Dietary Consumption Among Youth with Antipsychotic-Induced Weight Gain and Changes Following Healthy Lifestyle Education

Abstract

Background:

Youth treated with antipsychotic medications are high risk for weight gain, increased lipids/glucose, and development of metabolic syndrome. Little is known about the dietary intake/nutritional adequacy in this vulnerable population, and effect on weight gain. This secondary data analysis describes the baseline intake and changes in diet after receiving healthy lifestyle education/counseling over 6 months, in a sample of youth with antipsychotic-induced weight gain.

Methods:

The U.S. Department of Agriculture (USDA) Automated Multiple-Pass Method 24-hour dietary recall was administered to 117 youth at baseline, 3 months, and 6 months. Parent/child received personalized healthy lifestyle education sessions over 6 months. Baseline intake was compared with the USDA Recommended Daily Allowance using independent samples t-tests. Individual dietary covariates were examined for change over 6 months using longitudinal linear mixed modeling. Influence of each on body mass index (BMI) z-score change was tested in a pooled group analysis and then compared by treatment group.

Results:

Pooled analysis revealed baseline consumption high in carbohydrates, fat, protein, sugar, and refined grains, while low in fruit/vegetables, whole grains, fiber, and water. Change over 6 months demonstrated a statistically significant decrease in daily calories (p = 0.002), carbohydrates (p = 0.003), fat (p = 0.012), protein (p = 0.025), sugar (p = 0.008), refined grains (p = 0.008), total dairy (p = 0.049), and cheese (p = 0.027). Small increases in fruits/vegetables were not statistically significant, although the Healthy Eating Index subscores for total vegetables (p = 0.013) and dark green/orange vegetables (p = 0.034) were. No dietary covariates were predictors of change in BMI z-score. Nondietary predictors were parent weight/BMI and treatment group, with the metformin and switch groups experiencing significant decreases in BMI z-score.

Conclusions:

Further pediatric studies are necessary to assess the effects of antipsychotic medications on dietary intake, and test efficacy of healthy lifestyle interventions on change in nutrition. The relationship of nutrition to cardiometabolic health in this population must be further investigated.

Clinical Trial Registration number: NCT 02877823.

Introduction

Second-generation antipsychotic (SGA) medication treatment for pediatric serious mental illness has increased significantly over the past 15 years, with availability of newer agents and expanded pediatric research to guide the use of these medications in children and adolescents (Olfson et al. 2012; Olfson et al. 2015; Olfson et al. 2015). Food and Drug Administration-approved indications for pediatric SGA treatment include schizophrenia, bipolar mania, and irritability associated with autism. However, SGAs are used at least as often for “off label” treatment of severe irritability and aggression. Multiple pediatric studies have demonstrated SGA treatment to be quite efficacious in children/adolescents with autism, bipolar disorder, disruptive mood dysregulation disorder, and childhood schizophrenia by reducing debilitating psychiatric symptoms, resulting in sustained periods of psychiatric stability and improved functioning (Findling et al. 2008, 2010a, 2010b; Owen et al. 2009). However, these medications are also unfortunately associated with serious cardiometabolic side effects, including weight gain, dyslipidemias, and increased insulin resistance, leading to the development of metabolic syndrome, type 2 diabetes mellitus (T2DM), and premature cardiovascular disease (Deng 2013; Christian et al. 2015).

Of concern is the vulnerability to cardiometabolic side effects specifically in youth compared with adults. Side effects can develop early in treatment with children often experiencing increased triglycerides and weight gain >10% of their baseline weight within the first 3 months (Correll et al. 2009). They also can experience increased fasting triglycerides, low-density lipoprotein cholesterol, glucose, and insulin even without weight gain, making these youth particularly at risk for early onset of T2DM and cardiovascular disease (Correll et al. 2009; Almandil et al. 2013; Arango et al. 2014). Multiple studies in children have found decreased insulin sensitivity and increased fasting glucose within weeks of SGA treatment initiation, increasing the risk of T2DM by 50% in this population (Rubin et al. 2015). A meta-analysis found the incidence of T2DM in SGA treated youth to be three times higher compared with healthy controls, and psychiatric controls unexposed to SGAs (Galling et al. 2016).

Increased appetite and thirst are common side effects experienced with SGA treatment, resulting in excess food and beverage consumption. SGAs are thought to cause dysregulation of appetite hormones leptin and ghrelin, with resultant increased appetite and reduced satiety (Zhang et al. 2013; Lu et al. 2015). Additionally, increased triglycerides create a systemic inflammatory response, which contributes to increased appetite and fat storage (Fonseka et al. 2016). Adiponectin, which reflects nutritional and lifestyle health, and is a known predictor of metabolic syndrome, is considered a critical biomarker for determining metabolic health, although it is not routinely measured in clinical care (Kishida et al. 2014). Recent studies have begun to examine biomarkers, such as adiponectin, leptin, ghrelin, orexin, and insulin. A systematic review and meta-analysis of such studies in antipsychotic-naive patients (N = 1792) experiencing first-episode psychosis (FEP), compared with healthy controls (N = 1394), found appetite regulation to be impaired in the FEP group before starting an SGA (Błażej et al. 2019). This study found the FEP group to have higher insulin levels and lower leptin levels compared with the healthy controls, indicating that there may be dietary/lifestyle factors experienced by those with FEP, which makes them more vulnerable to weight gain associated with SGA treatment. Another review examined studies (included men, women, and children) testing specific dietary patterns and the effects on serum adiponectin levels, finding the Mediterranean diet, as well as diets high in low glycemic foods and fiber to be positively associated with increases in adiponectin (Izadi and Azadbakht 2015). These findings emphasize the need for future dietary studies to go beyond weight change as the primary outcome and to dig deeper by including metabolomic measures to evaluate nutritional intake.

There are many pediatric studies focused on identification of SGA side effects, but none have examined dietary intake and the effect on weight gain. Although there are several weight loss studies testing pharmacological interventions (e.g., metformin, topiramate) in youth treated with SGAs, none described dietary intake or change following an intervention (Klein et al. 2006; Arman et al. 2008; Zheng et al. 2015; Anagnostou et al. 2016; Correll et al. 2016). Several recent studies have examined healthy lifestyle strategies with youth receiving SGAs (Caemmerer et al. 2012), but only one study assessed the dietary intake at baseline and change over 12 weeks (Teasdale et al. 2016). The study design was cluster controlled with both treatment and control groups, but results were reported on only the treatment group, finding statistically significant decreases in calories, carbohydrates, protein, and discretionary foods, and increased vegetable intake (Teasdale et al. 2016). Given the impact of SGAs on appetite and metabolism, and the gap in the literature around dietary intake associated with SGA treatment, it is particularly important to conduct further research on dietary intake and nutritional composition. SGAs are used for chronic conditions; therefore, it is important to identify targets for dietary intervention as part of side effect management strategies. Research on healthy lifestyle strategies has also largely focused on obesity outcomes (e.g., weight classification, body mass index [BMI] percentile, BMI z-score) rather than investigating if children are meeting recommended nutrition for growth and development. Although antipsychotic-treated children are known to consume more calories, it is unclear if they have adequate nutrition. Given the challenges parents experience feeding children with serious mental illness (sensitivity to food taste and texture, parent–child conflict over food choices) and cost burden associated with increased food demand, children may not be consuming either the variety or amount of specific foods needed to meet nutrient targets. Additionally, studies examining dietary intake in relation to metabolomic biomarkers are needed to better understand the influence of specific dietary factors on nutritional and metabolic health. More comprehensive knowledge will lead to the development of dietary recommendations for prevention and/or mitigation of cardiometabolic side effects and ensure adequate nutrition.

There are numerous acceptable methods for dietary assessment, each with strengths and limitations, and subject to reporting bias and underreporting. Commonly used methods to quantify dietary intake of individuals are the food record and 24-hour dietary recall, both providing a valid measure of usual intake when collected over multiple days, although when only single-day recalls are collected on multiple different individuals, it will provide an accurate measure of intake for a group or population (Gibson and Rosalind 2005). The food frequency questionnaires identify patterns of intake and provide estimates of types of foods commonly consumed over longer periods of time, rather than a quantitative measure of usual daily intake. The U.S. Department of Agriculture (USDA) Automated Multiple-Pass Method (AMPM) 24-hour dietary recall has been considered the optimal method for accurately estimating usual daily intake in that it incorporates multiple steps for reporting intake, utilizes standardized interview methods and food models for accurate measurement, and report of total energy intake has been validated with the doubly labeled water method, finding <3% of underreporting (Moshfegh et al. 2008). A more recent computerized 24-hour recall method being utilized in dietary research is the Automated Self-Administered 24-hour recall (ASA24), which is a web-based free access tool available to researchers and their participants by the National Cancer Institute. It incorporates the multiple pass method and unique standardized aspects of the USDA AMPM, but additionally provides coding and analysis of foods/beverages/supplements reported and generates a comprehensive data set of 65 nutrients and 37 food groups (Suber et al. 2012). Several large-scale feasibility studies were conducted comparing the ASA24 to the USDA AMPM, finding equivalency in estimating intake of nutrients and food groups (Thompson et al. 2015). New technologies measuring intake in real time have become attractive to researchers as holding potential for increasing reporting accuracy. A recent review identified 43 unique technology-based tools being tested such as smart phone-based apps, photoimaging, and smart watches to capture ecological momentary assessment of dietary intake and physical activity (Eldridge et al. 2019).

The Improving Metabolic Parameters of Antipsychotic Child Treatment Study (IMPACT), a National Institute of Mental Health (NIMH)-funded multisite randomized control trial (RCT), tested the effectiveness of two pharmacological interventions (add metformin to current SGA or switch to a lower risk antipsychotic) on weight loss for children with SGA-induced weight gain (Reeves et al. 2013). This study found a statistically significant reduction in the BMI z-scores for both intervention groups, with addition of metformin being superior, and no significant changes were found in the control group (Correll et al. 2020). All participants of each study condition also received healthy lifestyle education (HLE). A personalized parent–child-focused approach was utilized to deliver nutrition and physical activity education/counseling to the child and parent based on the American Medical Association (AMA) 2007 Healthy Lifestyle Guidelines (Barlow 2001). HLE was reviewed at each study visit over the duration of subject participation, and dietary intake was measured at baseline, 3-month, and 6-month study visits with standardized methods. The purpose of this present supplemental secondary data analysis was to (AIM 1) describe the baseline dietary intake of the sample and compare with the USDA Recommended Daily Allowance (RDA), (AIM 2) examine changes in dietary intake over 6 months to determine participant integration of AMA healthy lifestyle guidelines, and (AIM 3) identify if specific dietary and nondietary variables were predictors of changes in BMI z-score. It is hypothesized that participants will demonstrate improved dietary intake, specifically by increased fruit/vegetable consumption and reduced sugar intake. It is also predicted that participants taking metformin will demonstrate a significant reduction in carbohydrate intake. Furthermore, it is hypothesized that reduced sugar intake is associated with reduced BMI z-score.

Methods

Sample

The sample includes 117 youth aged 8–18 years (at the time of enrollment), all treated with an SGA for at least 2 months before IMPACT study enrollment, resulting in clinically significant weight gain, and a BMI at the 85th percentile or higher. Additionally, cognitive screening was administered to ensure the child participant's ability to comprehend and engage in HLE. A full list of inclusion criteria can be found in the IMPACT methodology article (Reeves et al. 2013). Community psychiatric providers referred participants to one of four sites located at urban major medical/education centers in Maryland, North Carolina, and New York.

Design

IMPACT participants were randomized to one of three study arms (metformin, switch, control) with the primary outcome being weight loss (decreased BMI z-score) based on pharmacological interventions. All participants continued SGA treatment throughout the study, with the metformin and control groups remaining on their baseline antipsychotic, and the switch group changed to aripiprazole or perphenazine (if currently on or had an unsuccessful previous trial with aripiprazole), both of which are considered lower risk for weight gain. Although change in diet was not a primary outcome, all participants and their parent received HLE based on the AMA Healthy Lifestyle Guidelines 2007 (Barlow 2001) over 6 months, and dietary assessments were conducted at baseline, 3-month, and 6-month study visits. Detailed information related to SGA dosing/cross titration can be found in the IMPACT study primary outcomes publication (Correll et al. 2020). The institutional review board at each site approved the IMPACT study, and the University of Maryland Baltimore approved the current secondary data analysis.

Interventions

Healthy lifestyle education

HLE was based on the 2007 Stage 1 AMA Healthy Lifestyle Guidelines, which emphasizes intake of five fruit/vegetable servings per day, eating breakfast daily, eating meals together as a family at least five times per week, eliminating sugar sweetened beverages, reducing portion sizes, limiting foods/beverages high in sugar and fat, limiting food from restaurants/fast food to only once per week, 60 minutes of physical activity per day, and <2 hours of screen time daily (Rao 2020). All participants received this education at each of the nine study visits scheduled for weeks 1, 2, 4, 6, 8, 12, 16, 20, and 24. Although physical activity and screen time were assessed, and included in HLE, they are not addressed in this analysis. At the baseline visit, participant lifestyle behaviors were assessed by reviewing each guideline with the child and parent to determine current status and identify areas in need of improvement. Based on those findings, a personalized plan was developed with the youth and parent that included specific lifestyle goals, and strategies for attainment. AMA Stages 2 and 3 were implemented with participants who gained >7% and 10% of their baseline weight during the study. All research team members were trained by a certified diabetes educator nurse practitioner on teaching AMA Healthy Lifestyle Guidelines at study start-up, and each study site coordinator was responsible for continued oversight of team members' implementation and documentation of guidelines, and training new team members. Concerns regarding individual participants, specifically those requiring Stage 2 and 3 interventions, were discussed on the weekly study team meeting, which included all sites and principal investigators. Further detail on AMA stages and training fidelity can be found in the IMPACT Study methodology article (Reeves et al. 2013).

Measures

Demographics

Demographic information used in this study was obtained by the parent interview at the screening visit, which occurred 2 weeks before the baseline visit to determine the eligibility in the IMPACT study. Variables included in this analysis are youths' age, sex, race, household income, psychiatric diagnosis, antipsychotic medication, and number of months on SGA.

Anthropometry

Height and weight were measured using standardized procedures and equipment at the General Clinical Research Center of each site at the screening, baseline, 3-month, and 6-month study visits. All participants were fasting ≥8 hours, and weight was measured with the Tanita TBF-300 scale (participant in lightweight clothing), and height was measured three times using the Seca 264 digital stadiometer (average calculated). Parent height and weight were also measured at baseline, 3-month, and 6-month study visits using the same equipment and standardized procedures. Child BMI percentile (to determine eligibility at screening) and z-scores were determined using the Baylor School of Medicine Children's Nutrition Research Center BMI percentile for age calculator (https://www.bcm.edu/cnrc-apps/bodycomp/bmiz2.html), and parent BMI with the Centers for Disease Control and Prevention (CDC) BMI calculator for Adults (https://www.cdc.gov/healthyweight/assessing/bmi/adult_bmi/english_bmi_calculator/bmi_calculator.html).

Twenty-four-hour dietary recall

The USDA AMPM was used to collect 24-hour dietary recalls of all food, beverage, and supplement intake at the baseline, 3-month, and 6-month study visits. The recalls were conducted directly with the youth, although the parent was present to provide details related to food preparation, ingredients, amounts served/eaten, and other details the youth may have not been able to recall or did not know. The AMPM is a computerized structured interview instrument developed in 1999 by the USDA Food Surveys Research Group to establish standardized research methodology for assessment of high-quality dietary data. The dietary recall has multiple steps and standardized probes for forgotten foods and eating occasion, as well as visual aids provided in a food model booklet for precise determination of portion size. Other unique features are the extensive food lists, which include ethnic foods, coding of food for comprehensive data analysis, and features that only could be accomplished with computer software (Raper et al. 2004; Blanton et al. 2006; Moshfegh et al. 2008). The AMPM software generates a comprehensive data set of 64 nutrients and MyPyramid Equivalent Food Groups based on the USDA Food and Nutrient Database for Dietary Studies (USDA 2008). This standardized method, considered the “gold standard” measure of dietary intake, has been utilized since 2002 by the CDC National Health and Nutrition Examination Survey (NHANES), administered directly to adults, proxy assisted with children aged 6–11 years and by proxy in children younger than 6 years (CDC 2019).

Specific study team members were selected to conduct dietary recalls at each site, and all attended a week-long training, provided by a registered dietician USDA trainer, on administration of the AMPM consistent with NHANES procedures. This writer attended the training and provided ongoing oversight on collection of dietary intake data and fidelity monitoring, which involved three supervised intakes to ensure standardized procedures. The USDA trainer served as an ongoing consultant over the course of the study and assisted in addressing any procedural challenges, and preparation of the dietary data for analysis.

Healthy Eating Index

The Healthy Eating Index (HEI) is a method of quantifying diet quality based on standards established by the USDA Center for Nutrition Policy and Promotion to assess the U.S. population wide nutritional quality and conformance with the Dietary Guidelines for Americans. The IMPACT study was initiated in 2009 while the HEI 2005 version was in use (rather than the current HEI 2015 version); therefore, this analysis is based on the HEI 2005. The HEI 2005 utilizes food group standards found in the MyPyramid recommendations and a density-based approach for scoring (amount per 1000 calories). The HEI includes subscores for 12 components and a total HEI composite score rating overall diet quality (≤50 poor, 51–79 needs improvement, ≥80 good). The HEI can be used to assess diet quality of the population, specific groups, individuals, and to test the effectiveness of dietary interventions (Guenther et al. 2007). The HEI 2005 has been tested in children aged ≥2 years and adults (n = 8650) and found to be a valid and reliable measure of diet quality with good internal consistency (Cronbach's coefficient α = 0.43) (Guenther et al. 2007).

AMA healthy lifestyle knowledge questionnaire

This questionnaire was developed by the IMPACT team and is composed of questions based on the AMA guidelines reviewed with the parent–child at each visit. Validity and reliability of this measure has not been tested since it was created specifically for this study to test knowledge rather than actual behavior. The questionnaire is composed of 10 questions related to diet and physical activity, all with dichotomous outcomes of correct or incorrect. Only the five questions pertaining to diet were examined for this analysis.

Analysis

Baseline statistics were conducted with the whole sample and included categorical/continuous demographic variables, dietary intake of macronutrients/food groups, and HEI total/subscores. Continuous variables were examined for range of values, mean/standard deviation, and presence of outliers, and normality was assessed with skew/kurtosis values and histograms. Missingness was determined by conducting the Little's Missing Completely at Random (MCAR) test, and an intraclass correlation coefficient (ICC) was calculated for each continuous dietary variable to determine the need for linear mixed model (LMM) procedures. Null models of the dependent variable (BMI z-score) were conducted, first with fixed and random intercepts, then with random slopes, to determine the best fit to the data based on the lowest Akaike information criteria (AIC) value. Fit was further determined by testing covariance structures and comparing effects on AIC values. Analyses were conducted with SPSS version 25.

Description of sample and dietary intake

Baseline demographic characteristics and dietary intake were determined, followed by comparison of the samples' baseline nutrition to the USDA dietary RDA by age and gender subcategories to examine nutritional adequacy. Significance of difference was determined with calculation of independent t-statistics and p-values. To calculate the t-statistic (mean difference/standard error), the variance of the RDA was assumed to be the same as the sample.

Changes in dietary intake over time

Longitudinal LMM procedures were used to explore the following changes over 6 months: (1) individual macronutrient and food components, (2) bivariate association of each continuous dietary covariate with change in BMI z-score, and (3) interaction of change in individual continuous dietary covariates and nondietary continuous/categorical variables (i.e., demographic variables), with change in BMI z-score over 6 months. Interaction of change within each covariate with change in BMI z-score over time was further examined by creating bivariate models, which included the individual covariate, time, and interaction between the covariate and time. Additionally, comparison of change in dietary intake was also examined by pharmacological treatment group. The mean and standard deviation for each dietary variable was calculated at baseline, 3 months, and 6 months by treatment group for comparison. Additionally, individual LMM were conducted to determine the significance of change for each variable by group.

Results

The sample was composed of 117 youth aged 8–19 years, with a majority being male (64%) and white (53%). The mean age was 13.5 years, evenly split between those aged 8–12 years (43.6%) and 13–17 years (43.6%), with the remaining 18–19 years (12.8%). The mean time period of SGA treatment before study enrollment was 21 months for treatment of disruptive mood dysregulation disorder (62%), bipolar disorder (23%), schizophrenia (10%), and psychotic depression (5%). Comorbid psychiatric diagnoses included the following: attention-deficit/hyperactivity disorder (35%), autism (26%), anxiety disorders (25%), and oppositional defiant disorder (21%). The current antipsychotic treatments were primarily aripiprazole (46%) and risperidone (39%). Socioeconomic status was evenly distributed over a range of <20K to >100K, with the majority being middle income 40–60K. All participants were overweight or obese (≥85th BMI percentile), with a mean BMI z-score of 2.1 (standard deviation = 0.049). Table 1 reflects baseline characteristics of the sample. Total missing data over 6 months were 13.3%, with strong evidence that the data were missing completely at random based on Little's MCAR test (p = 1). Much of this can be accounted for by participants not completing a dietary recall at one of the three assessment time points.

Table 1.

Demographic Description of Sample

Variable	N	%	$\bar{X}$ (σ)	Range	Median
Sex
Male	75	64
Female	42	36
Age, years			13.5 (3.28)	8–19 years	13.00
8–13	61	52
14–18	50	43
19	6	5
Race
White	62	53
Black	32	27
Asian	2	2
Indian/Alaskan	1	1
HA/PA Islander	1	1
>1 Race	19	16
Ethnicity
Hispanic	15	13
Not Hispanic	102	87
Income
<20K	21	18
20–40K	22	19
40–60K	27	23
60–80K	15	13
80–100K	15	13
>100K	16	14
Missing	1
Psychiatric diagnosis
Bipolar disorder	27	23
Severe mood dysregulation	72	62
Schizophrenia	12	10
Major depressive disorder	6	5
Antipsychotic medication
Aripiprazole	54	46
Risperidone	45	39
Ziprasidone	9	8
Olanzapine	5	4
Quetiapine	4	3
Months on SGA medication			21.6 (20.6)	2–107	15
Child percent overweight			35.5% (11.5)	12–61%	34.8
Child BMI z-score			2.01 (.492)	0.93–3.1	2.07
Parent BMI	62		32.9 (8.3)	17.4–58.9	31.6
Missing	55

BMI, body mass index; HA/PA, Hawaiian or Pacific Islander; SGA, second-generation antipsychotic.

Description of sample and dietary intake

Examination of the samples' dietary intake revealed only one participant having a total HEI score within the good range (80–100), 55% needing improvement (50–79), and 44% with poor (<50) quality nutrition. Mean values of macronutrients and food groups were high in carbohydrates, protein, fat, sugar, and refined grains, while deficient in fruits, vegetables (particularly dark green and orange), whole grains, water, and fiber. The mean caloric intake was 2323, with most participants (66%) consuming between 1000 and 3000 kcal daily, whereas 6.8% were <1000, 13.7% were 3001 to 4000, and 8.6% were >4000. Comparison of the samples' intake to the USDA RDA by age/gender found all groups to be significantly deficient in vegetables, whole grains, water, and fiber, while consuming significantly excess carbohydrates, protein, fat, and sugar (Table 2). Subgroups with the poorest nutrition were boys aged 8–13 years and girls aged 14–19 years, who also had deficiencies in meat and dairy intake.

Table 2.

Intake of Food Groups by Sample Subgroups by Age and Gender Compared with the U.S. Department of Agriculture Recommendations

USDA food groups and macronutrients	RDA girls 8–13 years	Sample $\bar{X}$ (σ) N = 16	t	p	RDA boys 8–13 years	Sample $\bar{X}$ (σ) N = 44	t	p	RDA girls 14–20 years	Sample $\bar{X}$ (σ) N = 26	t	p	RDA boys 14–20 years	Sample $\bar{X}$ (σ) N = 31	t	p
Fruits (cup)	1.5	1.13 (0.85)	1.61	0.065	1.5	1.11 (1.1)	2.29	0.014	1.5	1.2 (1.2)	1.22	<0.001	2	1.3 (1.7)	2.26	0.016
Vegetables (cup)	2	1.09 (0.95)	3.64	0.001	2.5	0.72 (0.73)	15.8	<0.001	2.5	0.69 (0.63)	14.03	<0.001	3	1.6 (1.4)	5.49	<0.001
Refined grains (oz)	5	6.13 (4.9)	−0.889	0.195	6	7.62 (4.5)	−2.33	0.012	6	6.0 (5.4)	−0.019	0.493	8	8.7 (6.2)	−0.619	0.270
Whole grains (oz) (50% of total grains)	3	1.71 (2.1)	2.39	0.016	3	0.67 (1.0)	15.1	<0.001	3	1.5 (2.8)	2.62	0.008	4	0.88 (1.2)	14.2	<0.001
Meat/poultry/seafood/eggs/soy/beans/nuts/seeds (oz)	5	5.56 (4.9)	−0.441	0.333	5	3.79 (2.4)	3.27	0.001	5	3.5 (3.0)	2.45	0.011	6.5	5.49 (3.8)	1.46	0.078
Dairy (cups)	3	2.95 (2.0)	0.098	0.462	3	2.21 (1.6)	3.24	0.001	3	1.7 (2.1)	3.03	0.003	3	2.9 (2.6)	0.211	0.417
Energy (kcal)RDA based on a sedentary to moderate activity level	1400–2000	2485 (880)	−2.14	0.025	1400–2200	2138 (835)	0.481	0.317	1800–2200	1888 (1375)	1.11	0.139	2000–2800	2850 (1416)	−0.193	0.424
Carbohydrates (g)	130	311 (99)	−7.07	<0.001	130	298 (131)	8.32	<0.001	130	250 (192)	−3.05	0.003	130	348 (167)	−7.15	<0.001
Protein (g)	34	99 (56)	−4.48	<0.001	34	75 (33)	−8.04	<0.001	46	73 (48)	−2.75	0.006	52	106 (55)	−5.38	<0.001
Fat (g)	ND	97 (45)	−6.45	<0.001	ND	75 (31)	−10.7	<0.001	ND	69 (57)	−3.86	<0.001	ND	119 (72)	6.69	<0.001
Water (oz)	72	60 (24)	1.94	0.036	80	62 (27)	4.32	<0.001	80	58 (24)	4.48	<0.001	112	82 (50)	3.29	0.001
Fiber (g)	26	17 (9)	3.88	<0.001	31	14 (7)	15.7	<0.001	26	13 (11)	5.78	<0.001	38	19 (10)	10.44	<0.001
Sugar (g)		163 (58)	−7.53	<0.001		148 (80)	−7.56	<0.001		116 (94)	−3.17	0.002		144 (85)	−4.77	<0.001
Added sugar (tsp)	<10% kcal	27 (14)	−4.01	<0.001	<10% kcal	23 (15)	−4.00	<0.001	<10% kcal	17 (20)	−0.795	0.217	<10% kcal	22 (18)	−1.37	0.091
Solid fat (g)	<10% kcal	50 (27)	−4.01	<0.001	<10% kcal	48 (23)	−6.76	<0.001	<10% kcal	46 (43)	−2.50	0.01	<10% kcal	69 (46)	−4.53	<0.001

Bold indicates significance p < .05.

ND, not determined; RDA, Recommended Daily Allowance; USDA, U.S. Department of Agriculture.

Changes in dietary intake over time

Change in nutritional intake, in the pooled sample, over 6 months was examined by testing each individual dietary variable as a dependent variable with LMM (Table 3), which found significantly decreased consumption of the following nutrients over 6 months: −395 calories (p = 0.004), −51 g of carbohydrates (p = 0.007), −11 g of protein (p = 0.041), −17 g of total fat (p = 0.010), −4.5 g of saturated fat (p = 0.033), −27 g of total sugar (p = 0.008), −6 tsp of added sugar (p = 0.002), −9.6 g of solid fat (p = 0.025), and −1.6 oz of refined grains (p = 0.011). There were small insignificant increases in whole fruit by 1 oz (p = 0.49), total vegetables by 0.8 oz (p = 0.61), dark green vegetables by 0.02 oz (p = 0.49), and orange vegetables by 0.04 oz. (p = 0.22). There were statistically significant increases in HEI subscores for total vegetables from 2.6 to 5.3 oz per 1000 calories (p = 0.013) and dark green/orange vegetables from 3.2 to 3.4 oz per 1000 calories (p = 0.034). No change in the mean total HEI score or category was found over the 6 months. Healthy lifestyle knowledge scores for both parent and child participants increased slightly without statistical significance.

Table 3.

Change in Dietary Intake and Anthropometric Measures from Baseline to 6 Months by Treatment Group and Pooled Sample with Linear Mixed Model Procedures

Variable	Control group				Metformin group				Switch group				Pooled results
Variable	Baseline N = 41 $\bar{X}$ (σ)	6 months N = 31 $\bar{X}$ (σ)	Est	p	Baseline N = 42 $\bar{X}$ (σ)	6 months N = 32 $\bar{X}$ (σ)	Est	p	Baseline N = 28 $\bar{X}$ (σ)	6 months N = 20 $\bar{X}$ (σ)	Est	p	Baseline N = 111 $\bar{X}$ (σ)	6 months N = 83 $\bar{X}$ (σ)	Est	p
Energy (kcal)	2254 (1036)	1986 (888)	−139	0.208	2388 (1215)	1943 (852)	−154	0.023	2330 (1386)	1827 (1368)	−270	0.186	2323 (1190)	1931 (1000)	−195	0.004
Carbohydrate (g)	294 (129)	254 (123)	−20	0.186	311 (162)	264 (129)	−15	0.179	302 (180)	230 (182)	−37	0.152	303 (155)	252 (140)	−25	0.007
Protein (g)	87 (49)	82 (37)	−3.43	0.47	87 (49)	68 (33)	−8.23	0.012	82 (45)	74 (59)	−3.73	0.605	86 (48)	75 (42)	−5.73	0.041
Total fat (g)	84 (48)	74 (38)	−5.27	0.261	91 (53)	71 (35)	−8.14	0.040	91 (69)	70 (57)	−11.4	0.236	89 (56)	72 (42)	−8.40	0.010
Saturated fat (g)	29 (16)	27 (15)	−1.92	0.224	31 (19)	26 (15)	−2.64	0.125	31 (21)	25 (23)	−3.38	0.322	30.4 (18.5)	25.9 (17.2)	−2.66	0.033
Solid fat	51 (33)	48 (30)	−2.46	0.426	54 (37)	42 (25)	−5.37	0.077	56.1 (41.1)	43.3 (44.3)	−8.80	0.215	53.7 (36.4)	44.1 (32)	−5.09	0.025
Sugar (g)	136 (66)	105 (60)	−17.3	0.007	147 (87)	122 (64)	−8.86	0.218	141 (99)	121 (104)	−10.6	0.403	142 (82.3)	115 (74)	−12.85	0.008
Added sugar	21 (13)	15 (10)	−3.35	0.010	24 (18)	17 (11)	−2.27	0.618	20.5 (21.3)	15.8 (15.9)	−2.08	0.407	22 (17.1)	16.1 (11.9)	−2.79	0.002
Fiber (g)	16 (9)	16 (11)	0.133	0.909	16 (10)	15 (9)	−0.412	0.618	15 (9)	12 (10)	−1.46	0.251	15.5 (9.2)	15 (10)	−5.74	0.340
Water (oz)	69 (28)	63 (32)	−4.13	0.204	65 (43)	55 (32)	−3.38	0.253	64 (32)	63 (41)	1.17	0.788	66 (35)	60 (34)	−3.27	0.102
Whole grains (oz)	1.2 (2.2)	0.80 (.97)	−0.220	0.385	1.1 (1.5)	1.1 (1.7)	−0.008	0.959	0.75 (1.4)	0.16 (.31)	−0.294	0.065	1.1 (1.8)	0.78 (1.2)	−0.150	0.134
Refined grains (oz)	7.1 (5.2)	5.8 (3.7)	−0.587	0.274	7.7 (5.3)	6.5 (4.2)	−0.621	0.130	7.3 (5.7)	4.6 (4.9)	−1.27	0.091	7.4 (5.3)	5.8 (4.2)	−0.800	0.011
Total vegetables (oz)	6.3 (6.6)	10.3 (10.1)	0.234	0.047	9.0 (7.4)	6.6 (5.3)	−0.141	0.602	9.4 (11)	10.1 (9.3)	−0.016	0.914	1.0 (1.0)	1.1 (1.1)	0.034	0.610
Vegetables dark green (oz)	0.20 (.80)	0.55 (1.7)	0.026	0.402	0.24 (.80)	0.28 (1.2)	0.004	0.850	0.39 (1.3)	0.34 (.97)	−0.001	0.967	0.03 (.12)	0.05 (.17)	0.011	0.492
Vegetables orange (oz)	0.36 (1.3)	0.95 (2.0)	0.035	0.093	0.66 (2.0)	0.17 (.53)	−0.031	0.155	0.04 (.19)	1.2 (2.4)	0.070	0.050	0.05 (.19)	0.09 (.22)	0.017	0.219
Vegetables potato (oz)	1.7 (3.2)	2.6 (5.2)	0.052	0.425	2.14 (4.5)	2.4 (3.7)	0.018	0.783	3.6 (8.5)	2.6 (3.8)	−0.055	0.606	0.29 (.68)	0.32 (.55)	0.006	0.888
Fruit total (oz)	9.3 (8.9)	7.3 (6.8)	−0.117	0.244	7.8 (8.8)	11.2 (12.9)	0.171	0.203	11.8 (12.8)	10.1 (13)	−0.148	0.413	1.2 (1.3)	1.2 (1.4)	−0.009	0.904
Fruit juice (oz)	3.8 (5.4)	3.0 (4.1)	−0.038	0.560	3.6 (6.8)	4.2 (7.1)	0.028	0.695	7.1 (10.7)	3.9 (7.3)	−0.220	0.037	0.57 (.95)	0.46 (.77)	−0.056	0.203
Fruit whole (oz)	5.4 (7.5)	4.3 (6.3)	−0.03	0.323	4.2 (7.0)	7.0 (10.5)	0.142	0.212	2.6 (6.1)	6.2 (10.3)	0.025	0.862	0.60 (.86)	0.73 (1.1)	0.044	0.490
Dairy total (oz)	19.1 (15.0)	19.7 (18.7)	0.009	0.967	20.0 (17.0)	13.1 (10.0)	−0.408	0.017	18.2 (18.8)	15.4 (22.2)	−0.258	0.414	2.4 (2.1)	2.0 (2.1)	−0.214	0.106
Dairy milk (oz)	9.9 (10.2)	9.3 (9.8)	−0.034	0.797	9.5 (10.8)	7.6 (7.2)	−0.130	0.226	7.5 (8.2)	9.5 (14)	0.073	0.129	1.1 (1.2)	1.1 (1.3)	−0.041	0.615
Dairy cheese (oz)	8.6 (10.1)	9.5 (13.7)	0.010	0.952	9.6 (14.5)	4.7 (6.4)	−0.303	0.060	10.3 (14.5)	5.7 (12.5)	−0.340	0.129	1.28 (1.6)	0.84 (1.4)	−0.187	0.076
Meat/poultry/fish/nuts/legumes (oz)	4.9 (3.9)	3.8 (3.2)	−0.555	0.139	4.2 (3.2)	4.0 (2.4)	−0.106	0.764	4.3 (3.0)	4.5 (3.5)	0.172	0.694	4.4 (3.4)	4.0 (3.0)	−0.193	0.340
Processed meat (oz)	1.0 (1.6)	0.71 (1.2)	−0.150	0.451	0.90 (1.4)	0.98 (1.6)	0.031	0.859	0.93 (1.5)	0.50 (1.3)	−0.207	0.306	0.95 (1.5)	0.77 (1.4)	−0.091	0.370
Child weight (lb)	172.4 (49.8)	185.5 (53.2)	3.99	<0.001	174.6 (63.9)	170.1 (64.5)	−0.379	0.540	179.7 (65.2)	181.6 (70.6)	−0.138	0.883	175 (59.0)	178.4 (61.6)	1.34	0.008
Child height (in)	62.3 (5.4)	64.1 (5.1)	0.389	<0.001	62.5 (5.9)	62.7 (5.6)	0.246	<0.001	62.6 (5.4)	63.7 (5.1)	0.249	<0.001	62.5 (5.6)	63.5 (5.3)	0.303	0.037
Child BMI z-score	2.02 (.49)	1.97 (.49)	0.006	0.659	2.00 (.48)	1.85 (.55)	−0.058	0.001	2.02 (.5)	1.82 (.7)	−0.055	0.003	2.01 (.49)	1.89 (.56)	0.033	0.002
Child percent overweight	36.3 (12.0)	35.1 (11.7)	0.187	0.512	35.3 (11.4)	32.3 (12.0)	−1.13	0.001	34.7 (11.2)	30.7 (12.8)	−0.885	0.007	35.5 (11.5)	33.0 (12.1)	0.559	0.003
Parent weight	196.7 (53.7)	199.6 (61.4)	1.07	0.101	204.2 (55.9)	202.9 (55.3)	0.952	0.703	201.7 (40.4)	204 (39.4)	−1.79	0.120	201 (51.0)	203.2 (54.1)	0.524	0.626
Parent BMI	31.4 (6.9)	31.7 (7.1)	0.787	0.091	34.6 (10.1)	33.9 (10.3)	−0.136	0.710	32.3 (6.9)	33.9 (6.9)	−0.052	0.840	32.9 (8.3)	33.1 (8.6)	0.264	0.267

Bold indicates significance p < .05.

Examination by treatment group (Table 3) found the control group to have had statistically significant reductions in both total sugar (p = 0.007) and added sugar (p = 0.010), and a significant increase in total vegetables (p = 0.047). Although this group continued to have significant increases in weight (p < 0.001) and height (p < 0.001), they experienced nonsignificant reductions in BMI z-score and percent overweight. The metformin group demonstrated significant reductions in kcal (p = 0.023), protein (p = 0.012), total fat (p = 0.040), and total dairy (p = 0.017). They had weight loss (non significant [NS]) with increasing height (p < 0.001), resulting in reduced BMI z-score (p = 0.001) and percent overweight (p = 0.001). The switch group only had a significant decrease in fruit juice (p = 0.037) and increase in orange vegetables (p = 0.050). They continued to demonstrate weight gain (NS) and height (p < 0.001) but had significant reduction in BMI z-score (p = 0.003) and percent overweight (p = 0.007).

Identification of predictors of reduced BMI z-score over time

Analysis of change in child BMI z-score (dependent variable) was conducted with longitudinal LMM based on a null model with an ICC of 0.3, indicating the need for LMM procedures. Comparison of the null model with fixed and random intercepts, then random slopes, found the AIC to be lowest in the random intercept/slopes model using an unstructured covariance type and reflecting high variability between individual intercepts. Individual dietary variables and nondietary covariates were then tested in bivariate LMMs, with fixed slopes and random intercepts, to determine association of each with change in BMI z-score over 6 months. Only orange vegetables (p = 0.031) were found to be significantly associated with change in BMI z-score. Bivariate models (which included the individual covariate, time, and interaction between the covariate and time) found a statistically significant interaction between treatment group and change in BMI z-score over time. Treatment group was identified in this analysis as the strongest predictor of change in BMI z-score, with the metformin and switch groups having statistically significant reductions in BMI z-scores compared to the control group (metformin p < 0.001, switch p = 0.003), which is consistent with the IMPACT primary outcome results (Correll et al. 2020). Both parent weight (p = 0.04) and parent BMI (p = 0.05) were also found to be significant predictors of change in child BMI s-score, although there were no significant changes in parent weight nor BMI. As parent weight/BMI increases so does child BMI z-score.

Discussion

This analysis provides insight into the nutritional intake of a sample of children and adolescents treated with SGAs who have gained clinically significant weight. Baseline characteristics reveal a diet excessively high in carbohydrates, fat, and sugar with deficiencies in vegetables/fruits, whole grains, and fiber. Although the mean caloric intake was within RDA range, the proportion of macronutrient composition was in excess, with consumption being two to three times higher than the RDA for carbohydrates, protein, and fat. The sample also exceeded the percent of calories from fat and sugar.

All participants in this study received regular individualized HLE sessions related to diet and physical activity, and although the healthy lifestyle knowledge scores for the child participants did not increase substantially, their healthy eating behaviors did. This finding is much like results of a study with children aged 7–13 years where the effect of nutritional knowledge on food choices was explored (Tarabashkina et al. 2016). Results of the study concluded that factors of social acceptability and appealing taste had moderating effects on consumption and weakened the relationship of child nutritional knowledge with healthy food choices. Also, parent knowledge was positively associated with improved food choices, which may account for improved nutrition in our sample since we included the parent in the healthy lifestyle sessions. Parent healthy lifestyle knowledge scores increased in our study but were not statistically significant, although parents did report change in their own behaviors, such as making fruits/vegetables available in the home, reducing access to sugary beverages, and providing less fast food meals to their children. This is consistent with findings from another study with children aged 8–14 years and their parent, which evaluated a family-centered approach to education on diet and exercise (Jinks et al. 2013). The study found effectiveness of lifestyle changes stemmed from inclusion of family members in the process and efforts to change habits of the family, not just the individual. Similarly, our study took a parent-focused approach, which may account for reductions in sugar and fat in their child's diet and increase in vegetables. Also, of interest is our finding that parent weight and BMI were predictors of their child's weight, which is consistent in the literature related to children in the general population (Wrotniak et al. 2004).

Given dietary education/counseling was implemented with all participants, regardless of study arm randomization, it cannot be concluded that HLE directly resulted in changes in dietary intake without a comparison group. However, substantial changes in eating behaviors were observed. Overall, the entire study sample had significant decreases in calories, carbohydrates, protein, and fat, while increasing HEI subscores for total vegetables and dark green/orange vegetables. There were also significant decreases in refined grains and total/added sugar, indicating the need for further studies, which not only examine nutrient intake but also explore changes in specific types of food/beverage consumption, for example, sugary beverages, sugary dessert foods, types of vegetables and fruits, and so on.

Although this sample's baseline nutritional intake was deficient in fruits/vegetables and excess in fat and sugar, it is similar to the general population of children who, according the 2003–2004 and 2005–2006 CDC NHANES reports, are also deficient in fruits/vegetables and excess in fat/sugar consumption (Reedy et al. 2011). A primary focus of the HLE was for participants to decrease intake of sugar and fat in the forms of sweetened beverages, sugary desserts, and fast food. It is known that the most common foods consumed by children aged 2–18 years associated with high energy density and contributing significantly to childhood obesity are the following: soda, fruit drinks, dairy/grain desserts, pizza, and whole milk (USDA 2015). Other contributing factors include oversized portions and eating meals away from home. Our goal was to encourage subjects to replace sugary snack foods, such as cookies, with whole fruits and raw vegetables, such as apples and carrots. Youth with an unhealthy diet are likely to be more at risk for weight gain; therefore, dietary assessment and basic HLE are essential aspects of care at the initiation of SGA treatment. More recent nutrition studies in prevention/treatment of childhood obesity have revealed evidence that a Mediterranean diet (high in fruits/vegetables, whole grains, fiber, fish, lean meats, legumes, and low/nonfat dairy products) is effective in the prevention of unhealthy weight and metabolic syndrome in children (Katsagoni et al. 2020). A diet plentiful in fruits/vegetables, fish, whole grains, and low in sodium and sugary foods/beverages is also recommended by the American Heart Association 2020 Strategic Impact Goals for promotion of cardiovascular health in children (Steinberger et al. 2016). A key element to insulin sensitivity and glucose regulation are the inclusion of high-fiber and low glycemic foods in the daily diet (Zafar et al. 2019; Reynolds et al. 2020). Our sample's diet was initially high in refined grains, which have a high glycemic index, and quite low in fiber. Although intake of refined grains decreased, there was no increase in whole grains or fiber over the 6 months. Development of an intervention which includes the AMA healthy lifestyle guidelines and more detailed nutritional information/guidance on a Mediterranean like diet could lead to not only improved nutrition but possibly normalize metabolic parameters, such as lipids, hemoglobin A1C, and fasting glucose/insulin by mechanisms of improved leptin, ghrelin, and adiponectin regulation.

A specialized/individualized approach which involves the parent/caregiver is necessary in this population, given that often the youth's psychiatric symptoms impose limitations/barriers not encountered in the general pediatric population. For example, youth with autism often have a very narrow range of foods they will consume, which is often related to textural and taste sensitivities. Assisting parents in creative ways to include fruits/vegetables (e.g., putting fruits and vegetables in foods the youth likes) can be an effective strategy to improve nutritional composition. A qualitative study conducted with caregivers of children treated with SGA medication found parent-reported barriers to healthy lifestyle behaviors were often related to their child's serious mental illness (Nicole et al. 2016). Barriers such as extreme irritability and resistance to consuming new foods, sensory problems around tastes of foods, and social limitations affecting participation in mainstream group physical activities were identified as the primary challenges (Nicole et al. 2016). Therefore, an individualized and personalized approach is required with this unique group of youth to address prevention of obesity and associated comorbid illnesses at an early age.

A primary aim of this study was to identify predictors of reduction in BMI z-score, and in this sample, the most influential predictor was addition of metformin with their current SGA, likely a result of improved insulin sensitivity and glucose metabolism. Other factors to consider in addition to weight loss or mitigation of weight gain, given that children/adolescents are actively growing, include changes in percent body fat, lipids, glucose/insulin levels, and abdominal obesity. The IMPACT study examined metabolic parameters finding only the metformin group had significant reductions in fasting glucose/insulin and insulin resistance (Homeostatic Model Assessment for Insulin Resistance) at 3 months, and hemoglobin A1C at 6 months (Correll et al. 2020). All three groups had a statistically insignificant decrease in triglycerides at 3 months and no other change in lipids or C-reactive protein over the 6 months, although this may be due to attrition in the sample (Correll et al. 2020). Despite significant dietary changes occurring, a longer time period with sustained dietary intake patterns is likely necessary to see significant and lasting improvements in metabolic parameters. The IMPACT study demonstrated a clear effect of metformin on metabolism and reduced adiposity in this sample of obese youth. Further investigation of the effects of dietary interventions in prevention of weight gain in nonoverweight/obese youth starting an SGA, in comparison to combined metformin with dietary interventions, is indicated. Causation of dietary change cannot be attributed to addition of metformin, rather only to reduction in BMI z-score given that each treatment group demonstrated improvements in dietary intake. Despite continued weight gain in the control group, they did experience nonsignificant reductions in BMI z-scores and percent overweight, which could be attributed to dietary changes. It is notable that all three groups experienced some degree of decreased adiposity.

Further dietary research is necessary to fully understand the factors that moderate adiposity, and specific foods/beverages that may contribute to the development and/or prevention of metabolic syndrome in youth treated with SGAs. This analysis provides some initial groundwork for future dietary research in this population. A longer study (12 months) with more frequent time points for dietary assessment, initiated at the start of SGA treatment, is necessary to capture more detailed nutritional intake, eating patterns, and specific foods consumed to further determine the effects on weight, metabolism, and lipid synthesis regulation.

Limitations

While this analysis led to notable findings, specific limitations must be addressed in future studies. First, since HLE was provided to all participants, it could not be truly tested as an intervention and limits the inferential assessment of the results. Second, the HLE teaching was very broad, focusing only on the AMA guidelines for reducing sugar sweetened beverages, eating five fruits/vegetables daily, eating together as a family daily, eating breakfast daily, and reducing fast food to only once per week. There was no actual nutritional education on the components of food groups and eating a balanced diet. Furthermore, exploration of specific foods eaten would be informative (e.g., specific snack foods and beverages). Third, the 24-hour dietary recalls were administered only three times over the course of 6 months. Although the “Gold Standard” USDA AMPM was utilized to assess dietary intake, full NHANES procedures which involve capturing 24-hour dietary recalls over both a weekday and a weekend were not implemented in this study (CDC 2019). Future studies should follow current NHANES procedures and assess dietary intake more frequently than every 3 months to accurately capture changes (CDC 2019). Fourth, better measurement of healthy lifestyle knowledge with a standardized/validated tool would likely more accurately capture changes in knowledge over time. In addition, actual healthy behaviors should be measured to examine the discrepancies between knowledge and behavior. Fifth, all participants had been on an SGA for at least 2 months, but it is not known if their diet changed after starting the medication. Last, this secondary analysis did not include physical activity, which would shed light on the combined effect of physical activity and reduced caloric intake on changes in weight.

Conclusions

This new information contributes to understanding eating behaviors in youth with SGA-induced weight gain and the potential targets for improved dietary patterns, thereby providing a foundation for establishing practice guidelines related to dietary assessment and education in this population. Well-designed studies are needed to better assess dietary intake of macronutrients, micronutrients, food components, and specific foods that increase risk for weight gain in this population. Qualitative studies are also crucial in developing an accurate understanding of the barriers and unique needs of this special group of youth. RCTs testing healthy lifestyle interventions related to diet are needed to further evaluate and determine effectiveness. Ultimately, such information will contribute to the development of evidenced-based strategies for prevention of SGA-induced weight gain in youth.

Clinical Significance

This article provides a unique opportunity to examine the dietary intake and nutritional adequacy in a sample of youth aged 8–18 years who experienced clinically significant weight gain due to antipsychotic treatment. A detailed nutritional analysis was conducted on participants' baseline dietary intake, and after receiving HLE over 6 months finding a substantial improvement in diet quality. Given few studies have been conducted in this population examining specific dietary intake and quality, this study provides new information that can be utilized in further studies and informing clinical care.

Footnotes

Acknowledgment

In memory of Patricia Harris, MS, CRNP, for her contribution to development and implementation of the HLE training for our study team.

Disclosures

K.B., G.R., C.U.C., M.A.R., and L.S. all received grant funding from the NIMH and Betty Huse Foundation for the IMPACT study. K.B., G.R., E.H., S.Z., and M.A.R.: None. C.U.C. has been a consultant and/or advisor to or have received honoraria from: Acadia, Alkermes, Allergan, Angelini, Axsome, Gedeon Richter, Gerson Lehrman Group, Indivior, IntraCellular Therapies, Janssen/J&J, Karuna, LB Pharma, Lundbeck, MedAvante-ProPhase, MedInCell, Medscape, Merck, Mylan, Neurocrine, Noven, Otsuka, Pfizer, Recordati, Rovi, Servier, Sumitomo Dainippon, Sunovion, Supernus, Takeda, and Teva. He provided expert testimony for Janssen and Otsuka. He served on a Data Safety Monitoring Board for Lundbeck, Rovi, Supernus, and Teva. He has received grant support from Janssen and Takeda. He is also a stock option holder of LB Pharma. L.S. is an unpaid consultant to Neuren Pharmaceuticals, an unpaid member of the publications and outreach committee for balovaptan studies for F. Hoffman-La Roche, Inc. A portion of her salary at Duke is paid for by contracts between Duke Clinical Research Institute and Akili Interactive and Tris Pharma related to pediatric clinical trials of their products. She has been a site study in licensing trials conducted by F. Hoffman-La Roche, Inc. and Curemark Pharmaceutical trials.

References

Almandil

, Liu

, Murray

, Besag

, Aitchison

, Wong

: Weight gain and other metabolic adverse effects associated with atypical antipsychotic treatment of children and adolescents: A systematic review and meta-analysis. Pediatr Drugs, 15:139–150, 2013.

Anagnostou

, Aman

, Handen

, Sanders

, Shui

, Hollway

, Brian

, Arnold

, Capanol

, Hellings

, Butter

: Metformin for treatment of overweight induced by atypical antipsychotic medication in young people with autism spectrum disorder: A randomized clinical trial. JAMA Psychiatry, 73:928–937, 2016.

Arango

, Giráldez

, Merchán-Naranjo

, Baeza

, Castro-Fornieles

, Alda

, Carmen

, de Andres

, Cuerda

, de la Serna

, Correll

, Fraguas

, Parellada

: Second-generation antipsychotic use in children and adolescents: A six-month prospective cohort study in drug-naïve patients. J Am Acad Child Adolesc Psychiatry, 53:1179–1190, 2014.

Arman

, Sadramely

, Nadi

, Koleini

: A randomized, double-blind, placebo-controlled trial of metformin treatment for weight gain associated with initiation of risperidone in children and adolescents. Saudi Med J, 29:1130–1134, 2008.

Barlow

: Expert Committee Recommendations Regarding the Prevention, Assessment, and treatment of child and adolescent overweight and obesity: Summary report. Pediatrics 120(Suppl. 4), 2001.

Blanton

, Moshfegh

, Baer

, Kretsch

: The USDA Automated Multiple-Pass Method accurately estimates group total energy and nutrient intake. J Nutr, 136:2594–2599, 2006.

Błażej

, Bartoli

, Stramecki

, Samochowiec

, Lis

, Kasznia

, Jarosz

, Stańczykiewicz

: Appetite regulating hormones in first-episode psychosis: A systematic review and meta-analysis. Neurosci Biobehav Rev, 102:362–370, 2019.

Caemmerer

, Correll

, Maayan

: Acute and maintenance effects of non-pharmacologic interventions for antipsychotic associated weight gain and metabolic abnormalities: A meta-analytic comparison of randomized controlled trials. Schizophr Res, 140:159–168, 2012.

CDC (Centers for Disease Control and Prevention): NHANES Interviewer Procedures Manual. 2019. Available at https://wwwn.cdc.gov/nchs/data/nhanes/2019-2020/manuals/2019-MEC-In-Person-Dietary-Interviewers-Manual-508.pdf (accessed December 18, 2020 ).

10.

Christian

, Gaynes

, Saavedra

, Sheitman

, Wines

, Jonas

, Viswanathan, Ellis

, Woodell

, Carey

: Use of antipsychotic medications in pediatric and young adult populations: Future research needs. J Psychiatr Pract, 21:26–36, 2015.

11.

Correll

, Maayan

, Kane

, Hert

, Cohen

: Efficacy for psychopathology and body weight and safety of topiramate-antipsychotic cotreatment in patients with schizophrenia spectrum disorders: Results from a meta-analysis of randomized controlled trials. J Clin Psychiatry, 77:e746–e756, 2016.

12.

Correll

, Manu

, Olshanskiy

, Napolitano

, Kane

, Malhotra

: Cardiometabolic risk of second-generation antipsychotic medications during first-time use in children and adolescents. JAMA, 302:1765–1773, 2009.

13.

Correll

, Sikich

, Reeves

, Johnson

, Keeton

, Hazzard

, Spanos

, Alderman

, Scheer

, Kapoor

, Sheridan

, Borner

, Bussell

, Pirmohamed

, Bethea

, Reinblatt

, Miller

, Chandrasekhar

, Edwards

, Hamer

, Riddle

: Metformin add-on vs. antipsychotic switch vs. continued antipsychotic treatment plus healthy lifestyle education in overweight or obese youth with severe mental illness: Results from the IMPACT trial. World Psychiatry, 19:69–80, 2020.

14.

Deng

: Effects of antipsychotic medications on appetite, weight, and insulin resistance. Endocrinol Metab Clin, 42:545–563, 2013.

15.

Eldridge

, Piernas

, Illner

, Gibney

, Gurinović

, De Vries

JHM

, Cade

: Evaluation of new technology-based tools for dietary intake assessment—An Ilsi Europe dietary intake and exposure task force evaluation. Nutrients, 11:55, 2019.

16.

Findling

, Johnson

, McClellan

, Frazier

, Vitiello

, Hamer

, Liberman

, Ritz

, McNamara

, Lingler

, Hlastala

, Pierson

, Puglia

, Maloney AE

Kauffman, Noyes

, Sikich

: Double-blind maintenance safety and effectiveness findings from the Treatment of Early-Onset Schizophrenia Spectrum (TEOSS) study. J Am Acad Child Adolesc Psychiatry, 49:583–594, 2010a.

17.

Findling

, Robb

, Nyilas

, Forbes

, Jin

, Ivanova

, Marcus

, McQuade

, Iwamoto

, Carson

: A multiple-center, randomized, double-blind, placebo-controlled study of oral aripiprazole for treatment of adolescents with schizophrenia. Am J Psychiatry, 165:1432–1441, 2008.

18.

Findling

, Youngstrom

, Fristad

, Birmaher

, Kowatch

, Arnold

, Frazier

, Axelson

, Neal

, Demeter

, Gill

, Fields

, Depew

, Kennedy

, Marsh

, Rowles

, Horwitz

: Characteristics of children with elevated symptoms of mania: The Longitudinal Assessment of Manic Symptoms (LAMS) study. J Clin Psychiatry, 71:1664, 2010b.

19.

Fonseka

, Müller

, Kennedy

: Inflammatory cytokines and antipsychotic-induced weight gain: Review and clinical implications. Mol Neuropsych, 2:1–14, 2016.

20.

Galling

, Roldán

, Nielsen

, Gerhard

, Carbon

, Stubbs

, Vancampfort

, De Hert

, Olfson

, Kahl

, Martin

, Guo

, Lane

, Sung

, Liao

, Arango

, Correll

: Type 2 diabetes mellitus in youth exposed to antipsychotics: A systematic review and meta-analysis. JAMA Psychiatry, 73:247–259, 2016.

21.

Gibson

: Principles of Nutritional

Assessment

. Oxford University Press, USA, 2005.

22.

Guenther

, Reed

, Krebs-Smith

, Reeve

, Basiotis

: Development and evaluation of the healthy eating index-2005. U.S. Department of Agriculture, Center for Nutrition Policy and Promotion, 2007. Available at www.cnpp.usda.gov/HealthyEatingIndex.htm (accessed December 18, 2020 ).

23.

Izadi

, Azadbakht

: Specific dietary patterns and concentrations of adiponectin. J Res Med Sci, 20:178, 2015.

24.

Jinks

, English

, Coufopoulos

: Evaluation of a family-centered children's weight management intervention. Health Educ, 113:88–101, 2013.

25.

Katsagoni

, Psarra

, Georgoulis

, Tambalis

, Panagiotakos

, Sidossis

, the EZHN Study

Group

: High and moderate adherence to Mediterranean lifestyle is inversely associated with overweight, general and abdominal obesity in children and adolescents: The MediLIFE-index. Nutr Res, 73:38–47, 2020.

26.

Kishida

, Tohru

, Iichiro

: Adiponectin as a routine clinical biomarker. Best Pract Res Clin Endocrinol Metab, 28:119–130, 2014.

27.

Klein

, Cottingham

, Sorter

, Barton

, Morrison

: A randomized, double-blind, placebo-controlled trial of metformin treatment of weight gain associated with initiation of atypical antipsychotic therapy in children and adolescents. Am J Psychiatry, 163:2072–2079, 2006.

28.

, Wang

, Lin

, Shao

, Chang

, Chou

, Ho

, Liao

, Chen

VCH

: Differential effects of olanzapine and clozapine on plasma levels of adipocytokines and total ghrelin. Progr Neuropsychopharmacol Biol Psychiatry, 58:47–50, 2015.

29.

Moshfegh

, Rhodes

, Baer

, Murayi

, Clemens

, Rumpler

, Paul

, Sebastian

, Kucsynski

, Ingwersen

, Staples

, Cleveland

: The US Department of Agriculture Automated Multiple-Pass Method reduces bias in the collection of energy intakes. Am J Clin Nutr, 88:324–332, 2008.

30.

Nicol

, Worsham

, Haire-Joshu

, Schweiger

, Yingling

, Lenze

: Getting more effective weight management treatment in anti-psychotic treated youth: A survey of barriers and preferences. Childhood Obesity, 12:70–76, 2016.

31.

Olfson

, Blanco

, Liu

, Wang

, Correll

: National trends in the office-based treatment of children, adolescents, and adults with antipsychotics. Arch Gen Psychiatry, 69:1247–1256, 2012.

32.

Olfson

, Druss

, Marcus

: Trends in mental health care among children and adolescents. N Engl J Med, 372:2029–2038, 2015.

33.

Olfson

, King

, Schoenbaum

: Treatment of young people with antipsychotic medications in the United States. JAMA Psychiatry, 72:867–874, 2015.

34.

Owen

, Sikich

, Marcus

, Corey-Lisle

, Manos

, McQuade

, Carson

, Findling

: Aripiprazole in the treatment of irritability in children and adolescents with autistic disorder. Pediatrics, 124:1533–1540, 2009.

35.

Raper

, Perloff

, Ingwersen

, Steinfeldt

, Anand

: An overview of USDA's dietary intake data system. J Food Comp Analysis, 17:545–555, 2004.

36.

Reedy

, Krebs-Smith

: Dietary sources of energy, solid fats, and added sugars among children and adolescents in the United States. J Am Dietet Assoc, 110:1477–1484, 2011.

37.

Reeves

, Keeton

, Correll

, Johnson

, Hamer

, Sikich

, Hazzard

, Alderman

, Scheer

, Mabe

, Kapoor

, Sheridan

, Borner

, Bussell

, Pirmohamed

, Bethea

, Chekuri

, Gottfried

, Reinblatt

, Santana

, Riddle

: Improving metabolic parameters of antipsychotic child treatment (IMPACT) study: Rationale, design, and methods. Child Adolesc Psychiatry Mental Health, 7:31, 2013.

38.

Reynolds

, Akerman

, Mann

: Dietary fibre and whole grains in diabetes management: Systematic review and meta-analyses. PLoS Med, 17:e1003053, 2020.

39.

Rao

: Childhood obesity: Highlights of AMA Expert Committee recommendations. Am Fam Phys, 78:56–63, 2020.

40.

Rubin

, Kreider

, Matone

, Huang

, Feudtner

, Ross

, Localio

: Risk for incident diabetes mellitus following initiation of second-generation antipsychotics among Medicaid-enrolled youths. JAMA Pediatr, 169:e150285, 2015.

41.

Steinberger

, Daniels

, Hagberg

, Isasi

, Kelly

, Lloyd-Jones

, Russell

, Pratt

, Slay

, Towbin

, Urbina

, Van Horn

, Zachariah

: Cardiovascular health promotion in children: Challenges and opportunities for 2020 and beyond: A scientific statement from the American Heart Association. Circulation, 134:e236–e255, 2016.

42.

Subar

, Kirkpatrick

, Mitt

, Zimmerman

, Thompson

, Bingley

, Willis

, Islam

, Baranowski

, McNutt

, Potischman

: The Automated Self-Administered 24-hour dietary recall (ASA24): A resource for researchers, clinicians, and educators from the National Cancer Institute. J Acad Nutr Dietet, 112:1134–1137, 2012.

43.

Tarabashkina

, Quester

, Crouch

: Exploring the moderating effect of children's nutritional knowledge on the relationship between product evaluations and food choice. Soc Sci Med, 149:145–152, 2016.

44.

Teasdale

, Ward

, Rosenbaum

, Watkins

, Curtis

, Kalucy

, Samaras

: A nutrition intervention is effective in improving dietary components linked to cardiometabolic risk in youth with first-episode psychosis. Br J Nutr, 115:1987–1993, 2016.

45.

Thompson

, Dixit-Joshi

, Potischman

, Dodd

, Kirkpatrick

, Kushi

, Alexander

, Coleman

, Zimmerman

, Sundaram

, Clancy

: Comparison of interviewer-administered and automated self-administered 24-hour dietary recalls in 3 diverse integrated health systems. Am J Epidemiol, 181:970–978, 2015.

46.

USDA (U.S. Department of Agriculture), Agricultural Research Service, Food Surveys Research Group: USDA Food and Nutrient Database for Dietary Studies, 3.0. Beltsville, MD, 2008. Available at https://www.ars.usda.gov/ARSUserFiles/80400530/pdf/fndds/fndds3_doc.pdf (accessed December 18, 2020 ).

47.

USDA: 2015–2020 Dietary Guidelines for Americans. U.S. Department of Health and Human Services, Washington (DC), 2015.

48.

Wrotniak

, Epstein

, Paluch

, Roemmich

: Parent weight change as a predictor of child weight change in family-based behavioral obesity treatment. Arch Pediatr Adolesc Med, 158:342–347, 2004.

49.

Zafar

, Mills

, Zheng

, Regmi

, Hu

, Gou

, Chen

: Low-glycemic index diets as an intervention for diabetes: A systematic review and meta-analysis. Am J Clin Nutr, 110:891–902, 2019.

50.

Zhang

, Deng

, Huang

: The role of ghrelin signaling in second-generation antipsychotic-induced weight gain. Psychoneuroendocrinology, 38:2423–2438, 2013.

51.

Zheng

, Li

, Tang

, Xiang

, Wang

, de Leon

: Metformin for weight gain and metabolic abnormalities associated with antipsychotic treatment: Meta-analysis of randomized placebo-controlled trials. J Clin Psychopharmacol, 35:499–509, 2015.