSPAC Ethics and Law Committee and the Local Research Ethics Committees. Informed consent for the use of data collected via questionnaires and clinics was obtained from participants following the recommendations of the ALSPAC Ethics and Law Committee at the time. This report follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.
ALSPAC is a birth cohort study that recruited 14 541 pregnant women in the former region of Avon, UK, with expected delivery dates from April 1, 1991, to December 31, 1992. Of these pregnancies, 14 062 (96.1%) resulted in live births and 13 988 children (93.8%) were alive at 1 year.26,27 In this study, we included children from this original sample who had data available on all the exposures. In the case of twins, we retained 1 child at random to avoid potential overestimation of associations due to clustering of environmental and genetic risk.
Outcomes
We used 3 different outcomes capturing behavioral and cognitive symptoms of eating disorders. We defined disordered eating as a binary variable based on whether adolescents reported any binge eating, purging, or restrictive eating (excessive dieting and fasting) or none of these behaviors in the previous 12 months at ages 14, 16, and 18 years using modified questions from the validated Youth Risk Behavior Surveillance System questionnaire,28 which has been used in previous literature.29 We used binge eating, purging, and restrictive eating individually as a secondary outcome to investigate their specific associations with socioeconomic position.
Body dissatisfaction was self-reported by adolescents at age 14 years using 11 items from the body satisfaction scale.30 Individual items were summed and ranged from 11 to 55, with higher scores indicating greater body dissatisfaction.
Weight and shape concerns were self-reported by adolescents at ages 14 and 18 years using 2 items from the McKnight Risk Factor survey.31 We summed the 2 items. The total score ranged from 0 to 6, with higher scores indicating greater concerns. Further details on all outcomes are provided in eMethods 1 in Supplement 1.
Exposures
We derived highest parental occupation (professional, managerial, skilled nonmanual, skilled manual, and semiskilled or unskilled manual) and highest parental education (university degree, A-level, compulsory education) from individual paternal and maternal measures as reported by the mother at 32 weeks of gestation. If either parent had missing data on these variables, or in cases of single-parent households, we used data on the available parent.
At 32 weeks’ gestation, mothers were also asked how difficult they were finding it to afford food, heating, clothing, rent or mortgage, and items for their baby. Scores ranging from 0 to 15, with higher scores representing greater financial hardship.
Mothers reported weekly family income when the study child was ages 33 months and 47 months. We calculated mean parental income across these time points and weighted income by number of people within the household according to their age and estimated housing benefits using the Organisation for Economic Co-operation and Development modified scale and split into fifths.32
Mothers also provided residential post codes throughout gestation. These were linked to enumeration district codes (small census areas) and 1991 Census data Townsend deprivation index scores,33 which were subsequently standardized. Higher values indicate greater deprivation (eMethods 2 in Supplement 1).
Confounders
We identified potential confounders based on literature-informed a priori assumptions and using direct acyclic graphs to model our assumptions. In main analyses, we mutually adjusted each exposure for all other indicators of socioeconomic position, given their interconnectedness and intergenerationally (eFigure 1 in Supplement 1). We did not include child characteristics, as those are on the causal pathway between the socioeconomic position and eating disorder risk. We also hypothesized that maternal characteristics could be on the causal pathway between exposures and outcomes, as socioeconomic position in pregnancy could reflect earlier socioeconomic position and this could affect subsequent maternal socioeconomic indicators (eFigure 2 in Supplement 1). However, in sensitivity analyses, we tested competing causal assumptions and confounding structures. First, we hypothesized that some socioeconomic indicators could affect others, for instance, that education could affect subsequent income (eFigure 3 in Supplement 1).34 Second, we hypothesized that maternal characteristics35,36 could affect subsequent socioeconomic position; for instance, maternal history of eating disorders could affect the mother’s educational attainment, since peak age onset of eating disorders coincides with adolescence (eFigure 4 in Supplement 1). In sensitivity analyses, given the small number of participants from minoritized ethnic groups (eg, Black African, Indian), we also additionally adjusted analyses for child’s ethnicity as a proxy of parental ethnicity, as people from minoritized ethnic groups face additional barriers that can affect their socioeconomic status and eating disorder risk (eMethods 3 in Supplement 1).34 The child was classified as being of minoritized ethnicity if either the mother or partner or both self-reported being from 1 of the following backgrounds: Bangladeshi, Black African, Black Caribbean, Chinese, Indian, Other Black, Other (not including White), and Pakistani.
Statistical Analysis
Protocol deviations are available in eMethods 4 in Supplement 1. We described sample characteristics overall and by levels of exposures using frequencies with proportions and means with SDs. For participants with complete exposure data, we compared the distribution of exposures and potential confounders between the full sample and participants who had missing outcome measures.
To investigate the association between each socioeconomic indicator and eating disorder symptoms, we used univariable and multivariable multilevel logistic (any and each individual disordered eating) and linear (for weight and shape concerns) regression models with time of outcome assessment nested within individuals. First, we ran an unconditional model only including a mean-centered indicator of age at outcome measurements to describe how the latter changed across adolescence. For disordered eating, where we had 3 measurements available, we also added a quadratic term for age to test for nonlinear associations with age, retaining this in the models if there was evidence of an association. Subsequently, we ran a univariable model for each exposure and a multivariable model, adjusting each exposure for all other indicators of socioeconomic position. For the any disordered eating and weight and shape concerns outcomes, we subsequently included an interaction between each exposure and age to investigate whether there were differential associations with the exposure based on timing of outcome measurement for any disordered eating and weight and shape concerns. We stratified results by age when we found evidence of an interaction.
To investigate associations with body dissatisfaction at age 14 years, we used univariable and multivariable linear regression models mutually adjusting each socioeconomic indicators for all other indicators. We imputed missing confounder and outcome data using multiple imputation by chained equations for participants with complete data on exposure. We imputed 50 datasets on the assumption that the data were missing at random using all the variables included in the final models and auxiliary variables (eMethods 5 in Supplement 1).
We ran 4 sets of sensitivity analyses. First, we further adjusted the main multivariable models for maternal marital status and history of eating disorders and depression. Second, we reran the main multivariable models adjusting parental occupation for parental education, family income for parental occupation and education, and financial hardship for family income and highest parental occupation and education. We also adjusted all the main analyses for participants’ ethnicity to explore how ethnicity may confound our associations. To explore whether missing data patterns affected our effect sizes and estimates, we ran all our main analyses in a sample of participants with complete exposures and outcome (for body dissatisfaction) or at least 1 time point of outcome measurement available (for disordered eating and weight and shape concerns). Findings were interpreted with reference to effect sizes and CIs. Analyses were conducted using Stata software version 17 (StataCorp) from October 1, 2022, to November 25, 2024.
Results
Sample Characteristics
From the total sample of 13 988 ALSPAC children alive at 1-year, 7824 children (55.9%) had complete data on all exposures after removing 1 twin and were included in the analytical sample. There were 4003 (51.1%) male children; 294 children (3.8%) were from a minoritize ethnicity and 7420 children (96.2%) were White.
A large proportion of participants’ parents had a managerial occupation (3404 children [43.5%]) and compulsory education as their highest educational qualification (3127 children [40.0%]) (Table 1) and were in the highest fifth of income categories (1704 children [21.8%]) (Table 2). Most families did not experience financial hardship (5981 children [76.4%]) and lived in areas of low deprivation (576 children [73.7%]) during pregnancy (Table 2).
Table 1. Sample Characteristics by Parental Occupation and Education.
| Characteristic | Participants, No. (%) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Analytical sample | Highest parental occupationa | Highest parental educationb | ||||||||
| Professional | Managerial | Skilled nonmanual | Skilled manual | Semiskilled or unskilled manual | University degree | Advanced level | Compulsory education | |||
| Total | 7824 (100) | 1153 (14.7) | 3404 (43.5) | 1996 (25.5) | 883 (11.3) | 388 (5.0) | 1954 (25.0) | 2743 (35.1) | 3127 (40.0) | |
| Sex | ||||||||||
| Male | 4003 (51.1) | 593 (14.8) | 1731 (43.2) | 1036 (22.9) | 452 (11.3) | 191 (4.8) | 989 (24.7) | 1402 (35.0) | 1612 (40.3) | |
| Female | 3821 (48.9) | 560 (14.7) | 1673 (43.8) | 960 (25.1) | 431 (11.3) | 197 (5.2) | 965 (25.3) | 1341 (35.1) | 1515 (39.6) | |
| Ethnicity | ||||||||||
| Minoritized ethnicityc | 294 (3.8) | 59 (20.1) | 118 (40.1) | 65 (22.1) | 37(12.6) | 15 (5.1) | 93 (31.6) | 92 (31.3) | 109 (37.1) | |
| White | 7420 (96.2) | 1086 (14.6) | 3257 (43.9) | 1895 (25.5) | 830 (11.2) | 352 (4.7) | 1850 (24.9) | 2626 (35.4) | 2944 (39.7) | |
| Maternal history of eating disordersd | ||||||||||
| No | 7411 (96.4) | 1100 (14.9) | 3219 (43.4) | 1909 (25.8) | 822 (11.1) | 361 (4.9) | 1850 (25.0) | 2603 (35.1) | 2958 (39.9) | |
| Yes | 277 (3.6) | 38 (13.7) | 139 (50.2) | 55 (19.9) | 30 (10.8) | 15 (5.4) | 84 (30.3) | 101 (36.5) | 92 (33.2) | |
| Maternal marital statuse | ||||||||||
| Married | 6208 (80.2) | 1043 (16.8) | 2808 (45.2) | 1555 (25.1) | 582 (9.4) | 220 (3.5) | 1700 (27.4) | 2245 (36.2) | 2263 (36.4) | |
| Not married | 1530 (19.8) | 106 (6.9) | 568 (37.1) | 417 (27.3) | 277 (18.1) | 162 (10.6) | 244 (16.0) | 476 (31.1) | 810 (52.9) | |
| Maternal age at birth of study child, mean (SD), y | 28.7 (4.6) | 31.0 (3.8) | 29.8 (4.3) | 27.7 (4.3) | 27.5 (4.7) | 27.1 (4.9) | 31.3 (3.7) | 29.1 (4.2) | 27.4 (4.4) | |
| Maternal depressive symptoms, mean (SD)e | 6.59 (4.7) | 5.58 (4.1) | 6.21 (4.5) | 6.49 (4.6) | 7.3 (4.9) | 7.72 (4.6) | 5.8 (4.3) | 6.3 (4.5) | 6.8 (4.8) | |
Among 2-parental households, 9.7% were missing data for 1 parent.
Among 2-parental households, 3.0% were missing data for 1 parent.
Child was classified as being of minoritized ethnicity if either the mother or partner or both reported being from 1 of the following backgrounds: Bangladeshi, Black African, Black Caribbean, Chinese, Indian, Other Black, Pakistani, and other (not including White).
Maternal report at 12 weeks of pregnancy.
Maternal report at 8 weeks of pregnancy.
Total Edinburgh Postnatal Depression Scale score at 12 weeks of pregnancy.
Table 2. Sample Characteristics by Parental Income, Financial Hardship, and Area-Level Deprivation.
| Characteristic | Participants, No. (%) | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Fifths of equivalized parental income | Financial hardshipa | Area-level deprivationb | |||||||
| Highest | Second | Third | Fourth | Lowest | No | Yes | Low deprivation | High deprivation | |
| Total | 1704 (21.8) | 1670 (21.3) | 1560 (19.9) | 1535 (19.6) | 1355 (17.3) | 5981 (76.4) | 1843 (23.6) | 5762 (73.7) | 2062 (26.4) |
| Sex | |||||||||
| Male | 859 (21.4) | 839 (21.0) | 813 (20.3) | 779 (19.5) | 713 (17.8) | 3.056 (76.3) | 947 (22.7) | 2956 (73.8) | 1047 (26.2) |
| Female | 845 (22.1) | 831 (21.7) | 747 (19.6) | 756 (19.8) | 642 (16.8) | 2925 (76.5) | 896 (23.5) | 2806 (73.4) | 1015 (26.6) |
| Ethnicityc | |||||||||
| Minoritized ethnicity | 63 (21.4) | 53 (18.0) | 47 (16.0) | 56 (19.0) | 75 (25.5) | 193 (65.7) | 101 (34.4) | 151 (51.4) | 143 (48.6) |
| White | 1632 (22.0) | 1602 (21.6) | 1491 (20.1) | 1456 (19.6) | 1239 (16.7) | 5733 (77.3) | 1687 (22.7) | 5539 (74.6) | 1881 (25.4) |
| Maternal history of eating disordersd | |||||||||
| No | 1625 (21.9) | 1591 (21.5) | 1491 (20.1) | 1458 (19.7) | 1246 (16.8) | 5699 (76.9) | 1712 (23.1) | 1928 (26.0) | 5483 (74.0) |
| Yes | 64 (23.1) | 60 (21.7) | 43 (15.5) | 46 (16.6) | 64 (23.1) | 197 (71.1) | 80 (28.9) | 87 (31.4) | 190 (68.6) |
| Maternal marital statuse | |||||||||
| Married | 1492 (24.0) | 1435 (23.1) | 1298 (20.9) | 1164 (18.8) | 819 (13.2) | 4949 (79.7) | 1259 (20.3) | 1344 (21.6) | 4864 (78.4) |
| Not married | 201 (13.1) | 224 (14.7) | 250 (16.3) | 353 (23.1) | 502 (32.8) | 970 (63.4) | 560 (36.6) | 683 (44.6) | 847 (55.4) |
| Maternal age at birth of study child, mean (SD) | 30.8 (3.8) | 29.7 (4.2) | 28.5 (4.3) | 28.4 (4.5) | 27.6 (5.0) | 28.2 (5.0) | 29.5 (4.2) | 29.4 (4.4) | 28.2 (4.6) |
| Maternal depressive symptoms, mean (SD)f | 5.44 (4.2) | 6.09 (4.4) | 6.34 (4.3) | 6.66 (4.5) | 7.77 (5.0) | 5.81 (4.3) | 8.33 (4.8) | 6.07 (4.4) | 7.15 (4.7) |
For descriptive table purposes we defined experiencing high financial hardship as scoring 5 or above (75th percentile of scores of the total sample) on the financial hardship scale.
For descriptive table purposes we defined high area-level deprivation as having a standardized Townsend score equal or lower than 0.36, which was the mean deprivation score in the UK in 1990.
Child was classified as being of minoritized ethnicity if either the mother or partner or both reported being from 1 of the following backgrounds: Bangladeshi, Black African, Black Caribbean, Chinese, Indian, Other Black, Pakistani, and Other (not including White).
Maternal report at 12 weeks of pregnancy.
Maternal report at 8 weeks of pregnancy.
Total Edinburgh Postnatal Depression Scale score at 12 weeks of pregnancy.
The distribution of participants in terms of sex assigned at birth was comparable across all socioeconomic position indicators. Parents of children from minoritized ethnic background had lower income, experienced more financial hardship, and lived in higher deprivation areas. A greater proportion of participants with unmarried mothers reported more deprivation across all indicators. Mean levels of maternal depressive symptoms were progressively higher and mean maternal age lower in categories denoting more deprived backgrounds (Table 1 and Table 2). Outcome measurements were more commonly missing among participants from more deprived backgrounds across all socioeconomic indicators as well as in those with younger and single mothers as well as mothers with greater depressive symptoms. (eTable 1 in Supplement 1)
Descriptive Data of Eating Disorder Symptom
At 14 years, 338 participants (7.9%) experienced disordered eating. This proportion increased at age 16 years (574 participants [15.9%]) and at age 18 years (462 participants [18.9%]). At age 14 years, the mean (SD) weight and shape concern score was 1.7 (1.2) and 2.0 (1.5) at age 16 years. The mean body dissatisfaction score at age 14 years was 25.6 (10.3) (eTable 2 in Supplement 1).
Early-Life Socioeconomic Position and Adolescent Disordered Eating
All unconditional models are in eTable 3 and eTable 4 in Supplement 1. In univariable models, participants whose parents had only completed compulsory education (OR, 1.80; 95% CI, 1.46-2.23), had a semiskilled or unskilled occupation (OR, 2.09; 95% CI, 1.32-3.34), and whose income was in the lowest fourth (OR, 1.35; 95% CI, 1.04-1.74) and fifth (OR, 1.34; 95% CI, 1.01-2.79) of income distribution had higher odds of experiencing disordered eating compared with adolescents whose parents had university-level education or a professional occupation and who were in the highest fifth of income distribution, respectively (Table 3). A 1-point increase in financial hardship score (OR, 1.07; 95% CI, 1.04-1.10) and a 1-SD increase in area-level deprivation (OR, 1.05; 95% CI, 1.02-1.09) were associated with higher odds of experiencing disordered eating in adolescence.
Table 3. Univariable and Multivariable Multilevel Logistic and Linear Regression Models for Any Behavioral Eating Disorder Symptoms and Weight and Shape Concerns at Ages 14, 16, and 18 Years According To Parental Socioeconomic Position.
| Parental socioeconomic position | Disordered eating | Weight and shape concerns | ||||||
|---|---|---|---|---|---|---|---|---|
| Univariable model | Multivariable model | Univariable model | Multivariable model | |||||
| OR (95% CI) | P value | OR (95% CI) | P value | Mean difference (95% CI) | P value | Mean difference (95% CI) | P value | |
| Highest parental education | ||||||||
| University degree | 1 [Reference] | NA | 1 [Reference] | NA | 0 [Reference] | NA | 0 [Reference] | NA |
| A-level | 1.35 (1.09 to 1.68) | .007 | 1.31 (1.03 to 1.66) | .03 | 0.07 (−0.02 to 0.15) | .12 | 0.02 (−0.07 to 0.12) | .67 |
| Compulsory education | 1.80 (1.46 to 2.23) | <.001 | 1.64 (1.24 to 2.16) | .001 | 0.08 (−0.002 to 0.18) | .06 | 0.02 (−0.09 to 0.14) | .70 |
| Highest parental occupation | ||||||||
| Professional | 1 [Reference] | NA | 1 [Reference] | NA | 0 [Reference] | NA | 0 [Reference] | NA |
| Managerial | 1.27 (1.00 to 1.62) | .05 | 1.05 (0.81 to 1.36) | .73 | 0.10 (0.01 to 0.28) | .03 | 0.08 (−0.03 to 0.19) | .14 |
| Skilled nonmanual | 1.54 (1.17 to 2.02) | .002 | 1.09 (0.78 to 1.51) | .61 | 0.13 (0.03 to 0.22) | .01 | 0.08 (−0.04 to 0.20) | .20 |
| Skilled manual | 1.54 (1.08 to 2.19) | .02 | 0.99 (0.65 to 1.51) | .97 | 0.12 (−0.04 to 0.27) | .14 | 0.05 (−0.13 to 0.23) | .61 |
| Semiskilled or unskilled | 2.09 (1.32 to 3.34) | .002 | 1.26 (0.75 to 2.11) | .38 | 0.18 (−0.05 to 0.40) | .12 | 0.09 (−0.15 to 0.35) | .43 |
| Fifths of equivalized family income | ||||||||
| Highest 20% | 1 [Reference] | NA | 1 [Reference] | NA | 0 [Reference] | NA | 0 [Reference] | NA |
| Second | 0.99 (0.77 to 1.30) | .99 | 0.84 (0.65 to 1.10) | .21 | 0.02 (−0.07 to 0.12) | .63 | −0.01 (−0.11 to 0.09) | .79 |
| Third | 1.08 (0.82 to 1.42) | .60 | 0.79 (0.59 to 1.04) | .10 | 0.07 (−0.04 to 0.19) | .21 | 0.006 (−0.12 to 0.13) | .93 |
| Fourth | 1.35 (1.04 to 1.74) | .02 | 0.88 (0.66 to 1.16) | .36 | 0.10 (−0.01 to 0.21) | .09 | −0.0002 (−0.12 to 0.12) | .99 |
| Lowest 20% | 1.34 (1.01 to 1.79). | .04 | 0.75 (0.55 to 1.03) | .08 | 0.06 (−0.07 to 0.18) | .37 | −0.08 (−0.23 to 0.06) | .27 |
| Financial hardship score, per 1-point increase | 1.07 (1.04 to 1.10) | <.001 | 1.06 (1.03 to 1.09) | <.001 | 0.02 (0.01 to 0.04)a | .001 | 0.02 (0.01 to 0.04)a | .003 |
| Standardised area-level deprivation score, per 1-SD increase | 1.05 (1.02 to 1.09) | .002 | 1.03 (0.99 to 1.07) | .07 | 0.01 (0.001 to 0.03)a | .04 | 0.01 (−0.003 to 0.02)a | .12 |
Abbreviations: NA, not applicable; OR, odds ratio.
Expressed as coefficient (95% CI).
In the adjusted analyses, there was still strong evidence that adolescents whose parents only had compulsory education had higher odds of disordered eating compared to adolescents whose parents had university-level education (OR, 1.64; 95% CI, 1.24-2.16) and that a 1-point increase in financial hardship score (OR, 1.06; 95% CI, 1.03-1.09) was associated with higher odds of disordered eating (Table 3). The association between area-level deprivation and disordered eating was attenuated when adjusting for remaining socioeconomic conditions (OR, 1.03; 95% CI, 0.99-1.07). There was no evidence of an association for lower parental occupation and income. When investigating interactions between each exposure and age of outcome measurements, we found evidence of an interaction between income and age (P for interaction = .03), wherein children from lower income brackets experienced more disordered eating when they were younger (eTable 5 in Supplement 1).
Early-Life Socioeconomic Position and Adolescent Weight and Shape Concerns and Body Dissatisfaction
In univariable models, a 1-point increase in parental financial hardship score was associated with an increase in weight and shape concern score (coefficient, 0.02; 95% CI, 0.01-0.04) and body dissatisfaction score (coefficient, 0.24; 95% CI, 0.10-0.39) (Table 4). Associations for weight and shape concern (coefficient, 0.02; 95% CI, 0.01-0.04) and body dissatisfaction (coefficient, 0.22; 95% CI, 0.06-0.37) remained similar after adjusting. There was no evidence of an association between the remaining socioeconomic positions and cognitive eating disorder symptoms.
Table 4. Linear Regression Model for Body Dissatisfaction at Age 14 Years According to Parental Socioeconomic Position.
| Parental socioeconomic position indicators | Body dissatisfaction | |||
|---|---|---|---|---|
| Univariable model | Multivariable model | |||
| Mean difference (95% CI) | P value | Mean difference (95% CI) | P value | |
| Highest parental education | ||||
| University degree | 0 [Reference] | NA | 0 [Reference] | NA |
| A-level | 0.23 (−0.94 to 1.41) | .69 | −0.03 (−1.12 to 1.06) | .96 |
| Compulsory education | 1.08 (−0.08 to 2.23) | .07 | 0.57 (−0.65 to 1.79) | .36 |
| Highest parental occupation | ||||
| Professional | 0 [Reference] | NA | 0 [Reference] | NA |
| Managerial | 0.34 (−1.02 to 1.70) | .62 | 0.01 (−1.40 to 1.43) | .98 |
| Skilled nonmanual | 0.77 (−0.72 to 2.25) | .31 | 0.11 (−1.50 to 1.73) | .89 |
| Skilled manual | 1.49 (−0.69 to 3.66) | .18 | 0.48 (−1.89 to 2.85) | .69 |
| Semiskilled or unskilled | 0.93 (−1.95 to 3.81) | .52 | −0.27 (−3.25 to 2.72) | .86 |
| Fifths of equivalized family income | ||||
| Highest 20% | 0 [Reference] | NA | 0 [Reference] | NA |
| Second | 0.61 (−0.78 to 2.01) | .38 | 0.31 (−1.04 to 1.67) | .64 |
| Third | −0.35 (−2.09 to 1.39) | .69 | −0.94 (−2.69 to 0.81) | .29 |
| Fourth | 2.17 (0.45 to 3.91) | .01 | 1.27 (−0.48 to 3.01) | .15 |
| Lowest 20% | 0.72 (−1.07 to 2.50) | .43 | −0.61 (−2.58 to 1.37) | .54 |
| Financial hardship score, coefficient (95% CI) per 1-point increase | 0.24 (0.10 to 0.39) | .001 | 0.22 (0.06 to 0.37), | .007 |
| Standardised Area-level deprivation score, coefficient (95% CI) per 1-SD increase | 0.13 (−0.03 to 0.29) | .12 | 0.07 (−0.09 to 0.22) | .40 |
Abbreviation: NA, not applicable.
In univariable models, a 1-SD increase in area-level deprivation was associated with an increase in weight and shape concern scores (coefficient, 0.01; 95% CI, 0.00-0.03), which was completely attenuated in the adjusted analyses. We did not observe evidence of an interaction between socioeconomic position and age of weight and shape concern measurements.
Early-Life Socioeconomic Position and Individual Adolescent Disordered Eating
In the analyses of individual disordered eating, we found that lower education and greater financial hardship were associated with increased odds of restrictive eating. Higher area-level deprivation was associated with increased odds of binge eating and purging (eTable 6 in Supplement 1).
Sensitivity Analyses
Sensitivity analyses adjusting for ad hoc socioeconomic position indicators (eTable 7 and eTable 8 in Supplement 1), including maternal characteristics (eTable 9 and eTable 10 in Supplement 1), restricting to participants with complete data (eTable 11 and eTable 12 in Supplement 1), or adjusting for child ethnicity (eTable 13 and eTable 14 in Supplement 1) yielded comparable estimates to those observed in the main analyses, albeit with wider CIs in some cases. This may be due to the loss of power in these analyses.
Discussion
In this cohort study, we found that participants from more deprived backgrounds experienced greater eating disorder symptoms throughout adolescence. More severe financial hardship was associated with increased risk of disordered eating, weight and shape concerns, and body dissatisfaction. Lower parental educational attainment was strongly associated with increased odds of offspring’s disordered eating in adolescence.
Interpretation of Findings and Comparison With Previous Literature
Our study found opposite associations from those of register-based studies2,4,5,6,7,8,9,10 Anorexia nervosa is often overrepresented in clinical samples.4 If anorexia nervosa has a different pattern of association with socioeconomic position, we might not have been able to capture this using our restrictive eating measure, which may not represent severe fasting and dieting present in anorexia nervosa. However, register-based studies also find an association between high socioeconomic position and higher incidence of diagnosed bulimia nervosa and eating disorders not otherwise specified,4 which we should have captured more accurately with our outcome measures. Therefore, we hypothesize that the discrepancy between the socioeconomic patterning of clinical diagnoses and self-reported symptoms might be explained by inequalities in identification of eating disorders and access to services rather than measurement issues.
On the other hand, we expand on previous findings using self-reported symptoms16,17,18,19,20,21,22 by showing that the association between deprivation and eating disorder symptoms extends to the full adolescent period. This suggests persistent effects of childhood deprivation extending into the period of greatest risk for eating disorders symptoms.
Greater financial difficulties and lower parental educational attainment have the strongest associations with eating disorder symptoms. While these are small effect sizes, in general population settings, these can reflect large shifts in the number of individuals who experience eating disorder symptoms.37 It is possible that putative risk factors for eating disorders more commonly observed in individuals from lower socioeconomic positions, such as higher child BMI,38 increased food insecurity,39,40,41 and greater experience of childhood adversities,18,42 might explain these associations.
The lack of evidence for an association between low parental education and cognitive eating disorder symptoms is surprising, as the latter usually precede onset of behavioral symptoms.43 We might not have been able to observe those associations due to low statistical power, although other studies with smaller sample sizes have observed an association between lower socioeconomic position and these symptoms.15,19,20,21 If these findings are true, they could suggest different risk mechanisms among parental education, disordered eating, and cognitive eating disorder symptoms.
Limitations
This study has some limitations. There are high levels of attrition among ALSPAC respondents from lower socioeconomic backgrounds. This might have biased our results if those participants have differential risk of eating disorder symptoms. For example, we observed evidence of an association between income and disordered eating behaviors only at age 14. This could be explained by earlier onset of eating disorder symptoms in those from more deprived background. However, it could also be the result of missing data patterns, so replication in future studies is needed to help clarify the nature of this finding. However, in general, our results were comparable across complete record and imputed analyses.
Information from ALSPAC was collated in the 1990s and 2000s. Given the historical changes in economic contexts, such as the cost-of-living crisis, using a more recent cohort study may yield different associations than those observed in this study.
Using the data available for parental occupation and educational attainment may have introduced bias in our associations if missingness patterns relate to both the exposure and the outcome. Nearly 10% of 2-person households had data missing from 1 parent for occupation and 3.0% for education.
Our measurement of restrictive eating behaviors may not accurately capture extreme restrictive behaviors, such as typical symptoms of anorexia nervosa, as these are uncommon in general-population samples. Our outcome might have instead captured young people with more common restrictive eating behaviors, which could have different patterns of associations with socioeconomic status.
Previous research has shown that genetic susceptibility for a number of psychiatric conditions, including anorexia nervosa, is associated with increased probability of being born in more deprived environments, possibly as a result of intergenerational drift.33 Although we adjusted our models for maternal mental health difficulties in sensitivity analyses, we were unable to robustly account for the potential for genetic confounding, as polygenic risk scores currently explain limited phenotypical variance44 and the sample did not allow other genetically informed designs.
Conclusions
In this cohort study, we found that children from lower socioeconomic positions experienced greater levels of eating disorder symptoms throughout adolescence. People from more deprived backgrounds experience greater difficulties in accessing health care.45 However, eating disorders are one of the few conditions in which an association with deprivation is either not observed or reversed when using clinical registers, suggesting that there might be eating disorder–specific barriers in access to care. Individuals in lower socioeconomic groups with eating disorders could be less likely to seek help due to internalized, stigmatizing beliefs that eating disorders are a “disease of affluence”46,47 or because of differences in perceived need for treatment.48 Second, individuals with higher BMI are less likely to receive consultation for eating disorders,49 which might limit referrals for adolescents from lower socioeconomic positions who are more likely to have higher BMI.50
Identifying and addressing existing barriers that might prevent young people from deprived backgrounds from accessing eating disorder services should be research and policy priority. Provision of comprehensive medical training might facilitate identification of a broader spectrum of eating disorder presentations in primary care, particularly in populations who are more likely to be missed. Lastly, our findings add to the extensive evidence base calling for a reduction in socioeconomic inequalities as part of population-wide mental health prevention strategies.
eMethods 1. Outcome
eMethods 2. Exposures
eFigure 1. Direct Acyclic Graph hypothesizing relationship between socioeconomic indicators for our main analysis
eFigure 2. Direct Acyclic Graph hypothesizing maternal characteristics mediating the relationship between socioeconomic position indicators in childhood and eating disorder symptoms in adolescence.
eFigure 3. Direct Acyclic Graph hypothesizing different association between socioeconomic position indicators. In this graph, structural indicators (education and occupation) affect maternal indicators (income), which in turn affect perceptual indicators
eFigure 4. Direct Acyclic Graph hypothesizing maternal characteristics confounding the relationship between socioeconomic position indicators in childhood and eating disorder symptoms in adolescence.
eMethods 3. Confounders
eMethods 4. Deviation from protocol
eMethods 5. Auxiliary variables
eTable 1. Characteristics of respondents with at least one available outcome measurement or no available data on eating disorder outcomes among participants with complete exposure data
eTable 2. Frequency and means of adolescents who experience eating disorder symptoms from 14 to 18 years old for sample with complete exposures
eTable 3. Odds ratio for the association between age and eating disorder symptoms
eTable 4. Odds ratio for the association between age, age2 and eating disorder symptoms
eTable 5. Stratified odds ratio of adolescent disordered eating behaviour at age 14, 16, and 18 according to parental socieoconomic indicators
eTable 6. Multilevel logistic and linear regression models for binge eating, restrictive eating, and purging and their association with parental socioeconomic position at 14, 16, and 18
eTable 7. Multilevel logistic and linear regression models for disordered eating behaviours and weight and shape concerns at age 14, 16, and 18 according to parental socioeconomic position
eTable 8. Linear regression model for body dissatisfaction and weight and shape concerns at age 14, 16, and 18 according to parental socioeconomic position
eTable 9. Multilevel logistic and linear regression models adjusted for maternal characteristics for disordered eating behaviours at age 14, 16, and 18 according to parental socioeconomic position, adjusted for maternal characteristics
eTable 10. Linear regression models for body dissatisfaction at age 14 according to socioeconomic indicators, adjusted for maternal characteristics
eTable 11. Multilevel logistic and linear regression models for any behavioural eating disorder symptoms at age 14, 16, and 18 and weight and shape concerns at age 14 and 18 according to parental socioeconomic position
eTable 12. Linear regression model for body dissatisfaction at age 14 according to parental socioeconomic position
eTable 13. Multilevel logistic and linear regression models for disordered eating behaviours and its association to parental socioeconomic position at age 14, 16, and 18
eTable 14. Linear regression models for body dissatisfaction at age 14 according to socioeconomic indicators
Data Sharing Statement
References
- 1.McLaughlin KA, Breslau J, Green JG, et al. Childhood socio-economic status and the onset, persistence, and severity of DSM-IV mental disorders in a US national sample. Soc Sci Med. 2011;73(7):1088-1096. doi: 10.1016/j.socscimed.2011.06.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Hakulinen C, Mok PLH, Horsdal HT, et al. Parental income as a marker for socioeconomic position during childhood and later risk of developing a secondary care-diagnosed mental disorder examined across the full diagnostic spectrum: a national cohort study. BMC Med. 2020;18(1):323. doi: 10.1186/s12916-020-01794-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Joseph Rowntree Foundation . UK poverty 2024. Published January 23, 2024. Accessed July 21, 2025. https://www.jrf.org.uk/uk-poverty-2024-the-essential-guide-to-understanding-poverty-in-the-uk
- 4.Koch SV, Larsen JT, Plessen KJ, Thornton LM, Bulik CM, Petersen LV. Associations between parental socioeconomic-, family-, and sibling status and risk of eating disorders in offspring in a Danish national female cohort. Int J Eat Disord. 2022;55(8):1130-1142. doi: 10.1002/eat.23771 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wood S, Marchant A, Allsopp M, et al. Epidemiology of eating disorders in primary care in children and young people: a Clinical Practice Research Datalink study in England. BMJ Open. 2019;9(8):e026691. doi: 10.1136/bmjopen-2018-026691 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ahrén JC, Chiesa F, Koupil I, Magnusson C, Dalman C, Goodman A. We are family–parents, siblings, and eating disorders in a prospective total-population study of 250,000 Swedish males and females. Int J Eat Disord. 2013;46(7):693-700. doi: 10.1002/eat.22146 [DOI] [PubMed] [Google Scholar]
- 7.Goodman A, Heshmati A, Koupil I. Family history of education predicts eating disorders across multiple generations among 2 million Swedish males and females. PLoS One. 2014;9(8):e106475. doi: 10.1371/journal.pone.0106475 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Lindberg L, Hjern A. Risk factors for anorexia nervosa: a national cohort study. Int J Eat Disord. 2003;34(4):397-408. doi: 10.1002/eat.10221 [DOI] [PubMed] [Google Scholar]
- 9.McClelland L, Crisp A. Anorexia nervosa and social class. Int J Eat Disord. 2001;29(2):150-156. doi: [DOI] [PubMed] [Google Scholar]
- 10.Ahrén-Moonga J, Silverwood R, Klinteberg BA, Koupil I. Association of higher parental and grandparental education and higher school grades with risk of hospitalization for eating disorders in females: the Uppsala birth cohort multigenerational study. Am J Epidemiol. 2009;170(5):566-575. doi: 10.1093/aje/kwp166 [DOI] [PubMed] [Google Scholar]
- 11.Mulders-Jones B, Mitchison D, Girosi F, Hay P. Socioeconomic correlates of eating disorder symptoms in an Australian population-based sample. PLoS One. 2017;12(1):e0170603. doi: 10.1371/journal.pone.0170603 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Huryk KM, Drury CR, Loeb KL. Diseases of affluence: a systematic review of the literature on socioeconomic diversity in eating disorders. Eat Behav. 2021;43:101548. doi: 10.1016/j.eatbeh.2021.101548 [DOI] [PubMed] [Google Scholar]
- 13.Mikhail ME, Carroll SL, Clark DA, O’Connor S, Burt SA, Klump KL. Context matters: neighborhood disadvantage is associated with increased disordered eating and earlier activation of genetic influences in girls. J Abnorm Psychol. 2021;130(8):875-885. doi: 10.1037/abn0000719 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Power Y, Power L, Canadas MB. Low socioeconomic status predicts abnormal eating attitudes in Latin American female adolescents. Eat Disord. 2008;16(2):136-145. doi: 10.1080/10640260801887261 [DOI] [PubMed] [Google Scholar]
- 15.Deleel ML, Hughes TL, Miller JA, Hipwell A, Theodore LA. Prevalence of eating disturbance and body image dissatisfaction in young girls: an examination of the variance across racial and socioeconomic groups. Psychol Sch. 2009;46(8):767-775. doi: 10.1002/pits.20415 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Larson N, Loth KA, Eisenberg ME, Hazzard VM, Neumark-Sztainer D. Body dissatisfaction and disordered eating are prevalent problems among U.S. young people from diverse socioeconomic backgrounds: findings from the EAT 2010-2018 study. Eat Behav. 2021;42:101535. doi: 10.1016/j.eatbeh.2021.101535 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Larsen PS, Strandberg-Larsen K, Olsen EM, Micali N, Nybo Andersen AM. Parental characteristics in association with disordered eating in 11- to 12-year-olds: a study within the Danish National Birth Cohort. Eur Eat Disord Rev. 2018;26(4):315-328. doi: 10.1002/erv.2599 [DOI] [PubMed] [Google Scholar]
- 18.Micali N, De Stavola B, Ploubidis G, Simonoff E, Treasure J, Field AE. Adolescent eating disorder behaviours and cognitions: gender-specific effects of child, maternal and family risk factors. Br J Psychiatry. 2015;207(4):320-327. doi: 10.1192/bjp.bp.114.152371 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Paxton SJ, Eisenberg ME, Neumark-Sztainer D. Prospective predictors of body dissatisfaction in adolescent girls and boys: a five-year longitudinal study. Dev Psychol. 2006;42(5):888-899. doi: 10.1037/0012-1649.42.5.888 [DOI] [PubMed] [Google Scholar]
- 20.Parkinson KN, Drewett RF, Le Couteur AS, Adamson AJ; Gateshead Millennium Study core team . Earlier predictors of eating disorder symptoms in 9-year-old children: a longitudinal study. Appetite. 2012;59(1):161-167. doi: 10.1016/j.appet.2012.03.022 [DOI] [PubMed] [Google Scholar]
- 21.Rosenqvist E, Konttinen H, Berg N, Kiviruusu O. Development of body dissatisfaction in women and men at different educational levels during the life course. Int J Behav Med. 2024;31(5):718-729. doi: 10.1007/s12529-023-10213-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Nicholls D, Statham R, Costa S, Micali N, Viner RM. Childhood risk factors for lifetime bulimic or compulsive eating by age 30 years in a British national birth cohort. Appetite. 2016;105:266-273. doi: 10.1016/j.appet.2016.05.036 [DOI] [PubMed] [Google Scholar]
- 23.Swanson SA, Field AE. Commentary: considerations for the use of registry data to study adolescent eating disorders. Int J Epidemiol. 2016;45(2):488-490. doi: 10.1093/ije/dyw084 [DOI] [PubMed] [Google Scholar]
- 24.Solmi M, Radua J, Olivola M, et al. Age at onset of mental disorders worldwide: large-scale meta-analysis of 192 epidemiological studies. Mol Psychiatry. 2022;27(1):281-295. doi: 10.1038/s41380-021-01161-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Allen KL, Byrne SM, Oddy WH, Crosby RD. Early onset binge eating and purging eating disorders: course and outcome in a population-based study of adolescents. J Abnorm Child Psychol. 2013;41(7):1083-1096. doi: 10.1007/s10802-013-9747-7 [DOI] [PubMed] [Google Scholar]
- 26.Boyd A, Golding J, Macleod J, et al. Cohort profile: the ‘children of the 90s’—the index offspring of the Avon Longitudinal Study of Parents and Children. Int J Epidemiol. 2013;42(1):111-127. doi: 10.1093/ije/dys064 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Fraser A, Macdonald-Wallis C, Tilling K, et al. Cohort profile: the Avon Longitudinal Study of Parents and Children: ALSPAC mothers cohort. Int J Epidemiol. 2013;42(1):97-110. doi: 10.1093/ije/dys066 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Kann L, Warren CW, Harris WA, et al. Youth risk behavior surveillance–United States, 1995. MMWR CDC Surveill Summ. 1996;45(4):1-84. [PubMed] [Google Scholar]
- 29.Solmi F, Mascarell MC, Zammit S, Kirkbride JB, Lewis G. Polygenic risk for schizophrenia, disordered eating behaviours and body mass index in adolescents. Br J Psychiatry. 2019;215(1):428-433. doi: 10.1192/bjp.2019.39 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Stice E. A prospective test of the dual-pathway model of bulimic pathology: mediating effects of dieting and negative affect. J Abnorm Psychol. 2001;110(1):124-135. doi: 10.1037/0021-843X.110.1.124 [DOI] [PubMed] [Google Scholar]
- 31.Shisslak CM, Renger R, Sharpe T, et al. Development and evaluation of the McKnight Risk Factor Survey for assessing potential risk and protective factors for disordered eating in preadolescent and adolescent girls. Int J Eat Disord. 1999;25(2):195-214. doi: [DOI] [PubMed] [Google Scholar]
- 32.Gregg P, Propper C, Washbrook E. Understanding the relationship between parental income and multiple child outcomes: a decomposition analysis. Published January 2008. Accessed July 16, 2025. https://www.bristol.ac.uk/media-library/sites/cmpo/migrated/documents/wp193.pdf
- 33.Solmi F, Lewis G, Zammit S, Kirkbride JB. Neighborhood characteristics at birth and positive and negative psychotic symptoms in adolescence: findings from the ALSPAC Birth Cohort. Schizophr Bull. 2020;46(3):581-591. doi: 10.1093/schbul/sbz049 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Laaksonen M, Rahkonen O, Karvonen S, Lahelma E. Socioeconomic status and smoking: analysing inequalities with multiple indicators. Eur J Public Health. 2005;15(3):262-269. doi: 10.1093/eurpub/cki115 [DOI] [PubMed] [Google Scholar]
- 35.Johnson JG, Cohen P, Kasen S, Brook JS. Childhood adversities associated with risk for eating disorders or weight problems during adolescence or early adulthood. Am J Psychiatry. 2002;159(3):394-400. doi: 10.1176/appi.ajp.159.3.394 [DOI] [PubMed] [Google Scholar]
- 36.Tabler J, Utz RL. The influence of adolescent eating disorders or disordered eating behaviors on socioeconomic achievement in early adulthood. Int J Eat Disord. 2015;48(6):622-632. doi: 10.1002/eat.22395 [DOI] [PubMed] [Google Scholar]
- 37.Carey EG, Ridler I, Ford TJ, Stringaris A. Editorial perspective: when is a ‘small effect’ actually large and impactful? J Child Psychol Psychiatry. 2023;64(11):1643-1647. doi: 10.1111/jcpp.13817 [DOI] [PubMed] [Google Scholar]
- 38.Yilmaz Z, Gottfredson NC, Zerwas SC, Bulik CM, Micali N. Developmental premorbid body mass index trajectories of adolescents with eating disorders in a longitudinal population cohort. J Am Acad Child Adolesc Psychiatry. 2019;58(2):191-199. doi: 10.1016/j.jaac.2018.11.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Hazzard VM, Hooper L, Larson N, Loth KA, Wall MM, Neumark-Sztainer D. Associations between severe food insecurity and disordered eating behaviors from adolescence to young adulthood: findings from a 10-year longitudinal study. Prev Med. 2022;154:106895. doi: 10.1016/j.ypmed.2021.106895 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Abene JA, Tong J, Minuk J, Lindenfeldar G, Chen Y, Chao AM. Food insecurity and binge eating: a systematic review and meta-analysis. Int J Eat Disord. 2023;56(7):1301-1322. doi: 10.1002/eat.23956 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Bidopia T, Carbo AV, Ross RA, Burke NL. Food insecurity and disordered eating behaviors in children and adolescents: a systematic review. Eat Behav. 2023;49:101731. doi: 10.1016/j.eatbeh.2023.101731 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Larsen JT, Munk-Olsen T, Bulik CM, et al. Early childhood adversities and risk of eating disorders in women: a Danish register-based cohort study. Int J Eat Disord. 2017;50(12):1404-1412. doi: 10.1002/eat.22798 [DOI] [PubMed] [Google Scholar]
- 43.Stice E, Gau JM, Rohde P, Shaw H. Risk factors that predict future onset of each DSM-5 eating disorder: Predictive specificity in high-risk adolescent females. J Abnorm Psychol. 2017;126(1):38-51. doi: 10.1037/abn0000219 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Pingault JB, Allegrini AG, Odigie T, et al. Research review: how to interpret associations between polygenic scores, environmental risks, and phenotypes. J Child Psychol Psychiatry. 2022;63(10):1125-1139. doi: 10.1111/jcpp.13607 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Bristow K, Edwards S, Funnel E, et al. Help seeking and access to primary care for people from “hard-to-reach” groups with common mental health problems. Int J Family Med. 2011;2011:490634. doi: 10.1155/2011/490634 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Bruch H. Eating Disorders: Obesity, Anorexia Nervosa, and the Person Within. Basic Books; 1974. [Google Scholar]
- 47.Erickson MC, Mellinger C, Park B, Dimidjian S. Stereotypes about who is affected by eating disorders disadvantage risk perception for Black girls and women. Sex Roles. 2024;90(8):1075-1098. doi: 10.1007/s11199-024-01496-9 [DOI] [Google Scholar]
- 48.Sonneville KR, Lipson SK. Disparities in eating disorder diagnosis and treatment according to weight status, race/ethnicity, socioeconomic background, and sex among college students. Int J Eat Disord. 2018;51(6):518-526. doi: 10.1002/eat.22846 [DOI] [PubMed] [Google Scholar]
- 49.MacCaughelty C, Wagner R, Rufino K. Does being overweight or male increase a patient’s risk of not being referred for an eating disorder consult? Int J Eat Disord. 2016;49(10):963-966. doi: 10.1002/eat.22556 [DOI] [PubMed] [Google Scholar]
- 50.Calzo JP, Sonneville KR, Haines J, Blood EA, Field AE, Austin SB. The development of associations among body mass index, body dissatisfaction, and weight and shape concern in adolescent boys and girls. J Adolesc Health. 2012;51(5):517-523. doi: 10.1016/j.jadohealth.2012.02.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods 1. Outcome
eMethods 2. Exposures
eFigure 1. Direct Acyclic Graph hypothesizing relationship between socioeconomic indicators for our main analysis
eFigure 2. Direct Acyclic Graph hypothesizing maternal characteristics mediating the relationship between socioeconomic position indicators in childhood and eating disorder symptoms in adolescence.
eFigure 3. Direct Acyclic Graph hypothesizing different association between socioeconomic position indicators. In this graph, structural indicators (education and occupation) affect maternal indicators (income), which in turn affect perceptual indicators
eFigure 4. Direct Acyclic Graph hypothesizing maternal characteristics confounding the relationship between socioeconomic position indicators in childhood and eating disorder symptoms in adolescence.
eMethods 3. Confounders
eMethods 4. Deviation from protocol
eMethods 5. Auxiliary variables
eTable 1. Characteristics of respondents with at least one available outcome measurement or no available data on eating disorder outcomes among participants with complete exposure data
eTable 2. Frequency and means of adolescents who experience eating disorder symptoms from 14 to 18 years old for sample with complete exposures
eTable 3. Odds ratio for the association between age and eating disorder symptoms
eTable 4. Odds ratio for the association between age, age2 and eating disorder symptoms
eTable 5. Stratified odds ratio of adolescent disordered eating behaviour at age 14, 16, and 18 according to parental socieoconomic indicators
eTable 6. Multilevel logistic and linear regression models for binge eating, restrictive eating, and purging and their association with parental socioeconomic position at 14, 16, and 18
eTable 7. Multilevel logistic and linear regression models for disordered eating behaviours and weight and shape concerns at age 14, 16, and 18 according to parental socioeconomic position
eTable 8. Linear regression model for body dissatisfaction and weight and shape concerns at age 14, 16, and 18 according to parental socioeconomic position
eTable 9. Multilevel logistic and linear regression models adjusted for maternal characteristics for disordered eating behaviours at age 14, 16, and 18 according to parental socioeconomic position, adjusted for maternal characteristics
eTable 10. Linear regression models for body dissatisfaction at age 14 according to socioeconomic indicators, adjusted for maternal characteristics
eTable 11. Multilevel logistic and linear regression models for any behavioural eating disorder symptoms at age 14, 16, and 18 and weight and shape concerns at age 14 and 18 according to parental socioeconomic position
eTable 12. Linear regression model for body dissatisfaction at age 14 according to parental socioeconomic position
eTable 13. Multilevel logistic and linear regression models for disordered eating behaviours and its association to parental socioeconomic position at age 14, 16, and 18
eTable 14. Linear regression models for body dissatisfaction at age 14 according to socioeconomic indicators
Data Sharing Statement
