Inflammation is thought to play a critical role in the development of both metabolic and psychotic disorders. Researchers have documented patterns of inflammatory cytokine dysregulation and evidence of oxidative stress in individuals with schizophrenia (2, 6–11). An inflammatory paradigm may provide a theoretical framework to inform etiological questions that surround psychosis and provide insights into targeted interventions in psychotic illness.

While studies of inflammation in psychosis are promising, findings are heterogenous across the psychosis spectrum given multiple confounders such as metabolic disturbances, chronicity of illness, substance use, and varying levels of exposure to environmental stressors (6–8, 10–13). Long-term exposure to antipsychotic medication, common in chronic psychosis cohorts, limits understanding of how inflammatory processes impact illness severity. Antipsychotics have been shown to both mediate and exacerbate inflammation, in turn contributing to the presence of metabolic syndrome in individuals with psychosis (14–18). These findings underscore the need to study the earlier stages of illness in which individuals have less exposure to these factors.

Increased emphasis in recent years has been placed on studying first-episode psychosis (FEP) populations; these are individuals who have experienced a clinically significant episode of psychosis within the past 3–5 years (19). Robust clinical findings have suggested early intervention leads to improvements in symptom severity, social functioning and long-term clinical and cognitive outcomes for this population when employed within the first 5 years after the first-episode (12, 19–23). This is especially salient in the context of comorbidity of psychosis with inflammatory metabolic conditions. Individuals even in the earlier phases of psychotic illness show increased inflammatory biomarkers and higher risk for sedentary lifestyles (4, 16, 24–26). These findings suggest that psychosis could be viewed through the lens of a systemic illness in which inflammation can serve as a viable biomarker for disease risk or state. Several studies have indicated the presence of inflammatory abnormalities in both clinical high-risk (CHR) psychosis and FEP where individuals have had limited exposure to illness or medication (4, 16, 27–29).

A recent meta-analysis by Firth et al. (30) evaluated several studies (N = 28) measuring inflammatory characteristics of FEP populations and found that in addition to inflammatory dysregulation there is also evidence of nutritional deficiencies in FEP versus controls such as lower B vitamins (folate or B9 and cobalmin or B12), lower vitamin D, as well as some evidence of lower nutritional minerals such as zinc and magnesium. Cadenhead et al. (24) observed metabolic abnormalities among medication naive participants at CHR for psychosis as well as an association between diets low in Omega 3 and symptoms of subsyndromal psychosis. Unfortunately, current gold-standard treatment for early-course psychosis does little to consider environmental risk factors on overall health and disease progression (31). Given that the risk for severity of metabolic health issues appears to be positively associated with illness duration among individuals with FEP in particular, there is a need for anti-inflammatory intervention trials in this population that not only assess psychotic symptoms as target endpoints but as biomarkers of metabolic health (26).

Anti-inflammatory interventions have been explored across diverse patient populations. Significant research attention has been placed on inflammation as a modifiable treatment target that can be influenced through different modalities ranging from anti-inflammatory medications to changes in one’s diet and lifestyle (3, 9, 13, 32). While the mechanisms by which inflammation impacts the trajectory of psychiatric disorders is largely unknown, there are active research initiatives to trial low-harm interventions that have demonstrated anti-inflammatory benefit and potential impact on psychiatric symptoms (3, 6, 10, 33).

The use of non-steroidal anti-inflammatory drugs (NSAIDs) to reduce systemic inflammation among chronic patients has been documented with small studies indicating NSAID intervention leads to decreased peripheral inflammatory cytokine levels and decreased psychotic symptomatology (34–38). Findings from Zhang et al. (39) demonstrated the potential value of microglia inhibitors such as minocycline in the amelioration of cognitive deficits, negative symptoms, and overall systemic inflammation among a group of individuals with chronic schizophrenia (40) (N = 75). Recent studies have also evaluated the anti-inflammatory effects of Omega-3, Vitamin D, mindfulness, yoga, and cannabinoids in psychosis populations, highlighting the potential clinical utility of these interventions as they are easily modifiable and they promote health and well-being (16, 24, 41, 42). Two reviews of adjunctive nutritional interventions for schizophrenia by (43, 44) outlined that there is growing evidence supporting the possible benefits of supplementation of N-acetyl cysteine (NAC), alpha lipoic acid (ALA), Melatonin, B vitamins, L-thyanine, and essential polyunsaturated fatty acids (PUFAs) on metabolic health, sleep, and in some studies, negative symptoms when used in tandem with traditional antipsychotic intervention. Joseph et al. (45) similarly suggest there is substantial preclinical and clinical evidence in schizophrenia contexts showing that the adjunctive adoption of short chain fatty acids (SCFA), which have been shown to reduce neuroinflammation, through an anti-inflammatory diet may be a viable combined intervention in schizophrenia given its emphasis on adding whole foods rather than restricting foods or food groups.

Dietary interventions (DI) that emphasize reducing inflammation through the consumption of nutritious whole foods have been investigated across medical conditions including rheumatoid arthritis, depression, and chronic schizophrenia (2, 3, 46–51). In addition to the early metabolic changes, individuals with psychosis are also more likely to have diets that are high in processed carbohydrates and refined sugars, which can imbalance the gut microbiome and increase insulin resistance (2, 3, 32). Small studies suggest dietary interventions are low-harm and have a beneficial impact on improving psychiatric symptoms as well as overall metabolic health (2, 3, 9, 47, 52). Gilbert-Jaramillo et al. (47) investigated the efficacy of a 12-week ketogenic diet for chronic psychosis and found a significant reduction in positive and negative symptoms as well as improved liver function among participants. Based on these findings, similar diets may have utility to address symptoms and health for early psychosis groups. Specifically, anti-inflammatory diets emphasize a high intake of Omega-3 fatty acids, fibrous cruciferous greens, probiotics, and a low intake of processed carbohydrates and refined sugars. These diets can regulate metabolic processes through the production of SCFA, a hallmark dysregulation in a variety of inflammatory diseases and psychiatric conditions (2, 46, 53, 54). Some dietary interventional studies also address dysbiosis (an imbalance of colonies of microbacteria in the gut), an emerging theory behind inflammatory findings in several psychiatric disorders (54–56). Dietary intervention may help to recultivate and encourage the growth of healthy gut bacteria (50, 53, 54).

A recent review by Aucoin et al. (3) evaluated 718 articles (N = 52,634) that explored topics such as levels of individual nutrients and vitamins common in this population, the gut microbiome, dietary macronutrients and patterns, as well as food sensitivities in various psychosis populations (3). This review developed a comprehensive summary of evidence and informed dietary recommendations for individuals with psychosis, emphasizing a reduction in processed foods and refined carbohydrates. This summary also included a list of dietary factors to increase (e.g., fruits, vegetables, high fiber foods, Omega 3 fatty acids, sources of vitamin B12, folate, vitamin B6, zinc, protein sources that are high in glycine and lysine, and vitamin C) and recommended their sources (3). Despite only involving studies of chronic psychosis populations, this novel review provided valuable psychosis-specific insight that can be adapted for early psychosis populations.

Given socioeconomic disparities among individuals living with chronic psychotic disorders including limited access to affordable food, lack of psychosocial support to maintain dietary habits and unstable living conditions, DI are often difficult to implement and maintain overtime and its true effects are difficult to entangle from post-morbid effects associated with disease chronicity (2, 9, 32, 52, 57). The period after the FEP is considered a critical window for treatment intervention, prior to the onset of health problems, when individuals may have more access to family support networks who can assist with dietary monitoring and food preparation (22). Current recommended clinical intervention for FEP is generally limited to a combination of antipsychotics and cognitive behavioral therapy (CBT), and while these treatments are generally effective for reducing positive symptoms associated with psychosis, their impact on negative symptoms and comorbid health issues is generally negligible (58, 59). There is a need to incorporate dietary education and wellness intervention into treatment for psychosis in young people as they gain independence. Incorporation of health education is critical to the lifespan approach to identifying and treating psychosis, and there is a dearth of research aimed at studying the impact of nutritional assistance and education on clinical outcomes in this population (57, 60).

To our knowledge, only one study to date has trialed an anti-inflammatory diet among FEP specifically. Among a cohort of 33 FEP participants and aged-matched healthy controls, Vassilopoulou et al. (61) trialed adherence to a Mediterranean dietary intervention as adjunctive therapy with traditional antipsychotics. These investigators relied on self-reported intake of anti-inflammatory foods (olive oil, fermented foods) and measured baseline and endpoint metabolic markers such as blood pressure and blood glucose (61). This study found that adherence to the Mediterranean diet was lower among FEP compared to controls, however there was no difference in reported intake of fermented foods between diagnostic groups. Additionally, there were variations in blood glucose between individuals on different antipsychotic medication, warranting future study. Importantly, adherence to the Mediterranean diet was negatively associated with blood glucose among a subset of FEP who were taking olanzapine, azapine, clozapine, or risperidone specifically providing essential information related to the impact of adherence on metabolic changes in this population (61). This study provided valuable insight into the feasibility of an anti-inflammatory DI in FEP, however feasibility and acceptability were not the primary targets of this study, suggesting a need for future studies that aim to work collaboratively with patients to design a personalized dietary intervention.

Based on this background information, the present paper describes the methods to develop a DI informed by knowledge of the mechanistic and systemic biological changes observed in FEP and iterative feedback from patients and their families.

Here we describe the methods to address the following aims:

This novel DI, developed through a multi-phasic small-scale feasibility study, is intended to provide a framework for future nutritional intervention trials for FEP as well as their support systems that emphasizes improvements in diet, psychiatric symptoms, general health, and is guided by lived experience and perspectives.

The study will include participants aged 15–30 and their support figures (e.g., family and/or partners). Eligible participants will have received a diagnosis of either a schizophrenia-spectrum disorder or an affective psychotic disorder within the past 5 years as determined by the Structured Clinical Interview for DSM-5 Disorders (SCID-5) (64). All clinical screening assessments will be administered by trained diagnostic raters with master or doctoral level degrees in clinical psychology within the UCSD CARE program. Both affective and non-affective psychoses will be included given the heterogeneity of FEP diagnoses, and that a wide range of psychosis spectrum conditions are reflective of the diagnostic makeup of the UCSD CARE program (65, 66). Full inclusion and exclusion criteria are outlined in Table 1. Importantly, individuals will remain on their current prescribed psychiatric medication as the proposed project is testing the feasibility of an adjunctive therapy. If an individual is required to initiate antibiotics at any period during study participation in the Feasibility phase (discussed below), they will continue with the DI protocol if medically indicated and antibiotic use will be documented. Individuals currently taking antibiotic medication at study screening/ entry will not be included in the study or will be recruited once antibiotic use has ceased (see Table 1 for time-course permitted). Individuals will be representative of the greater San Diego, California community and will be reflective of diverse racial, ethnic, educational, and socioeconomic backgrounds as well as neurodivergent and sexual and gender minority statuses as is naturally observed within the UCSD CARE Program. Differences in sociodemographic factors will be evaluated following the end of this feasibility study, and any limitations or trends associated with these factors (e.g., skewed socioeconomic or educational representation) will inform future dietary interventions in this population.

Individuals will be recruited chronologically for participation in distinct phases of this study which will be described in detail below (Figure 1). Severity of psychosis symptoms will be assessed at the initial screening period through the Positive and Negative Symptom Scale (PANSS), and current mood will be evaluated through the Calgary Depression Scale for Schizophrenia (CDSS) (67, 68). The Eating Attitudes Test (EAT-26) will be administered to all participants to learn more about disordered eating behavior in this population, a topic poorly understood and understudied in psychosis-spectrum disorders (69).

In the Development phase, preliminary sample diets were dynamically created based on dietary preferences and physical needs that are specific and flexible to psychosis populations based on input from registered dieticians at the CIM as well as clinicians within the CARE clinic (Table 2). In addition to relying on research findings in this area, the delivery of this tentative DI was refined and strategized based on clinician views of perceived acceptability and feasibility of patients within the CARE clinic with whom clinicians had worked with or treated. The primary method of delivery for this DI is through remote modalities with the intention of determining whether these channels of communication are feasible and acceptable by research participants in the Formative phase of the study. During the COVID-19 pandemic, such methods of service delivery have been found to be effective in reaching diverse populations (70, 71). This method may be essential to those living with psychosis as several studies have highlighted poor in-person treatment and research adherence among FEP populations (72, 73).

The Formative phase involved four distinct focus groups (Table 3). Each focus group in this phase provided novel information on participants’ attitudes towards health, dietary preferences, accessibility to food, feasibility, and acceptability of a DI based on lived experience. Information collected during each focus group in this phase was used to refine the DI further in preparation for the Feasibility Phase. Participants in each focus group will be evaluated remotely via telehealth within their homes and may have a family member or support figure present depending on their comfort level/preference.

The primary aim of the Feasibility phase is to assess the feasibility and acceptability of the 5-week anti-inflammatory DI developed in the Formative phase among a small FEP cohort (81). A secondary aim is to naturalistically evaluate the impact of the 5-week DI on weight and various other metabolic markers of inflammation as well as on clinical symptoms. The DI’s duration of 5-weeks was both informed by experience within the CIM regarding an appropriate duration to see metabolic changes, as well as input from patients in the Formative Phase of an appropriate timeline for their availability and interest. Meta-analytic findings also demonstrate that short-term dietary interventions (longer than 2 weeks) have been shown to impact gut microbiota (82), and similar reviews have noted that anti-inflammatory dietary interventions longer than 4 weeks may have observable effects on blood-based inflammatory markers (83, 84) A projected sample size of 12 for this phase was proposed a priori based on similar feasibility studies within the CIM, and parameters that would be measured in the Feasibility phase were selected based on responses from participants in the Formative phase. Following the Formative phase, study staff from both the UCSD CARE Program and the CIM discussed participants’ qualitative and quantitative feasibility and acceptability ratings for all assessments administered as well as educational materials provided. Study staff then discussed and decided, based on these responses, which assessments and educational materials would be administered in the Feasibility phase and reduced assessments in which there was significant overlap of content or which participants had expressed difficulty completing or understanding (Table 4). Given that the objectives of the proposed study are to assess the feasibility and acceptability of the DI as well as to explore any impact on metabolic outcomes following the DI within an FEP population, healthy control participants will not be included in this pilot study as there is no methodological question proposed in this project regarding applicability or comparability to a healthy control group.

Given that participants are going to be recruited directly from the UCSD CARE Early Psychosis program, we anticipate retention across the 5-week DI as study staff have regular communication and interaction with patients in the clinic and research setting. If participants withdraw or leave the study at any point during the 5-week DI, this missing data will be incorporated into our holistic understanding of the feasibility and acceptability of the DI. Participants in this study are not randomized and primarily qualitative data is being collected over the course of the DI, therefore typical methods for managing attrition issues and missing data used in clinical trials such as intention to treat (ITT) or last observation carried forward (LOCF) analyses are not necessary. FEP participants recruited to this phase will importantly be different than those who participated in the Formative phase to ensure objective exposure to final DI material. Possible barriers to attrition are also addressed through the modality in which the DI will be delivered; participants will only come into the research center physically for the baseline and endpoint collection of bio-samples, and all clinical assessment appointments and weekly diet-coaching sessions will be held remotely.

For the clinical and metabolic measures collected in the Feasibility phase (discussed below), a series of exploratory analyses will be used to estimate effect sizes and directionality of changes from pre to post 5-week DI. If basic statistical assumptions are met, the preferred method of analysis will be a one sample paired t-test between baseline and end point measures. If the final data does not meet these assumptions, an alternative Wilcoxon signed-rank will be considered. Pearson’s correlation will be used to analyze the association of average adherence to secondary outcome measures, however if the data appears to be non-parametric, alternative Spearman’s correlations will be considered. For the analysis of feasibility, acceptability and participant satisfaction of the DI, descriptive statistics will be obtained for all variables and the appropriate statistical test will be employed to assess normality and homogeneity of variance.

Per the UCSD Health IRB guidelines, participant data will be collected under de-identified subject IDs. All data will be collected and simultaneously input into a secure REDCap database formulated specifically for this study (87). Clinical assessments, demographic data, and diet coaching materials will all be stored on this secure database. Study staff will have all completed the required Collaborative Institutional Training Initiative (CITI) certification for research in Human Subjects (CITI).

Height, weight, and other markers of obesity and metabolic health will be collected pre and post DI (Table 4). Plasma lipids, fasting glucose, hemoglobin A1C (HgA1C), C-Reactive Protein (CRP), and insulin levels will be collected via blood draw at baseline and week 5. The gut microbiome for each participant will be examined pre and post intervention through the collection of fecal samples at baseline and at endpoint. Based on preexisting fecal sample collection protocols from Sinha et al. (85), samples will be obtained using Fecal Occult Blood Tests (FOBT). Levels of immune-modulating gut species Lactobacillis and Bifidobactirium will be analyzed using Shotgun metagenomics sequencing (85, 86). During this initial visit, participants will be introduced to study staff and provided with instructional material on the DI as well as a 5–7 day meal plan that will be adjusted with each week of the DI. They will be provided with dried goods that are essential for the DI, and will be given instructions on how to complete questionnaires and will gain access to online instructional cooking material.

After baseline, participants will meet once per week for 5 weeks with study staff and complete questionnaires. During these weekly check-ins, participants will meet with the registered dietician and review weekly adherence to the DI, as well as barriers and facilitators related to adherence. The registered dietician will consult on modifications to the DI as well as review sample recipes with the participant and answer any questions (Supplementary material). Medication adherence and overall health and well-being will be assessed at these visits, and participants will be encouraged to explore the available online resources of cooking videos and recipes to assist them with the DI. After the 5-week DI, participants will complete final biomarker testing and clinical evaluation.

One aim of the Feasibility phase of this DI will be to assess patterns of participant recruitment and retention, participant satisfaction to the DI, and to evaluate the overall feasibility of a 5-week anti-inflammatory DI among FEP cohorts. Adherence to the overall research process will be evaluated based on the number of participants that complete all timepoints of the study. Adherence to the DI will incorporate this measure as well as qualitative weekly adherence to the DI as recorded by the registered dieticians in the weekly coaching sessions. At the end of the 5-week DI, feasibility of the DI will be informed by participant self-report (81, 89). Subjects will also be asked to complete the Research Participant’s Perception Survey-UP to provide feedback into their overall experiences as a research participant in this study (89). This phase will also elucidate the feasibility of collecting and monitoring naturalistic changes in certain biomarkers that can be used to measure metabolic health in early psychosis populations, which is an area of significant interest in the study of inflammation across psychiatric disorders. Similarly, psychiatric symptoms will be evaluated pre-and-post DI as an exploratory aim. Any changes in symptoms observed following adherence to the DI will inform the relevance of monitoring psychiatric symptoms in this context.

The proposed study has several potential limitations. Generalizability of findings from this study is limited by the singular site in which it was conducted, and findings are only representative of the participants who completed this DI and are not broadly applicable to all FEP contexts. Future clinical studies that may implement this DI, if feasibility and acceptability are determined in the proposed study, can include direct comparisons between FEP control groups receiving standard of care treatment to those receiving anti-inflammatory DI adjunctive to standard of care. These studies should ensure representation from a diverse range of FEP individuals, larger sample sizes for statistical power, as well as inclusion of multiple clinical sites from varying geographic regions. Future studies may also explore longitudinal differences in response to this DI; the duration of the proposed DI was determined a priori based on expertise among research staff, patient perspectives, and existing literature to be reflective of a relatively limited time-window. Findings regarding the feasibility of this adherence to this short-term DI do not speak to the long term sustainability of this intervention in this population. Importantly however, findings regarding the “appropriate” or “recommended” length or sustainability of anti-inflammatory dietary interventions are mixed (46, 49, 52).

Importantly, while evidence of the role of inflammation in the pathophysiology of psychosis is promising and highlights important biological processes associated with psychosis, it is critical to underscore that this research is in its infancy especially in the context of early-course psychosis. The aim of this study is solely to provide the preliminary blueprint of a DI for early psychosis that is formed and refined using data and feedback collected from participants, with the Feasibility phase assessing the feasibility and acceptability of the DI along with metabolic and inflammatory biomarker assessments as target endpoints. Changes in metabolic and psychiatric health in the Feasibility phase may provide insight into the potential clinical treatment model, however future clinical trials would be required to evaluate its utility beyond the present sample.