Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Apr 1.
Published in final edited form as: J Affect Disord. 2023 Jan 28;326:193–197. doi: 10.1016/j.jad.2023.01.090

Maternal FGF2 Levels Associated with Child Anxiety and Depression Symptoms Through Child FGF2 Levels

Eli R Lebowitz 1,2,+, Carla E Marin 1,2, Meital Orbach 1,2,+, Natalina Salmaso 3, Flora M Vaccarino 1,4,5,*, Wendy K Silverman 1,2,6,*
PMCID: PMC10104478  NIHMSID: NIHMS1873101  PMID: 36717031

Abstract

Background:

Recent research implicates fibroblast growth factor 2 (FGF2) in anxiety and depressive symptoms of childhood. This study is the first to examine an intergenerational pathway linking FGF2 levels in mothers to FGF2 levels in children, and to the children’s anxiety and depressive symptoms.

Methods:

We assayed serum FGF2 in 259 mothers and their children, with a range of anxiety and depressive symptoms: 194 were mothers of clinic-referred anxious and depressed children; 65 were mothers of non-referred children. We examined associations between FGF2 levels in mothers and children, and anxiety and depression symptoms. We used structural equation modeling (SEM) to examine associations between maternal and child FGF2 levels, and between maternal and child FGF2 levels and symptoms of anxiety and depression in and children.

Results:

FGF2 levels in mothers and children were significantly positively correlated. Children’s FGF2 levels were significantly negatively correlated with their ratings of anxiety and depression. Results of the SEM model showed that increases in maternal FGF2 levels were significantly associated with increases in child FGF2, which in turn was associated with decreases in child anxiety and child depression, controlling for maternal anxiety and depression.

Limitations:

We relied on self-reported ratings of anxiety and depression, and on a single measurement of FGF2 levels for each participant.

Conclusions:

Our results point to a role for FGF2 in the intergenerational transmission of risk for, and resilience to, anxiety and depression in youth.

Keywords: Anxiety, Children, Depression, Fibroblast Growth Factor 2

Introduction

Anxiety and depressive disorders are highly prevalent among youth, with rates that have doubled since the COVID-19 pandemic, surpassing 20% for anxiety and 25% for depression (Racine et al., 2021). They frequently co-occur, suggesting shared underlying mechanisms and pathophysiology. Research implicates fibroblast growth factor 2 (FGF2), as one common biological substrate (Deng, Deng, Zhang, & Tang, 2019; Salmaso et al., 2016; Turner, Akil, Watson, & Evans, 2006)

Lebowitz and colleagues (2021) recently reported the first evidence linking serum FGF2 levels to indicators of anxiety and depression in youth. Children referred for anxiety and depression had significantly lower FGF2, compared with non-referred children. FGF2 levels were significantly negatively correlated with indicators of anxiety and depression, and positively correlated with reward responsiveness, which is impaired in depression.

In addition to these novel and significant group differences and correlations, confirmatory factor analysis (CFA) supported the contribution of FGF2 to a latent variable representing anxiety/depression. A well-fitted model emerged with FGF2 loading negatively on anxiety/depression, explaining approximately 5% of its variance.

These data further align with research suggesting alterations in growth factor systems can impact risk of psychopathology (Kiraly et al., 2017; Terwisscha van Scheltinga, Bakker, & Kahn, 2010) through subtle neurodevelopmental alterations to brain structure. The data highlight the potential role of FGF2 in intergenerational transmission of, or buffering against, vulnerability to anxiety and depression. Indeed, it is well-established that rearing environment and childhood events impact neurodevelopment with long-term effects (Fumagalli, Bedogni, Slotkin, Racagni, & Riva, 2005); FGF2 may play a role in such impact.

Converging lines of evidence support our hypothesis of an intergenerational pathway. FGFs are deeply implicated in neurodevelopment, regulating neurogenesis and neuronal repair, (Reuss & von Bohlen und Halbach, 2003) while genetic alterations in the FGF2 system lead to structural brain changes that impact risk of psychopathology (O’Donovan et al., 2009). In rodents, FGF2 gene-knockout likewise alters brain structure, including in cortical regions (Turner et al., 2006). Studies of environmental factors and manipulation also support FGF2’s role in vulnerability to anxiety and depression. FGF2 mediates rodent’s neurobiological outcome following brain lesions, evidenced by heightened recovery in pups who received direct FGF2 administration, compared with lesioned pups who did not (Kolb & Gibb, 2007). Direct administration is not required however, for FGF2 to play a role in brain recovery. Repeated stroking, simulating parental rearing behavior, increased FGF2 release and improved recovery. Furthermore, stroking a pregnant mother rat attenuated the impact of future lesions on her offspring, as did direct administration of FGF2 to the pregnant rat (Kolb & Gibb, 2007). Additionally, prenatal stress in rat dams impacted offspring’s FGF2 expression in multiple brain regions and altered FGF2 gene expression regulation in response to stress (Fumagalli et al., 2005). Exposure to glucocorticoid stress hormones during late pregnancy also led to persistent changes in the regulatory machinery of FGF2 in rats (Molteni et al., 2001).

These findings demonstrate FGF2’s key roles influencing vulnerability to anxiety and depression through intergenerational processes. Here, we take an important next step toward elucidating these intergenerational pathways. Specifically, we examine pathways linking FGF2 levels in mothers to FGF2 levels in their children, and to their children’s anxiety and depression symptoms.

Method

Participants

Participants were 259 mothers (26 to 64 years, M = 42.64 years, SD = 6.39) and their children with a broad range of anxiety and depressive symptoms: 194 were mothers of clinic-referred anxious and/or depressed children (47% female assigned at birth, 6 to 11 years, M = 8.59 years, SD = 1.70; mothers were 26 to 64 years, M = 43.89 years, SD = 5.59); 65 were mothers of non-referred children without psychiatric history (48% female assigned at birth, 6 to 12 years, M = 8.65 years, SD = 1.95; mothers were 27 to 55 years, M = 39.00 years, SD = 7.19). 149 (35%) of the children in this study participated in Lebowitz and colleagues (2021). Non-referred participants were recruited from the community and compensated $50 for their participation. Table S1 (see supplement) summarizes participant demographic characteristics.

Measures

Anxiety

Child Anxiety was assessed with the child-report Screen for Child Anxiety Related Disorders (SCARED; Birmaher et al.,1997). SCARED contains 41 items, rated on a 3-point scale. SCARED is widely used with established psychometric properties including internal consistency, test-retest reliability and validity (Monga et al., 2000). Internal consistency in the current sample was .91 for both Cronbach’s α and McDonald’s ω.

Maternal Anxiety was assessed using the Beck Anxiety Inventory (BAI; Beck, Epstein, Brown, & Steer, 1988). BAI is a 21-item questionnaire rated on a four-point scale. Scores range from 0 to 63. BAI is widely used with well-established psychometric properties including internal consistency, test-retest reliability and validity (Fydrich, Dowdall, & Chambless, 1992). Internal consistency in the current sample was .96 for both Cronbach’s α and McDonald’s ω.

Depression

Child Depression was assessed using the child-reported Children’s Depression Inventory (CDI; Kovacs, 1980). CDI contains 27 items, each having three response options (e.g., ‘I am sad once in a while, I am sad many times, and I am sad all the time’). Total scores range from 0 to 54. CDI is widely used with established psychometric properties including internal consistency, test-retest reliability and validity (Logan & Goetsch, 1993). Internal consistency in the current sample was α = .88 and McDonald’s ω = .86.

Maternal Depression was assessed using the Beck Depression Inventory (BDI; Beck, Epstein, Brown, & Steer, 1988). BDI is a 21-item self-report questionnaire rated on a four-point rating scale. Scores range from 0 to 63. BDI is widely used with well-established psychometric properties including internal consistency, test-retest reliability and validity. Internal consistency for in the current sample was .90 for both α and ω.

Serum FGF2 Levels

Samples were collected using VACUETTE® serum clot activator tubes (456073P, GREINER BIO-ONE). Samples were collected between 12PM and 3PM, after participants had arrived at the lab and completed other study tasks. After allowing samples to clot for 30 minutes at room temperature they were centrifuged for 30 minutes at 4000 rpm, aliquoted and stored at −80°C. FGF2 levels were analyzed using a high-sensitivity enzyme linked immunosorbent assay kit (HSFB00D; R&D Systems). Measurements were performed in triplicates and concentrations calculated using Prism7 (Graph-Pad) according to relevant standard curves.

Procedures

The study was approved by the University’s Institutional Review Board. Following initial telephone-screening, children and mothers were invited for in-person evaluations. After providing informed assent/consent, children and mothers completed a standardized questionnaire battery and blood samples were collected by trained phlebotomists.

Data Analysis

To account for the significant differences between the clinic-and non-referred participants, we controlled for maternal age (t(248) = 5.59, p < .001), maternal (χ2(5) = 53.09, p < .001) and child race (χ2(5) = 58.26, p < .001), marital status (χ2(3) = 37.25, p < .001), and family income (χ2(2) = 47.17, p < .001), in relevant subsequent analyses. We used multiple regression with dummy coded variables (Jaccard & Tursi, 2003) to examine mean differences on study variables between the clinic-referred and non-referred samples. The regression coefficient associated with the ‘Clinic vs non-Referred’ variable represents the mean difference between the group scored 1 (non-referred sample) and the group scored 0 (clinic -referred sample). We used structural equation modeling (SEM) to examine associations between maternal and child FGF2 levels, and between maternal and child FGF2 levels and symptoms of anxiety and depression in and children. We followed recommendations by Kline (2016) to evaluate model fit.

Results

Preliminary Analyses

Robust multivariate outlier analyses were conducted using the projection-type method (Wilcox, 2011). This method revealed 15 outliers. Analyses were conducted with and without outliers and the statistical significance of all paths of interest remained the same. Analyses are thus presented with the outliers. To accommodate missing data and non-normality, we used full information maximum likelihood and a maximum likelihood estimator with robust standard errors, respectively as implemented in MPlus 8.5 (Muthén & Muthén, 1998–2012).

Main Analyses

Descriptive Statistics and Bivariate Correlations.

Descriptive statistics and correlations are presented in Table 1. We examined associations between maternal anxiety, depression, and FGF2, and their respective associations with child FGF2, anxiety and depression. Maternal and child FGF2 levels were significantly positively correlated. Maternal FGF2 levels were not significantly correlated with the other study variables. Children’s FGF2 levels were significantly negatively correlated with their self-rated anxiety and depression; children’s FGF2 levels were not significantly correlated with mothers’ anxiety or depression.

Table 1.

Descriptive Statistics and Correlations for Study Variables

M SD 1 2 3 4 5 6
1. FGF2 mother 4.81 4.43 -
2. FGF2 child 5.33 5.32 .37*** -
3. BAI 7.65 9.36 .03 −.13 -
4. BDI 6.66 6.95 .07 −.09 .59*** -
5. SCARED 27.36 14.08 −.05 −.24*** .08 .11 -
6. CDI 20.62 9.34 −.09 −.28*** .15* .08 .39*** -
Open in a new tab

Note. BAI = Beck Anxiety Inventory; BDI = Beck Depression Inventory; SCARED = Screen for Child Anxiety and Related Emotional Disorders; CDI = Children’s Depression Inventory.

*

p < .05;

**

p < .01;

***

p < .001

Multiple regression showed that controlling for maternal age, maternal and child race, marital status, and family income, maternal FGF2 levels were not significantly different between the clinic-referred and the non-referred samples (β = 0.86, SE = .79, t = 1.09, p = .28). There were also no significant differences between the clinic- and non-referred samples on children’s FGF2 levels (β = 1.60, SE = 1.11, t = 1.45, p = .15), or maternal depression (β = −0.92, SE = 1.11, t = −0.82, p = .41). Clinic-referred mothers rated themselves significantly higher in anxiety than non-referred mothers (β = −3.65, SE = 1.52, t = −2.41, p = .02). Clinic referred children rated themselves significantly higher in anxiety (β= −13.69, SE = 2.30, t = −5.94, p < .001) and depression (β= −17.67, SE = −0.84, t = −17.30, p < .001) than non-referred children.

Structural Model

Fit statistics for the SEM model (see Figure 1) indicated excellent model fit (i.e., chi square = 1.05, p = .90; RMSEA = .000; CFI = 1.0; SRMR = .02). Results showed that increases in maternal FGF2 (path coefficient = 0.43, 95% CI 0.16, 0.69, p = .002) were significantly associated with increases in child FGF2, which in turn was associated with decreases in child anxiety (path coefficient = −0.53, 95% CI −0.89, −0.18, p = .003) and depression (path coefficient = −0.38, 95% CI −0.72, −0.04, p = .03), controlling for maternal anxiety and depression. Pathways linking maternal symptoms (anxiety, depression) to child FGF2 and child symptoms were not significant. Findings indicate an indirect linkage between maternal FGF2 and child anxiety and depression via child FGF2 levels.

Fig. 1.

Fig. 1.

Open in a new tab

Note: SCARED: Screen for Child Anxiety Related Emotional Disorders; CDI: Child Depression Inventory; FGF2: Fibroblast Growth Factor 2; BAI: Beck Anxiety Inventory; BDI: Beck Depression Inventory.

Discussion

We found maternal and child FGF2 levels are significantly positively correlated and increased maternal FGF2 is predictive of increased child FGF2, which in turn predicts lower levels of both anxiety and depressive symptoms in the child. This is the first study, though preliminary, to test an intergenerational pathway linking FGF2 and psychiatric symptoms in mothers to anxiety and depression in youth, through youth FGF2 levels. Our findings provide important and novel insight into the cross-generational transmission of anxiety and depression and the possible role for FGF2 in this complex process.

These findings align with the growing body of research surrounding the role of FGFs in the risk for psychiatric problems. Together, they indicate that higher maternal FGF2 may serve as a protective factor, buffering children against such risk and increasing resilience in the face of other risk factors, such as maternal anxiety and/or depression.

One potentially fruitful direction for future research relates to the genetic heritability of variations in FGFs, in particular FGF2. Research has begun to identify specific genetic markers implicated in the risk for anxiety disorders (Olfson et al., 2022). The current study cannot say whether, and to what extent, the correlation between maternal and child FGF2 levels results from genetic heritability or other, non-genetic, factors.

The intergenerational pathway also has potential translational implications. For example, given findings from animal research showing that environmental factors can continue to impact FGF2 postnatally, early life interventions in humans that increase FGF2 may reduce risk of anxiety and depression, with long-term benefits even across the life span.

It is notable that no significant correlation emerged between maternal symptoms of anxiety and depression and maternal FGF2 levels. More research, including with larger samples, is required to determine whether such a correlation does exist or whether the association is specific to youth. If the association is specific to youth, it will be important to determine why this is, as well to clarify the age at which this link tapers off.

Limitations include reliance on child-reported symptoms of anxiety and depression. Multi-informant data and direct behavioral observations provide a more comprehensive picture of youth’s clinical presentation. FGF2 levels were assessed once, where repeated measurement may yield better indications of FGF2 functioning. This is mitigated by Lebowitz et al’s (2021) findings indicating strong significant correlations between FGF2 measurements taken six months apart. Also of note, FGF2 levels were collected after participants had spent some time in the lab, completing other study tasks such as responding to questionnaires. Given evidence suggesting that FGF2 levels may increase in response to stress (Bland et al., 2007; Bryant, Richardson, & Graham, 2022; Turner, Eren-Koçak, Inui, Watson, & Akil, 2016), it is possible that the association between mother and child FGF2 was increased due to the shared experience of completing the other tasks before collection. This could represent either a genetic or non-genetic pathway as the FGF2 response to stress may itself be genetically heritable. This study along with that of Lebowitz and colleagues’ (2021) underscore the need to conduct further research into FGF2’s role in the cross-generation transmission of risk for, and resilience to, anxiety and depression in youth.

Supplementary Material

1

Highlights:

  • Maternal FGF2 levels are positively correlated with youth FGF2 levels

  • Maternal FGF2 levels are associated with youth anxiety symptoms through youth FGF2 levels

  • Maternal FGF2 levels are associated with youth depressive symptoms through youth FGF2 levels

  • Results demonstrate a novel cross-generational pathway for the role of FGF2 in the risk for, and resilience to, youth anxiety and depression

Acknowledgement:

Funding:

NIMH Grant R21MH113946

NIMH Grant 1R61MH115113

Footnotes

Conflict of Interest:

The authors have no conflict of interest to disclose. Funding has been disclosed in the Author Statement.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Bland ST, Tamlyn JP, Barrientos RM, Greenwood BN, Watkins LR, Campeau S, . . . Maier SF (2007). Expression of fibroblast growth factor-2 and brain-derived neurotrophic factor mrna in the medial prefrontal cortex and hippocampus after uncontrollable or controllable stress. Neuroscience, 144(4), 1219–1228. doi: 10.1016/j.neuroscience.2006.11.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bryant EM, Richardson R, & Graham BM (2022). The association between salivary fgf2 and physiological and psychological components of the human stress response. Chronic Stress (Thousand Oaks), 6, 24705470221114787. doi: 10.1177/24705470221114787 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Deng Z, Deng S, Zhang MR, & Tang MM (2019). Fibroblast growth factors in depression. Front Pharmacol, 10, 60. doi: 10.3389/fphar.2019.00060 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Fumagalli F, Bedogni F, Slotkin TA, Racagni G, & Riva MA (2005). Prenatal stress elicits regionally selective changes in basal fgf-2 gene expression in adulthood and alters the adult response to acute or chronic stress. Neurobiol Dis, 20(3), 731–737. doi: 10.1016/j.nbd.2005.05.005 [DOI] [PubMed] [Google Scholar]
  5. Fydrich T, Dowdall D, & Chambless DL (1992). Reliability and validity of the beck anxiety inventory. Journal of anxiety disorders, 6(1), 55–61. doi: Doi 10.1016/0887-6185(92)90026-4 [DOI] [Google Scholar]
  6. Kiraly DD, Horn SR, Van Dam NT, Costi S, Schwartz J, Kim-Schulze S, . . . Murrough JW (2017). Altered peripheral immune profiles in treatment-resistant depression: Response to ketamine and prediction of treatment outcome. Transl Psychiatry, 7(3), e1065. doi: 10.1038/tp.2017.31 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Kolb B, & Gibb R. (2007). Brain plasticity and recovery from early cortical injury. Dev Psychobiol, 49(2), 107–118. doi: 10.1002/dev.20199 [DOI] [PubMed] [Google Scholar]
  8. Logan AC, & Goetsch VL (1993). Attention to external threat cues in anxiety states. Clinical Psychology Review, 13, 541–559. doi: 10.1016/0272-7358(93)90045-N [DOI] [Google Scholar]
  9. Molteni R, Fumagalli F, Magnaghi V, Roceri M, Gennarelli M, Racagni G, . . . Riva MA (2001). Modulation of fibroblast growth factor-2 by stress and corticosteroids: From developmental events to adult brain plasticity. Brain Res Brain Res Rev, 37(1–3), 249–258. doi: 10.1016/s0165-0173(01)00128-x [DOI] [PubMed] [Google Scholar]
  10. Monga S, Birmaher B, Chiappetta L, Brent D, Kaufman J, Bridge J, & Cully M. (2000). Screen for child anxiety-related emotional disorders (scared): Convergent and divergent validity. Depress Anxiety, 12(2), 85–91. doi: [DOI] [PubMed] [Google Scholar]
  11. Muthén LK, & Muthén B. (1998–2012). Mplus user’s guide. Seventh edition. Los Angeles, CA: Muthén & Muthén. [Google Scholar]
  12. O’Donovan MC, Norton N, Williams H, Peirce T, Moskvina V, Nikolov I, . . . Corvin A. P. O. h. o. o. (2009). Analysis of 10 independent samples provides evidence for association between schizophrenia and a snp flanking fibroblast growth factor receptor. Molecular psychiatry,. 14(1), pp. doi: 10.1038/mp.2008.10818813210 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Olfson E, Lebowitz ER, Hommel G, Pashankar N, Silverman WK, & Fernandez TV (2022). Whole-exome DNA sequencing in childhood anxiety disorders identifies rare de novo damaging coding variants. Depress Anxiety. doi: 10.1002/da.23251 [DOI] [PMC free article] [PubMed]
  14. Racine N, McArthur BA, Cooke JE, Eirich R, Zhu J, & Madigan S. (2021). Global prevalence of depressive and anxiety symptoms in children and adolescents during covid-19: A meta-analysis. JAMA Pediatr, 175(11), 1142–1150. doi: 10.1001/jamapediatrics.2021.2482 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Reuss B, & von Bohlen und Halbach O. (2003). Fibroblast growth factors and their receptors in the central nervous system. Cell & Tissue Research, 313(2), 139–157. [DOI] [PubMed] [Google Scholar]
  16. Salmaso N, Stevens HE, McNeill J, ElSayed M, Ren Q, Maragnoli ME, . . . Vaccarino FM (2016). Fibroblast growth factor 2 modulates hypothalamic pituitary axis activity and anxiety behavior through glucocorticoid receptors. Biol Psychiatry, 80(6), 479–489. doi: 10.1016/j.biopsych.2016.02.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Terwisscha van Scheltinga AF, Bakker SC, & Kahn RS (2010). Fibroblast growth factors in schizophrenia. Schizophr Bull, 36(6), 1157–1166. doi: 10.1093/schbul/sbp033 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Turner CA, Akil H, Watson SJ, & Evans SJ (2006). The fibroblast growth factor system and mood disorders. Biol Psychiatry, 59(12), 1128–1135. doi: 10.1016/j.biopsych.2006.02.026 [DOI] [PubMed] [Google Scholar]
  19. Turner CA, Eren-Koçak E, Inui EG, Watson SJ, & Akil H. (2016). Dysregulated fibroblast growth factor (fgf) signaling in neurological and psychiatric disorders. Semin Cell Dev Biol, 53, 136–143. doi: 10.1016/j.semcdb.2015.10.003 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1873101-supplement-1.docx (18.9KB, docx)

RESOURCES