Advertisement

Persistent Childhood and Adolescent Anxiety and Risk for Psychosis: A Longitudinal Birth Cohort Study

  • Isabel Morales-Muñoz
    Correspondence
    Address correspondence to Isabel Morales-Muñoz, Ph.D.
    Affiliations
    Institute for Mental Health, School of Psychology, University of Birmingham, Birmingham, United Kingdom

    Department of Public Health Solutions, Finnish Institute for Health and Welfare, Helsinki, Finland
    Search for articles by this author
  • Edward R. Palmer
    Affiliations
    Institute for Mental Health, School of Psychology, University of Birmingham, Birmingham, United Kingdom

    Birmingham and Solihull Mental Health Foundation Trust, Birmingham, United Kingdom
    Search for articles by this author
  • Steven Marwaha
    Affiliations
    Institute for Mental Health, School of Psychology, University of Birmingham, Birmingham, United Kingdom

    Specialist Mood Disorders Clinic, Zinnia Centre, Birmingham, United Kingdom

    Barberry National Centre for Mental Health, Birmingham, United Kingdom
    Search for articles by this author
  • Pavan K. Mallikarjun
    Affiliations
    Institute for Mental Health, School of Psychology, University of Birmingham, Birmingham, United Kingdom

    Early Intervention Service, Birmingham Women’s and Children’s NHS Trust, Birmingham, United Kingdom
    Search for articles by this author
  • Rachel Upthegrove
    Affiliations
    Institute for Mental Health, School of Psychology, University of Birmingham, Birmingham, United Kingdom

    Early Intervention Service, Birmingham Women’s and Children’s NHS Trust, Birmingham, United Kingdom
    Search for articles by this author
Open AccessPublished:December 12, 2021DOI:https://doi.org/10.1016/j.biopsych.2021.12.003

      Abstract

      Background

      Persistent anxiety in childhood and adolescence could represent a novel treatment target for psychosis, potentially targeting activation of stress pathways and secondary nonresolving inflammatory response. Here, we examined the association between persistent anxiety through childhood and adolescence with individuals with psychotic experiences (PEs) or who met criteria for psychotic disorder (PD) at age 24 years. We also investigated whether C-reactive protein mediated any association.

      Methods

      Data from the Avon Longitudinal Study of Parents and Children (ALSPAC) were available in 8242 children at age 8 years, 7658 at age 10 years, 6906 at age 13 years, and 3889 at age 24 years. The Development and Well-Being Assessment was administered to capture child and adolescent anxiety. We created a composite score of generalized anxiety at ages 8, 10, and 13. PEs and PD were assessed at age 24, derived from the Psychosis-like Symptoms Interview. The mean of C-reactive protein at ages 9 and 15 years was used as a mediator.

      Results

      Individuals with persistent high levels of anxiety were more likely to develop PEs (odds ratio 2.02, 95% CI 1.26–3.23, p = .003) and PD at age 24 (odds ratio 4.23, 95% CI 2.27–7.88, p < .001). The mean of C-reactive protein at ages 9 and 15 mediated the associations of persistent anxiety with PEs (bias-corrected estimate −0.001, p = .013) and PD (bias-corrected estimate 0.001, p = .003).

      Conclusions

      Persistent high levels of anxiety through childhood and adolescence could be a risk factor for psychosis. Persistent anxiety is potentially related to subsequent psychosis via activation of stress hormones and nonresolving inflammation. These results contribute to the potential for preventive interventions in psychosis, with the novel target of early anxiety.

      Keywords

      Psychosis is a heterogeneous illness with several risk and protective factors (
      • Radua J.
      • Ramella-Cravaro V.
      • Ioannidis J.P.A.
      • Reichenberg A.
      • Phiphopthatsanee N.
      • Amir T.
      • et al.
      What causes psychosis? An umbrella review of risk and protective factors.
      ). Psychotic disorders (PDs) are among the leading causes of disability (
      • Whiteford H.A.
      • Degenhardt L.
      • Rehm J.
      • Baxter A.J.
      • Ferrari A.J.
      • Erskine H.E.
      • et al.
      Global burden of disease attributable to mental and substance use disorders: findings from the Global Burden of Disease Study 2010.
      ), with a mean incidence of 31.7 per 100,000 people in England (
      • Kirkbride J.B.
      • Errazuriz A.
      • Croudace T.J.
      • Morgan C.
      • Jackson D.
      • Boydell J.
      • et al.
      Incidence of schizophrenia and other psychoses in England, 1950-2009: A systematic review and meta-analyses.
      ) and a 12-month prevalence of 1.1% among the U.S. population (
      • Cloutier M.
      • Aigbogun M.S.
      • Guerin A.
      • Nitulescu R.
      • Ramanakumar A.V.
      • Kamat S.A.
      • et al.
      The economic burden of schizophrenia in the United States in 2013.
      ). However, the precise etiology of psychosis remains to be determined (
      • Fusar-Poli P.
      • Rutigliano G.
      • Stahl D.
      • Davies C.
      • Bonoldi I.
      • Reilly T.
      • et al.
      Development and validation of a clinically based risk calculator for the transdiagnostic prediction of psychosis.
      ). Evidence suggests that this involves genetic and environmental risk factors, along with an interaction between the two (
      • van Os J.
      • Rutten B.P.
      • Poulton R.
      Gene-environment interactions in schizophrenia: Review of epidemiological findings and future directions.
      ). Well-recognized environmental factors include deprivation, childhood trauma, and minority status (
      • Dean K.
      • Murray R.M.
      Environmental risk factors for psychosis.
      ), which are stressful exposures in early development. Further, individuals who develop psychosis are more likely to experience socioemotional and behavioral problems in childhood (
      • Rutter M.
      • Kim-Cohen J.
      • Maughan B.
      Continuities and discontinuities in psychopathology between childhood and adult life.
      ). However, whether childhood anxiety is specifically associated with subsequent psychosis is still underinvestigated.
      Anxiety may have a role in the development of specific psychopathology relevant to psychosis, such as hypervigilance and increased threat to self-network leading to persecutory delusions (
      • Correll C.U.
      • Hauser M.
      • Auther A.M.
      • Cornblatt B.A.
      Research in people with psychosis risk syndrome: A review of the current evidence and future directions.
      ), or as a driving factor for misinterpreting anomalous experiences leading to hallucinations (
      • Garety P.A.
      • Kuipers E.
      • Fowler D.
      • Freeman D.
      • Bebbington P.E.
      A cognitive model of the positive symptoms of psychosis.
      ,
      • Freeman D.
      • Fowler D.
      Routes to psychotic symptoms: Trauma, anxiety and psychosis-like experiences.
      ). After the onset of psychosis, anxiety is common, with prevalence rates for anxiety disorders in psychosis ranging from 42% to 74% (
      • Ciapparelli A.
      • Paggini R.
      • Marazziti D.
      • Carmassi C.
      • Bianchi M.
      • Taponecco C.
      • et al.
      Comorbidity with axis I anxiety disorders in remitted psychotic patients 1 year after hospitalization.
      ,
      • Bosanac P.
      • Mancuso S.G.
      • Castle D.J.
      Anxiety symptoms in psychotic disorders: Results from the Second Australian National Mental Health Survey.
      ). However, existing studies examining the longitudinal relationship between anxiety and psychosis are scarce, and the results are inconsistent. For instance, associations between changes in depression/anxiety and coexisting psychotic-like experiences over 1 year in adolescence have been reported (
      • Yamasaki S.
      • Usami S.
      • Sasaki R.
      • Koike S.
      • Ando S.
      • Kitagawa Y.
      • et al.
      The association between changes in depression/anxiety and trajectories of psychotic-like experiences over a year in adolescence.
      ). Further, anxiety was a strong predictor of paranoid thinking at 18-month follow-up in 16- to 74-year-old individuals (
      • Freeman D.
      • Stahl D.
      • McManus S.
      • Meltzer H.
      • Brugha T.
      • Wiles N.
      • et al.
      Insomnia, worry, anxiety and depression as predictors of the occurrence and persistence of paranoid thinking.
      ), and having panic attacks precedes psychotic symptoms at ages 18 and 21 (
      • Goodwin R.D.
      • Fergusson D.M.
      • Horwood L.J.
      Panic attacks and psychoticism.
      ). To date, previous studies have investigated anxiety at a single time point and in adulthood rather than investigating the continued influence of anxiety in childhood and adolescence. The period of childhood and adolescence is the core risk phase for developing anxiety disorders (
      • Beesdo-Baum K.
      • Knappe S.
      Developmental epidemiology of anxiety disorders.
      ), and anxiety at these ages is a risk factor for general mental disorders in later adolescence and adulthood (
      • Copeland W.E.
      • Angold A.
      • Shanahan L.
      • Costello E.J.
      Longitudinal patterns of anxiety from childhood to adulthood: the Great Smoky Mountains Study.
      ,
      • Copeland W.E.
      • Shanahan L.
      • Costello E.J.
      • Angold A.
      Childhood and adolescent psychiatric disorders as predictors of young adult disorders.
      ,
      • Doering S.
      • Lichtenstein P.
      • Gillberg C.
      • Middeldorp C.M.
      • Bartels M.
      • Kuja-Halkola R.
      • et al.
      Anxiety at age 15 predicts psychiatric diagnoses and suicidal ideation in late adolescence and young adulthood: Results from two longitudinal studies.
      ). However, whether persistent high levels of anxiety might be associated with risk of psychosis in the longer term is currently unknown.
      If persistent childhood and adolescent anxiety is part of the pathway to psychosis, this may be related to a chronic activation of acute-phase proteins and nonresolving inflammation. C-reactive protein (CRP) may increase proinflammatory cytokines, which could have a direct impact on microglial and astrocytic function that is linked to brain change (
      • Upthegrove R.
      • Khandaker G.M.
      Cytokines, oxidative stress and cellular markers of inflammation in schizophrenia.
      ). More specifically, recent hypotheses are emerging suggesting that astrocytic reactivity may reflect impaired restraint by hypofunctional regulatory T cells, leading to subsequent structural and functional brain changes in schizophrenia (
      • Corsi-Zuelli F.
      • Deakin B.
      Impaired regulatory T cell control of astroglial overdrive and microglial pruning in schizophrenia.
      ). Further, there is some indication of association between anxiety and higher CRP levels (
      • Liukkonen T.
      • Räsänen P.
      • Jokelainen J.
      • Leinonen M.
      • Järvelin M.R.
      • Meyer-Rochow V.B.
      • et al.
      The association between anxiety and C-reactive protein (CRP) levels: Results from the Northern Finland 1966 birth cohort study.
      ,
      • Bullmore E.T.
      • Lynall M.E.
      Immunologic therapeutics and psychotic disorders.
      ,
      • Bankier B.
      • Barajas J.
      • Martinez-Rumayor A.
      • Januzzi J.L.
      Association between anxiety and C-reactive protein levels in stable coronary heart disease patients.
      ,
      • Gegenava T.
      • Gegenava M.
      • Kavtaradze G.
      C-reactive protein level correlation with depression and anxiety among patients with coronary artery disease.
      ), with previous evidence reporting that anxiety symptoms alone increase the probability for elevated CRP levels (
      • Liukkonen T.
      • Räsänen P.
      • Jokelainen J.
      • Leinonen M.
      • Järvelin M.R.
      • Meyer-Rochow V.B.
      • et al.
      The association between anxiety and C-reactive protein (CRP) levels: Results from the Northern Finland 1966 birth cohort study.
      ). Recent meta-analyses have also reported a high prevalence of elevated CRP levels in schizophrenia (
      • Fernandes B.S.
      • Steiner J.
      • Bernstein H.G.
      • Dodd S.
      • Pasco J.A.
      • Dean O.M.
      • et al.
      C-reactive protein is increased in schizophrenia but is not altered by antipsychotics: Meta-analysis and implications.
      ,
      • Miller B.J.
      • Culpepper N.
      • Rapaport M.H.
      C-reactive protein levels in schizophrenia: A review and meta-analysis.
      ), and previous research indicates an association between higher CRP levels in adolescence and schizophrenia at follow-up until age 27 years (
      • Metcalf S.A.
      • Jones P.B.
      • Nordstrom T.
      • Timonen M.
      • Mäki P.
      • Miettunen J.
      • et al.
      Serum C-reactive protein in adolescence and risk of schizophrenia in adulthood: A prospective birth cohort study.
      ). Therefore, evidence exists of an association of CRP with both anxiety and psychotic outcomes. While this may indicate a common inflammatory cause of both anxiety and psychosis (
      • Barker D.J.
      The fetal and infant origins of adult disease.
      ), an alternative hypothesis might be that chronic anxiety potentiates stress pathways, with downstream impact on risk for psychosis, and as such CRP could potentially mediate the association between anxiety and psychosis.
      Understanding the nature of the associations between anxiety in childhood and adolescence with subsequent psychosis would improve knowledge of potential underlying mechanisms of psychosis and provide novel treatment targets. To date, no studies have examined the prospective associations between persistent anxiety across childhood and adolescence and psychosis in adulthood. Further, there is no previous research investigating the mediating role of inflammatory markers in this association. Here, we examined the associations between persistent anxiety across childhood and adolescence with psychosis at age 24 years. Additionally, we investigated whether CRP measured at ages 9 and 15 years mediated these associations. We hypothesized that persistent high levels of anxiety would be a risk factor for psychosis in young adulthood and that CRP levels would mediate these associations.

      Methods and Materials

      Participants

      The Avon Longitudinal Study of Parents and Children (ALSPAC) is a birth cohort study in the United Kingdom examining the determinants of development, health, and disease during childhood and beyond (
      • Boyd A.
      • Golding J.
      • Macleod J.
      • Lawlor D.A.
      • Fraser A.
      • Henderson J.
      • et al.
      Cohort Profile: The ‘children of the 90s’—the index offspring of the Avon Longitudinal Study of Parents and Children.
      ,
      • Fraser A.
      • Macdonald-Wallis C.
      • Tilling K.
      • Boyd A.
      • Golding J.
      • Davey Smith G.
      • et al.
      Cohort Profile: The Avon Longitudinal Study of Parents and Children: ALSPAC mothers cohort.
      ). The ALSPAC study website contains details of all the data available (http://www.bristol.ac.uk/alspac/researchers/our-data/). The initial number of pregnancies enrolled was 14,541. Further details of this cohort are described in the Supplement. Ethical approval was obtained from the ALSPAC Law and Ethics Committee and local research ethics committees. Informed consent was obtained from participants following the recommendations of the ALSPAC Ethics and Law Committee at the time.

      Measures

      Persistent Anxiety Across Childhood and Adolescence

      The Development and Well-Being Assessment (DAWBA) (
      • Goodman R.
      • Ford T.
      • Richards H.
      • Gatward R.
      • Meltzer H.
      The Development and Well-Being Assessment: Description and initial validation of an integrated assessment of child and adolescent psychopathology.
      ) was administered as a parent-report questionnaire to capture child and adolescent psychopathology that corresponds with ICD-10 and DSM-IV criteria. Full details of the DAWBA are provided in the Supplement. We used the generalized anxiety dimension at ages 8, 10, and 13 years as a consistent measure of childhood and early adolescence anxiety, unlikely to be confounded by early symptoms of prodromal psychosis, which typically emerges later in adolescence or early adulthood (
      • Gogtay N.
      • Vyas N.S.
      • Testa R.
      • Wood S.J.
      • Pantelis C.
      Age of onset of schizophrenia: Perspectives from structural neuroimaging studies.
      ). The generalized anxiety dimension comprises two scores: 1) the generalized anxieties total score, representing the cognitive dimension (e.g. “child worries about disasters,” “child worries about health”); and 2) the generalized anxieties symptoms score, representing the symptomatology dimension (e.g., “child’s general anxieties lead to restlessness,” “child’s general anxieties lead to concentration problems”). We created a composite score of anxiety: (generalized anxieties + generalized anxieties symptoms)/2. We used this composite score to allow capture of two dimensions of anxiety (cognitive and symptomatology) rather than only one (see Table 1 for descriptive values for the variables of interest of this study, including the DAWBA composite score of anxiety).
      Table 1Descriptive Values of Sociodemographic and Health-Related Variables, Anxiety Measures, and Psychotic Disorder for the Study Sample
      8 Years Old10 Years Old13 Years Old24 Years Old
      n/n or Mean%/% or SDn/n or Mean%/% or SDn/n or Mean%/% or SDn/n or Mean%/% or SD
      Sociodemographic Factors
       Sex, females/males4007/423548.6%/51.4%3869/393849.6%/50.4%3444/346249.9%/50.1%2429/145862.5%/37.5%
       Ethnicity, non-White/White145/78081.8%/98.2%125/71061.7%/98.3%104/63231.6%/98.4%78/34052.2%/97.8%
       Birth weight, kg3.420.553.420.543.430.543.410.53
       FAI, total score4.064.103.934.013.864.013.613.84
       Maternal age at childbirth, years28.434.8229.034.5728.044.6829.454.56
       Gestational age, weeks39.072.3939.411.8539.442.3139.491.80
      Clinical Scores
       Meeting criteria of PD at 24 years old, Yes/No47/38421.2%/98.8%
       PEs at 24 years, Yes/No120/38423.1%/96.9%
       DAWBA anxiety composite score1.161.592.371.612.401.71
      9 Years Old15 Years Old
      Mean (SD)RangeMean (SD)Range
      Inflammatory Markers
       CRP, mg/L0.80 (2.88)67.431.26 (3.97)72.48
       CRP z-transformed scores0.00 (1.06)24.760.00 (1.05)19.15
       IL-61.29 (1.59)20.04
       IL-6 z-transformed scores0.00 (1.00)12.60
      n (%)n (%)n (%)n (%)
      Substance Use
       Cannabis use at 15 years old1347 (25.3%)3968 (74.7%)
      CRP, C-reactive protein; DAWBA, Development and Well-Being Assessment; FAI, Family Adversity Index; IL-6, interleukin 6; PD, psychotic disorder; PEs, psychotic experiences.

      Psychotic Outcomes at 24 Years

      Psychotic experiences (PEs) were identified through the semistructured Psychosis-like Symptom Interview (
      • Horwood J.
      • Salvi G.
      • Thomas K.
      • Duffy L.
      • Gunnell D.
      • Hollis C.
      • et al.
      IQ and non-clinical psychotic symptoms in 12-year-olds: Results from the ALSPAC birth cohort.
      ). PEs occurring in the past 6 months included the three main positive symptom domains: hallucinations, delusions, and thought interference. Interviewers rated PEs as not present, suspected, or definitely present. Cases of PEs were defined as individuals with definite PEs.
      Following previous research (
      • Sullivan S.A.
      • Kounali D.
      • Cannon M.
      • David A.S.
      • Fletcher P.C.
      • Holmans P.
      • et al.
      A population-based cohort study examining the incidence and impact of psychotic experiences from childhood to adulthood, and prediction of psychotic disorder.
      ,
      • Perry B.I.
      • Stochl J.
      • Upthegrove R.
      • Zammit S.
      • Wareham N.
      • Langenberg C.
      • et al.
      Longitudinal trends in childhood insulin levels and body mass index and associations with risks of psychosis and depression in young adults.
      ), we identified PD (i.e., meeting criteria for PD) at 24 years, a more restricted phenotype, which was defined as follows: 1) being rated as having definite PEs not attributable to sleep or fever; 2) having recurred regularly over the previous 6 months; 3) being reported as very distressing or having very negative impact on their social/occupational functioning. We selected the psychotic outcomes at age 24 to give a clear 11-year time frame between the last measure of anxiety (age 13) and subsequent psychosis.

      Inflammatory Markers at 9 and 15 Years

      Blood samples were collected from nonfasting participants during clinic assessment at age 9 around the same time of the day. At age 15, blood was drawn while fasting at a largely consistent time of day (i.e., at least 6 hours), limiting potential for diurnal effects on traits such as inflammatory markers (
      • Bell J.A.
      • Hamer M.
      • Richmond R.C.
      • Timpson N.J.
      • Carslake D.
      • Davey Smith G.
      Associations of device-measured physical activity across adolescence with metabolic traits: Prospective cohort study.
      ). Samples were immediately spun, frozen, and stored at −80 °C. There was no evidence of freeze-thaw cycles during storage. High-sensitivity CRP was measured at one time point at the same laboratory by automated particle-enhanced immunoturbidimetric assay (Roche UK). Additionally, interleukin 6 (IL-6) was measured by single enzyme-linked immunosorbent assay (R&D Systems). All assay coefficients of variation were <5%. CRP was available at ages 9 and 15, and this was the main inflammatory marker selected for this study owing to the availability of this marker at two time points, which allowed us to treat this variable as a mediator. IL-6 was available only at age 9, and thus it could not be treated as a mediator, but only as a contributing factor. Therefore, this was not included in the main analyses of this study; results for IL-6 are, however, given in Figure S1.

      Confounders

      Multiple family risk factors were assessed using the Family Adversity Index (FAI) during pregnancy (long index), at 2 years (long index), and at 4 years (short index). FAI includes early parenthood, housing and family conditions, maternal education, financial difficulties, parents’ relationship, maternal psychopathology, parents’ substance abuse, partner support, and social network. Points were summed at each time point for a total FAI score. We included this variable as a confounder, as early adversity is a well-established risk factor for poor mental health (
      • Wadman R.
      • Hiller R.M.
      • St Clair M.C.
      The influence of early familial adversity on adolescent risk behaviors and mental health: Stability and transition in family adversity profiles in a cohort sample.
      ). We also controlled for cannabis use at age 15, as adolescent cannabis use is related to anxiety and is a risk factor for PD and PEs (
      • Semple D.M.
      • McIntosh A.M.
      • Lawrie S.M.
      Cannabis as a risk factor for psychosis: Systematic review.
      ,
      • Kiburi S.K.
      • Molebatsi K.
      • Ntlantsana V.
      • Lynskey M.T.
      Cannabis use in adolescence and risk of psychosis: Are there factors that moderate this relationship? A systematic review and meta-analysis.
      ).
      Finally, relevant socioeconomic factors selected as covariates were child’s sex, gestational age, and ethnicity and maternal age when the child was born. Child’s sex, gestational age, ethnicity and maternal age at childbirth were selected because of their impact on psychosis and anxiety (
      • Harron K.
      • Gilbert R.
      • Fagg J.
      • Guttmann A.
      • van der Meulen J.
      Associations between pre-pregnancy psychosocial risk factors and infant outcomes: A population-based cohort study in England.
      ).

      Statistical Analysis

      First, we conducted latent class growth analyses using Mplus (version 8; Muthén & Muthén) to potentially identify differing levels of anxiety symptoms across childhood and adolescence. The indicator variables were DAWBA composite score of anxiety at ages 8, 10, and 13 years. Several models were fitted by increasing the number of classes (
      • Jung T.
      • Wickrama K.A.S.
      An introduction to latent class growth analysis and growth mixture modeling.
      ). The best-fitting classification model was chosen according to fit indices (i.e., Bayesian information criteria [BIC] and Vuong-Lo-Mendell-Rubin [VLMR] test) (
      • Jung T.
      • Wickrama K.A.S.
      An introduction to latent class growth analysis and growth mixture modeling.
      ). Lower BIC values suggest better model fit. A significant VLMR value suggests that a k-class model fits the data better than a K − 1 class model. Entropy was additionally used to select the best model fit; entropy with values approaching 1 indicates clear delineation of classes. Finally, to decide the optimal class solution, an emphasis was placed on large enough group sizes. Missing values owing to attrition were handled by the full information maximum likelihood estimation method (
      • Jung T.
      • Wickrama K.A.S.
      An introduction to latent class growth analysis and growth mixture modeling.
      ).
      Second, we investigated the prospective associations between persistent high levels of anxiety, identified by latent class growth analyses, and psychosis at age 24 using SPSS version 25 (IBM Corp). We conducted multinomial logistic regression analyses. The derived latent classes from the latent class growth analyses were included as predictor (with class 1, which referred to persistent low levels of anxiety, as reference) and psychosis at age 24 as the outcome. We included the two psychotic outcomes (i.e., PEs and PD) in separate models. Further, we tested first unadjusted associations, and then we controlled for all the confounders in the adjusted model. To deal with missingness, we conducted logistic regressions to identify significant factors associated with attrition. The individuals associated with attrition at age 24 were more often boys and had younger mothers, shorter gestational age, lower weight at birth, and higher socioeconomic levels (Table S1). Using the variables associated with selective dropout as the factors, we fitted a logistic regression model to determine weights for each individual using the inverse probability of response.
      Finally, to examine the potential role of CRP at ages 9 and 15 as a mediator in the association between persistent high levels of anxiety in childhood and adolescence with psychosis at age 24, we conducted path analysis in SPSS Amos 27 (IBM Corp.). Two separate path analyses were conducted for each psychotic outcome (i.e., PEs and PD). CRP levels at each time point (9 and 15 years) were standardized (z-transformed), and then we computed the mean score of CRP (z-transformed) at 9 and 15 years. Therefore, this new mean score ([CRP9 + CRP15]/2) was included as a mediator in the path analyses. The independent variable was dichotomized: 1 = class referring to persistent high levels of anxiety across time points; 0 = the other classes. We controlled for FAI, sex, birth weight, and cannabis use at age 15. We used bootstrapped bias-corrected 95% confidence intervals and p values for assessing the significance of the standardized indirect associations. Missing data were dealt with using the full information maximum likelihood method. Further, additional analyses to examine the potential role of IL-6 at age 9 in the associations between persistent anxiety and psychotic outcomes at age 24 were conducted.

      Results

      Latent Classes of Anxiety

      Table 2 shows VLMR, BIC, and entropy for all models assessed (2–6 classes). Overall, a 3-class model offered the best fit. VLMR showed a statistically significant difference for the 2-class, 3-class, and 4-class models. The 5-class model did not offer a significantly better fit than the 4-class model at the p < .01 level. Further, BIC decreased with the addition of each class, indicating a better model fit for more classes. This pattern is typically found in large samples (
      • Wiggins J.L.
      • Mitchell C.
      • Hyde L.W.
      • Monk C.S.
      Identifying early pathways of risk and resilience: The codevelopment of internalizing and externalizing symptoms and the role of harsh parenting.
      ). However, decreases in BIC became considerably smaller in 3-class compared with 2-class models. Finally, compared with the 4-class model, the 3-class model reported higher entropy value (near 1.0), which refers to classification precision. As this is important when importing classes for further analyses (
      • Asparouhov T.
      • Muthén B.
      Auxiliary variables in mixture modeling: A 3-step approach using Mplus.
      ), the 3-class model with an entropy of 0.864 was selected. In addition, the 3-class model provided large enough group sizes for each class (class 1 = 6331; class 2 = 1882; class 3 = 469), whereas 2 or 4-class models produced mixed or smaller groups (Table S2). The 3 derived classes of anxiety from the 3-class model are shown in Figure 1. The derived classes from the 2- and 4-class models appear in Table S2. In the 3-class model, class 3 represented persistent high levels of anxiety (5.4% of the sample), class 2 represented persistent moderate levels of anxiety (21.7% of the sample), and class 1 represented persistent low levels of anxiety (72.9% of the sample). Descriptive values for the covariates, psychotic measures, and inflammatory markers for each of the 3 classes appear in Table S3.
      Table 2BIC, VLMR Likelihood Test p Values, and Entropy for Classes 2–6 of the DAWBA Composite Score of Anxiety
      Composite Score of General AnxietiesBICVLMR p ValueEntropy
      2 Classes47499.749<.0010.853
      3 Classes45831.340.00140.864
      4 Classes45640.278.00070.817
      5 Classes43290.281.04810.884
      6 Classes43317.489.26100.833
      BIC, Bayesian information criterion; DAWBA, Development and Well-Being Assessment; VLMR, Vuong-Lo-Mendell-Rubin.
      Figure thumbnail gr1
      Figure 1Growth trajectories of anxiety across childhood to adolescence. The latent class growth analyses detected a best model fit for 3 classes. Class 1 (blue line on the bottom) represents individuals with persistent low levels of anxiety across time points. Class 2 (red line in the middle) represents individuals with persistent intermediate levels of anxiety. Class 3 (green line on the top) represents individuals with persistent high levels of anxiety across time points, which was the main focus of this study.

      Associations Between Persistent High Levels of Anxiety and Psychosis at 24 Years

      The weighted adjusted logistic regression model showed that persistent high levels of anxiety across time points (i.e., class 3) were significantly associated with PEs at age 24 years (odds ratio [OR] 2.02, 95% CI 1.26–3.23, p = .003) and PD at age 24 (OR 4.23, 95% CI 2.27–7.88, p = .001) (Table 3). To show the consistency and robustness of these findings, logistic regression analyses with other psychosis-related outcomes at age 24 are reported in Tables S4 through S6, where similar trends were found for broader defined psychotic outcomes. Further, to explore the specificity of our findings, logistic regression analyses with other relevant mental disorders (e.g., depression, anxiety, hypomania, substance abuse) in young adulthood were also performed. We found that persistent anxiety was also associated with depression and generalized anxiety at age 24. The ORs of these associations (i.e., 1.99 and 2.15, respectively) were lower than the OR reported in the association between persistent anxiety and PD (i.e., 4.23). Further, persistent high anxiety was not associated with hypomania, phobias, or substance abuse in young adulthood. This supports the specificity of the associations reported here in relation to psychosis at age 24 (see Tables S7–S10 for further details).
      Table 3Associations of Latent Classes of Anxiety With Psychotic-like Symptoms and Meeting Criteria of Psychotic Disorder at 24 Years
      Unadjusted ModelAdjusted Model
      OR95% CIp ValueOR95% CIp Value
      Psychotic Experiences at 24 Years
      General Anxiety Class 1 (Reference).035.011
      General Anxiety Class 20.9080.680–1.214.5150.9890.708–1.381.949
      General Anxiety Class 31.698
      Significant value.
      1.096–2.628
      Significant value.
      .018
      Significant value.
      2.022
      Significant value.
      1.265–3.232
      Significant value.
      .003
      Significant value.
      Sex1.344
      Significant value.
      1.014–1.781
      Significant value.
      .040
      Significant value.
      Gestational Age1.0070.929–1.091.868
      FAI Total Score1.0030.972–1.035.857
      Ethnicity1.7340.817–3.681.152
      Maternal Age at Childbirth1.0170.986–1.049.280
      Cannabis Use Ever at 15 Years Old1.673
      Significant value.
      1.239–2.260
      Significant value.
      .001
      Significant value.
      Meeting Criteria of Psychotic Disorder at 24 Years
      General Anxiety Class 1 (Reference).004.000
      General Anxiety Class 21.0230.651–1.609.9201.3920.809–2.394.232
      General Anxiety Class 32.656
      Significant value.
      1.479–4.767
      Significant value.
      .001
      Significant value.
      4.229
      Significant value.
      2.268–7.885
      Significant value.
      < .001
      Significant value.
      Sex0.6590.415–1.048.078
      Gestational Age0.9360.831–1.055.279
      FAI Total Score1.0380.989–1.090.129
      Ethnicity1.6730.500–5.589.403
      Maternal Age at Childbirth1.0010.952–1.054.959
      Cannabis Use Ever at 15 Years Old0.8470.475–1.512.575
      In relation to psychotic experiences at 24 years old, in addition to the significant associations observed between persistent anxiety (class 3) and the outcome, we also found that being a boy and cannabis use at 15 years old both were related to psychotic experiences at 24 years. However, concerning meeting criteria of psychotic disorder at 24 years old, none of the covariates included were associated with the outcome at 24 years old.
      FAI, Family Adversity Index; OR, odds ratio.
      a Significant value.

      Mediating Effect of Inflammatory Markers

      In examining whether the mean level of CRP at ages 9 and 15 years mediated the association between persistent anxiety and PEs at age 24, path analysis model fit indices indicated good model fit (χ2 = 3.26, p = 0.66, root mean square error of approximation 0, comparative fit index 1.00). Consistent with the adjusted logistic regression, persistent high levels of anxiety across time points were directly and significantly associated with PEs at age 24 (β = 0.028, p < .001). Direct associations are shown in Figure 2A. Further, we observed an indirect effect of the mean of CRP at ages 9 and 15 in the association between exposure and outcome (bias-corrected estimate −0.001, 95% CI −0.002 to 0.000, p = .013). The fact that the bias-corrected estimate includes the value 0 indicates certain uncertainty of these results. The mediating role of the mean of CRP at ages 9 and 15 in the association between persistent anxiety and PD at age 24 also demonstrated a good model fit (χ2 = 2.80, p = .73, root mean square error of approximation 0, comparative fit index 1.00). Persistent anxiety was directly and significantly associated with PD at age 24 (β = 0.040, p < .001). Direct associations are shown in Figure 2B. Further, an indirect effect of the mean of CRP at 9 and 15 in the association between exposure and outcome was found (bias-corrected estimate 0.001, 95% CI 0.001–0.002, p = .003). Additionally, path analyses with IL-6 at age 9 as mediator are provided in Figure S1. IL-6 at age 9 did not mediate the associations between persistent anxiety and any of the psychotic outcomes.
      Figure thumbnail gr2
      Figure 2Path diagram showing the main direct associations. Only the direct associations of the independent variable, mediating factor, and dependent variable are shown. Persistent high levels of anxiety represent the independent variable; the mean of C-reactive protein (CRP) levels at ages 9 and 15 represent the mediating factor; and psychotic experiences (A) and meeting criteria for psychotic disorder (B) at age 24 represent the outcomes. The covariates also included in this path analysis were sex, family adversity, birth weight, and cannabis use at age 15. Significant pathways are indicated by solid arrows.

      Discussion

      To our knowledge, this is the first longitudinal study to examine whether persistent high levels of anxiety across childhood and adolescence are associated with PEs and PD at age 24 and to investigate potential inflammatory mechanisms mediating this association. First, we detected a group of individuals characterized by persistent high levels of anxiety across childhood and adolescence. Second, persistent high levels of anxiety were associated with PEs and PD at 24 years. Finally, CRP levels mediated the prospective associations between persistent high levels of anxiety and psychosis at age 24 (i.e., PEs and PD).
      We identified 3 different trajectories of anxiety across childhood and adolescence. These were persistent low levels of anxiety, comprising more than 70% of the sample; persistent moderate levels of anxiety, representing approximately 20% of the sample; and persistent high levels of anxiety, representing 5% of the sample. These findings support recent research reporting the presence of 3 classes of individuals with distinct anxiety symptom trajectories in early adolescence (
      • Ahlen J.
      • Ghaderi A.
      Dimension-specific symptom patterns in trajectories of broad anxiety: A longitudinal prospective study in school-aged children.
      ,
      • Bushnell G.A.
      • Talati A.
      • Wickramaratne P.J.
      • Gameroff M.J.
      • Weissman M.M.
      Trajectories of childhood anxiety disorders in two generations at high risk.
      ,
      • Gutman L.M.
      • McMaster N.C.
      Gendered pathways of internalizing problems from early childhood to adolescence and associated adolescent outcomes.
      ). It appears that it is possible to identify a group of children and adolescents who experience persistent and high anxiety levels, and these could represent individuals at higher risk for later mental disorders, including psychosis.
      To our knowledge, this is the first study to report the associations between persistent high levels of anxiety across childhood and adolescence with psychosis at age 24. These findings contribute novel evidence that supports the role of anxiety in early stages as a precursor of psychosis. Interestingly, we found more robust associations of persistent high levels of anxiety with PD than with PEs. This suggests that persistent anxiety across childhood and adolescent might constitute a better indicator of the development of future formal PD, while PEs, which are far more common in the population than PD (
      • Kelleher I.
      • Wigman J.T.
      • Harley M.
      • O’Hanlon E.
      • Coughlan H.
      • Rawdon C.
      • et al.
      Psychotic experiences in the population: Association with functioning and mental distress.
      ), constitute a more heterogeneous group. As such, PEs may be associated with a wider range of potential risk factors in young adulthood. This is supported in our study by the fact that none of our covariates were associated with PD, while sex and cannabis use were both associated with PEs. Our findings build on the literature of an association between anxiety and psychosis, previously related to psychotic symptom severity, distress of positive symptoms, and associated prognosis and relapse (
      • Hartley S.
      • Barrowclough C.
      • Haddock G.
      Anxiety and depression in psychosis: A systematic review of associations with positive psychotic symptoms.
      ). However, much of this previous work is cross-sectional, or focused on persistence and generation of positive presence of childhood anxiety many years before the onset of psychosis. In one previous large longitudinal study, anxiety at a single time point (age 15 years) was not associated with PD in young adulthood (
      • Doering S.
      • Lichtenstein P.
      • Gillberg C.
      • Middeldorp C.M.
      • Bartels M.
      • Kuja-Halkola R.
      • et al.
      Anxiety at age 15 predicts psychiatric diagnoses and suicidal ideation in late adolescence and young adulthood: Results from two longitudinal studies.
      ). Our study indicates that it is the persistence of high anxiety that might constitute a risk for future psychosis, even after adjusting for multiple important confounders. Therefore, our findings highlight the potential causal role of persistent anxiety in the development of psychosis and potentially support the notion that genetic risk for the disorder may be associated with increased anxiety long before the onset of psychosis (
      • Hall J.
      Schizophrenia—an anxiety disorder?.
      ,
      • Jones H.J.
      • Stergiakouli E.
      • Tansey K.E.
      • Hubbard L.
      • Heron J.
      • Cannon M.
      • et al.
      Phenotypic manifestation of genetic risk for schizophrenia during adolescence in the general population.
      ). Further, our findings indicate that the associations of persistent high levels of anxiety with psychosis at age 24 could be specific to this mental disorder, as no associations with other relevant disorders, such as hypomania, phobias, or substance abuse, were found. However, the potential role of other persistent symptoms in the findings, particularly persistent depression, which is usually comorbid with anxiety (
      • Lamers F.
      • van Oppen P.
      • Comijs H.C.
      • Smit J.H.
      • Spinhoven P.
      • van Balkom A.J.
      • et al.
      Comorbidity patterns of anxiety and depressive disorders in a large cohort study: The Netherlands Study of Depression and Anxiety (NESDA).
      ), could also explain some of our findings.
      The mediating role that CRP exerted in the association between persistent high levels of anxiety across childhood and adolescence and both PEs and PD at 24 years is noteworthy. While inflammation in childhood is a risk factor for depression (
      • Chu A.L.
      • Stochl J.
      • Lewis G.
      • Zammit S.
      • Jones P.B.
      • Khandaker G.M.
      Longitudinal association between inflammatory markers and specific symptoms of depression in a prospective birth cohort.
      ), anxiety (
      • Khandaker G.M.
      • Zammit S.
      • Lewis G.
      • Jones P.B.
      Association between serum C-reactive protein and DSM-IV generalized anxiety disorder in adolescence: Findings from the ALSPAC cohort.
      ), and psychosis (
      • Perry B.I.
      • Upthegrove R.
      • Thompson A.
      • Marwaha S.
      • Zammit S.
      • Singh S.P.
      • et al.
      Dysglycaemia, inflammation and psychosis: Findings from the UK ALSPAC Birth Cohort.
      ), our novel approach was to identify a persistent clinically relevant phenotype (anxiety) and demonstrate relevance for psychosis, with potentially CRP mediating these associations. We found that the mean of CRP levels at ages 9 and 15 mediated the associations between persistent high anxiety and psychotic outcomes at age 24 (i.e., PEs and PD). Persistently high levels of anxiety may lead to increased levels of stress hormones, which would activate a chronic, low-grade inflammatory state (
      • Singh B.
      • Chaudhuri T.K.
      Role of C-reactive protein in schizophrenia: An overview.
      ). Within this context, the CRP plasma levels may indicate nonresolving inflammation, activation of microglia, and subsequent downstream effects on brain structure and function via disordered synaptic pruning affecting crucial neurodevelopmental stages, ultimately leading to psychosis (
      • Fernandes B.S.
      • Steiner J.
      • Bernstein H.G.
      • Dodd S.
      • Pasco J.A.
      • Dean O.M.
      • et al.
      C-reactive protein is increased in schizophrenia but is not altered by antipsychotics: Meta-analysis and implications.
      ,
      • Miller B.J.
      • Culpepper N.
      • Rapaport M.H.
      C-reactive protein levels in schizophrenia: A review and meta-analysis.
      ). Our findings provide new evidence suggesting that persistent high levels of anxiety are associated with psychosis, and this may be mediated by elevated CRP levels. However, activation of stress pathways and inflammation is only one potential explanation of these associations, and further research should disentangle in more detail this potential mechanism, including interrogation of genetic factors and further assessment of other environmental factors. In addition, further interrogation of directionality of association would need to be explored in further observational and experimental medicine studies within inflammatory markers. Finally, we did not find a mediating role of IL-6 at age 9 in any of these associations, which is in contrast to increasing evidence of IL-6 in relation to schizophrenia with anxiety (
      • Osimo E.F.
      • Perry B.I.
      • Cardinal R.N.
      • Lynall M.E.
      • Lewis J.
      • Kudchadkar A.
      • et al.
      Inflammatory and cardiometabolic markers at presentation with first episode psychosis and long-term clinical outcomes: A longitudinal study using electronic health records.
      ). Studies including IL-6 at more than one time point and at later stages are required to further mechanistic understanding of persistent anxiety and psychosis at age 24.
      This study has several strengths, including the large population-based sample, the longitudinal design, and the inclusion of meeting criteria for both PD and PEs at age 24. There are also some limitations. First, other potential contributing factors, such as depression, cognition, social interactions, obstetric complications, or factors associated with CRP, such as viral infections, smoking, or body mass index, were left unexplored. Second, our study was unable to use anxiety data in individuals older than 13 years, and thus we cannot comment on a trajectory of future anxiety disorders after this age group. Third, the attrition rate was significant. However, we used procedures to ensure representativeness of our results. Fourth, anxiety identified at age 13 may be part of prodromal symptoms in an early-onset group; however, our premise is that persistence of anxiety was the exposure of interest, and it is unlikely that anxiety at age 9 is conflated with prodromal psychosis. Fifth, only 47 individuals presented with criteria for PD, comprising 1.2% of our sample at age 24. However, this prevalence is representative of the general population. Sixth, we focused on the impact of persistent high anxiety on psychosis, and the effect of other potential anxiety trajectories, such as worsening anxiety, were left unexplored. Future studies should focus on examining also how worsening anxiety might be associated with psychosis. Finally, our results concerning the mediating role of CRP were more robust in relation to PD compared with PEs. For instance, we found a positive indirect effect between persistent anxiety and PD, which was congruent with the positive direct effects observed between exposure, mediator, and outcome. However, in relation to PEs, we found a negative indirect effect, despite direct effects between the three variables being reported. Therefore, our results in relation to PEs should be cautiously interpreted. One possible explanation for the inconsistency of the direction of the indirect effect in PEs could be that other potential mediators would explain the negative indirect effect observed. As previously reported in the discussion, PEs are associated with a wider range of risk factors than PD, which would subsequently impact our findings. Further, the bias-corrected estimate in relation to the indirect effect for PEs contains the value 0, and thus there are limited conclusions that can be drawn. This supports the need for future studies to disentangle other potential mechanisms underlying this association.
      In summary, our findings showed that persistent high levels of anxiety across childhood and adolescence are associated with psychosis at age 24. Further, CRP levels at ages 9 and 15 had a mediating role in these prospective associations. These findings suggest that persistent high levels of anxiety predate and could be considered a potentially malleable phenotype for developing psychosis at age 24, which could inform targeted prevention strategies. Moreover, a nonresolving proinflammatory mechanism might partially explain how persistent anxiety might lead to subsequent psychosis and provides further support to the role of inflammation in mechanistic pathways to psychosis. These results could contribute to the design of more personalized and indicated prevention in psychosis, such as early diagnosis and management of adolescent anxiety and possibly novel treatments targeted at inflammation.

      Acknowledgments and Disclosures

      This work was supported by the Medical Research Council and Wellcome Trust (Grant No. 217065/Z/19/Z ). The University of Bristol provided core support for ALSPAC.
      We thank all the families who took part in this study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes interviewers, computer and laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists, and nurses.
      The authors report no biomedical financial interests or potential conflicts of interest.

      Supplementary Material

      References

        • Radua J.
        • Ramella-Cravaro V.
        • Ioannidis J.P.A.
        • Reichenberg A.
        • Phiphopthatsanee N.
        • Amir T.
        • et al.
        What causes psychosis? An umbrella review of risk and protective factors.
        World Psychiatry. 2018; 17: 49-66
        • Whiteford H.A.
        • Degenhardt L.
        • Rehm J.
        • Baxter A.J.
        • Ferrari A.J.
        • Erskine H.E.
        • et al.
        Global burden of disease attributable to mental and substance use disorders: findings from the Global Burden of Disease Study 2010.
        Lancet. 2013; 382: 1575-1586
        • Kirkbride J.B.
        • Errazuriz A.
        • Croudace T.J.
        • Morgan C.
        • Jackson D.
        • Boydell J.
        • et al.
        Incidence of schizophrenia and other psychoses in England, 1950-2009: A systematic review and meta-analyses.
        PLoS One. 2012; 7e31660
        • Cloutier M.
        • Aigbogun M.S.
        • Guerin A.
        • Nitulescu R.
        • Ramanakumar A.V.
        • Kamat S.A.
        • et al.
        The economic burden of schizophrenia in the United States in 2013.
        J Clin Psychiatry. 2016; 77: 764-771
        • Fusar-Poli P.
        • Rutigliano G.
        • Stahl D.
        • Davies C.
        • Bonoldi I.
        • Reilly T.
        • et al.
        Development and validation of a clinically based risk calculator for the transdiagnostic prediction of psychosis.
        JAMA Psychiatry. 2017; 74: 493-500
        • van Os J.
        • Rutten B.P.
        • Poulton R.
        Gene-environment interactions in schizophrenia: Review of epidemiological findings and future directions.
        Schizophr Bull. 2008; 34: 1066-1082
        • Dean K.
        • Murray R.M.
        Environmental risk factors for psychosis.
        Dialogues Clin Neurosci. 2005; 7: 69-80
        • Rutter M.
        • Kim-Cohen J.
        • Maughan B.
        Continuities and discontinuities in psychopathology between childhood and adult life.
        J Child Psychol Psychiatry. 2006; 47: 276-295
        • Correll C.U.
        • Hauser M.
        • Auther A.M.
        • Cornblatt B.A.
        Research in people with psychosis risk syndrome: A review of the current evidence and future directions.
        J Child Psychol Psychiatry. 2010; 51: 390-431
        • Garety P.A.
        • Kuipers E.
        • Fowler D.
        • Freeman D.
        • Bebbington P.E.
        A cognitive model of the positive symptoms of psychosis.
        Psychol Med. 2001; 31: 189-195
        • Freeman D.
        • Fowler D.
        Routes to psychotic symptoms: Trauma, anxiety and psychosis-like experiences.
        Psychiatry Res. 2009; 169: 107-112
        • Ciapparelli A.
        • Paggini R.
        • Marazziti D.
        • Carmassi C.
        • Bianchi M.
        • Taponecco C.
        • et al.
        Comorbidity with axis I anxiety disorders in remitted psychotic patients 1 year after hospitalization.
        CNS Spectr. 2007; 12: 913-919
        • Bosanac P.
        • Mancuso S.G.
        • Castle D.J.
        Anxiety symptoms in psychotic disorders: Results from the Second Australian National Mental Health Survey.
        Clin Schizophr Relat Psychoses. 2016; 10: 93-100
        • Yamasaki S.
        • Usami S.
        • Sasaki R.
        • Koike S.
        • Ando S.
        • Kitagawa Y.
        • et al.
        The association between changes in depression/anxiety and trajectories of psychotic-like experiences over a year in adolescence.
        Schizophr Res. 2018; 195: 149-153
        • Freeman D.
        • Stahl D.
        • McManus S.
        • Meltzer H.
        • Brugha T.
        • Wiles N.
        • et al.
        Insomnia, worry, anxiety and depression as predictors of the occurrence and persistence of paranoid thinking.
        Soc Psychiatry Psychiatr Epidemiol. 2012; 47: 1195-1203
        • Goodwin R.D.
        • Fergusson D.M.
        • Horwood L.J.
        Panic attacks and psychoticism.
        Am J Psychiatry. 2004; 161: 88-92
        • Beesdo-Baum K.
        • Knappe S.
        Developmental epidemiology of anxiety disorders.
        Child Adolesc Psychiatr Clin N Am. 2012; 21: 457-478
        • Copeland W.E.
        • Angold A.
        • Shanahan L.
        • Costello E.J.
        Longitudinal patterns of anxiety from childhood to adulthood: the Great Smoky Mountains Study.
        J Am Acad Child Adolesc Psychiatry. 2014; 53: 21-33
        • Copeland W.E.
        • Shanahan L.
        • Costello E.J.
        • Angold A.
        Childhood and adolescent psychiatric disorders as predictors of young adult disorders.
        Arch Gen Psychiatry. 2009; 66: 764-772
        • Doering S.
        • Lichtenstein P.
        • Gillberg C.
        • Middeldorp C.M.
        • Bartels M.
        • Kuja-Halkola R.
        • et al.
        Anxiety at age 15 predicts psychiatric diagnoses and suicidal ideation in late adolescence and young adulthood: Results from two longitudinal studies.
        BMC Psychiatry. 2019; 19: 363
        • Upthegrove R.
        • Khandaker G.M.
        Cytokines, oxidative stress and cellular markers of inflammation in schizophrenia.
        Curr Top Behav Neurosci. 2020; 44: 49-66
        • Corsi-Zuelli F.
        • Deakin B.
        Impaired regulatory T cell control of astroglial overdrive and microglial pruning in schizophrenia.
        Neurosci Biobehav Rev. 2021; 125: 637-653
        • Liukkonen T.
        • Räsänen P.
        • Jokelainen J.
        • Leinonen M.
        • Järvelin M.R.
        • Meyer-Rochow V.B.
        • et al.
        The association between anxiety and C-reactive protein (CRP) levels: Results from the Northern Finland 1966 birth cohort study.
        Eur Psychiatry. 2011; 26: 363-369
        • Bullmore E.T.
        • Lynall M.E.
        Immunologic therapeutics and psychotic disorders.
        Biol Psychiatry. 2014; 75: 260-261
        • Bankier B.
        • Barajas J.
        • Martinez-Rumayor A.
        • Januzzi J.L.
        Association between anxiety and C-reactive protein levels in stable coronary heart disease patients.
        Psychosomatics. 2009; 50: 347-353
        • Gegenava T.
        • Gegenava M.
        • Kavtaradze G.
        C-reactive protein level correlation with depression and anxiety among patients with coronary artery disease.
        Georgian Med News. 2011; 194: 34-37
        • Fernandes B.S.
        • Steiner J.
        • Bernstein H.G.
        • Dodd S.
        • Pasco J.A.
        • Dean O.M.
        • et al.
        C-reactive protein is increased in schizophrenia but is not altered by antipsychotics: Meta-analysis and implications.
        Mol Psychiatry. 2016; 21: 554-564
        • Miller B.J.
        • Culpepper N.
        • Rapaport M.H.
        C-reactive protein levels in schizophrenia: A review and meta-analysis.
        Clin Schizophr Relat Psychoses. 2014; 7: 223-230
        • Metcalf S.A.
        • Jones P.B.
        • Nordstrom T.
        • Timonen M.
        • Mäki P.
        • Miettunen J.
        • et al.
        Serum C-reactive protein in adolescence and risk of schizophrenia in adulthood: A prospective birth cohort study.
        Brain Behav Immun. 2017; 59: 253-259
        • Barker D.J.
        The fetal and infant origins of adult disease.
        BMJ. 1990; 301: 1111
        • Boyd A.
        • Golding J.
        • Macleod J.
        • Lawlor D.A.
        • Fraser A.
        • Henderson J.
        • et al.
        Cohort Profile: The ‘children of the 90s’—the index offspring of the Avon Longitudinal Study of Parents and Children.
        Int J Epidemiol. 2013; 42: 111-127
        • Fraser A.
        • Macdonald-Wallis C.
        • Tilling K.
        • Boyd A.
        • Golding J.
        • Davey Smith G.
        • et al.
        Cohort Profile: The Avon Longitudinal Study of Parents and Children: ALSPAC mothers cohort.
        Int J Epidemiol. 2013; 42: 97-110
        • Goodman R.
        • Ford T.
        • Richards H.
        • Gatward R.
        • Meltzer H.
        The Development and Well-Being Assessment: Description and initial validation of an integrated assessment of child and adolescent psychopathology.
        J Child Psychol Psychiatry. 2000; 41: 645-655
        • Gogtay N.
        • Vyas N.S.
        • Testa R.
        • Wood S.J.
        • Pantelis C.
        Age of onset of schizophrenia: Perspectives from structural neuroimaging studies.
        Schizophr Bull. 2011; 37: 504-513
        • Horwood J.
        • Salvi G.
        • Thomas K.
        • Duffy L.
        • Gunnell D.
        • Hollis C.
        • et al.
        IQ and non-clinical psychotic symptoms in 12-year-olds: Results from the ALSPAC birth cohort.
        Br J Psychiatry. 2008; 193: 185-191
        • Sullivan S.A.
        • Kounali D.
        • Cannon M.
        • David A.S.
        • Fletcher P.C.
        • Holmans P.
        • et al.
        A population-based cohort study examining the incidence and impact of psychotic experiences from childhood to adulthood, and prediction of psychotic disorder.
        Am J Psychiatry. 2020; 177: 308-317
        • Perry B.I.
        • Stochl J.
        • Upthegrove R.
        • Zammit S.
        • Wareham N.
        • Langenberg C.
        • et al.
        Longitudinal trends in childhood insulin levels and body mass index and associations with risks of psychosis and depression in young adults.
        JAMA Psychiatry. 2021; 78: 416-425
        • Bell J.A.
        • Hamer M.
        • Richmond R.C.
        • Timpson N.J.
        • Carslake D.
        • Davey Smith G.
        Associations of device-measured physical activity across adolescence with metabolic traits: Prospective cohort study.
        PLoS Med. 2018; 15e1002649
        • Wadman R.
        • Hiller R.M.
        • St Clair M.C.
        The influence of early familial adversity on adolescent risk behaviors and mental health: Stability and transition in family adversity profiles in a cohort sample.
        Dev Psychopathol. 2020; 32: 437-454
        • Semple D.M.
        • McIntosh A.M.
        • Lawrie S.M.
        Cannabis as a risk factor for psychosis: Systematic review.
        J Psychopharmacol. 2005; 19: 187-194
        • Kiburi S.K.
        • Molebatsi K.
        • Ntlantsana V.
        • Lynskey M.T.
        Cannabis use in adolescence and risk of psychosis: Are there factors that moderate this relationship? A systematic review and meta-analysis.
        Subst Abus. 2021; 42: 527-542
        • Harron K.
        • Gilbert R.
        • Fagg J.
        • Guttmann A.
        • van der Meulen J.
        Associations between pre-pregnancy psychosocial risk factors and infant outcomes: A population-based cohort study in England.
        Lancet Public Health. 2021; 6: e97-e105
        • Jung T.
        • Wickrama K.A.S.
        An introduction to latent class growth analysis and growth mixture modeling.
        Social and Personality Psychology Compass. 2008; 2: 302-317
        • Wiggins J.L.
        • Mitchell C.
        • Hyde L.W.
        • Monk C.S.
        Identifying early pathways of risk and resilience: The codevelopment of internalizing and externalizing symptoms and the role of harsh parenting.
        Dev Psychopathol. 2015; 27: 1295-1312
        • Asparouhov T.
        • Muthén B.
        Auxiliary variables in mixture modeling: A 3-step approach using Mplus.
        (Available at:)
        • Ahlen J.
        • Ghaderi A.
        Dimension-specific symptom patterns in trajectories of broad anxiety: A longitudinal prospective study in school-aged children.
        Dev Psychopathol. 2020; 32: 31-41
        • Bushnell G.A.
        • Talati A.
        • Wickramaratne P.J.
        • Gameroff M.J.
        • Weissman M.M.
        Trajectories of childhood anxiety disorders in two generations at high risk.
        Depress Anxiety. 2020; 37: 521-531
        • Gutman L.M.
        • McMaster N.C.
        Gendered pathways of internalizing problems from early childhood to adolescence and associated adolescent outcomes.
        J Abnorm Child Psychol. 2020; 48: 703-718
        • Kelleher I.
        • Wigman J.T.
        • Harley M.
        • O’Hanlon E.
        • Coughlan H.
        • Rawdon C.
        • et al.
        Psychotic experiences in the population: Association with functioning and mental distress.
        Schizophr Res. 2015; 165: 9-14
        • Hartley S.
        • Barrowclough C.
        • Haddock G.
        Anxiety and depression in psychosis: A systematic review of associations with positive psychotic symptoms.
        Acta Psychiatr Scand. 2013; 128: 327-346
        • Hall J.
        Schizophrenia—an anxiety disorder?.
        Br J Psychiatry. 2017; 211: 262-263
        • Jones H.J.
        • Stergiakouli E.
        • Tansey K.E.
        • Hubbard L.
        • Heron J.
        • Cannon M.
        • et al.
        Phenotypic manifestation of genetic risk for schizophrenia during adolescence in the general population.
        JAMA Psychiatry. 2016; 73: 221-228
        • Lamers F.
        • van Oppen P.
        • Comijs H.C.
        • Smit J.H.
        • Spinhoven P.
        • van Balkom A.J.
        • et al.
        Comorbidity patterns of anxiety and depressive disorders in a large cohort study: The Netherlands Study of Depression and Anxiety (NESDA).
        J Clin Psychiatry. 2011; 72: 341-348
        • Chu A.L.
        • Stochl J.
        • Lewis G.
        • Zammit S.
        • Jones P.B.
        • Khandaker G.M.
        Longitudinal association between inflammatory markers and specific symptoms of depression in a prospective birth cohort.
        Brain Behav Immun. 2019; 76: 74-81
        • Khandaker G.M.
        • Zammit S.
        • Lewis G.
        • Jones P.B.
        Association between serum C-reactive protein and DSM-IV generalized anxiety disorder in adolescence: Findings from the ALSPAC cohort.
        Neurobiol Stress. 2016; 4: 55-61
        • Perry B.I.
        • Upthegrove R.
        • Thompson A.
        • Marwaha S.
        • Zammit S.
        • Singh S.P.
        • et al.
        Dysglycaemia, inflammation and psychosis: Findings from the UK ALSPAC Birth Cohort.
        Schizophr Bull. 2019; 45: 330-338
        • Singh B.
        • Chaudhuri T.K.
        Role of C-reactive protein in schizophrenia: An overview.
        Psychiatry Res. 2014; 216: 277-285
        • Osimo E.F.
        • Perry B.I.
        • Cardinal R.N.
        • Lynall M.E.
        • Lewis J.
        • Kudchadkar A.
        • et al.
        Inflammatory and cardiometabolic markers at presentation with first episode psychosis and long-term clinical outcomes: A longitudinal study using electronic health records.
        Brain Behav Immun. 2021; 91: 117-227

      Linked Article

      • Persistent Child and Adolescent Anxiety Predicts Development of Psychotic Disorders via Elevated Inflammation
        Biological PsychiatryVol. 92Issue 4
        • Preview
          Psychotic disorders (PDs) are a common group of conditions that include schizophrenia, schizoaffective disorder, and brief psychotic episodes. These disorders have a varied and complex etiology comprising genetic and environmental risk factors. PDs such as schizophrenia often have a prodromal stage during adolescence, and subsequent diagnosis hinges on symptoms presenting over the course of several years, but the manifestation of these symptoms is likely a consequence of brain changes in the years predating their onset.
        • Full-Text
        • PDF