Introduction
Social anxiety disorder (SAD) is a persistent and profoundly distressing mental condition marked by an intense fear or apprehension of being adversely evaluated in social contexts. 1 SAD is associated with substantial functional impairment and a reduced quality of life, typically appears early in life, during infancy or adolescence, and frequently has a permanent impact.Reference Kessler, Petukhova, Sampson, Zaslavsky and Wittchen 2 –Reference Patel, Knapp, Henderson and Baldwin 6 Fears associated with SAD typically pertain to a variety of situations, including talking to others or meeting new people, being observed while eating or drinking, and performing in front of an audience. 1Numerous aspects of a person’s life, such as relationships, social development, academic achievement, and career success, may be adversely affected by the disorder, which could ultimately result in serious functional deficiencies and a markedly reduced quality of life.Reference Stein and Kean 4 Although research into the identification, assessment, and treatment of SAD has expanded, the condition continues to be overlooked in various settings. Prevalence rates of SAD can vary significantly, ranging from 0.8% to 36%,Reference Ghanizadeh, Mohammadi and Yazdanshenas 7 , Reference Mattick and Clarke 8 with the 12-month prevalence in the United States being around 7%, while in Europe it is approximately 2.3%. 1 Despite its widespread occurrence, SAD remains one of the most under-treated mental health conditions. Several factors may contribute to this issue. Social anxiety is often not recognized, particularly in adolescents who may not seek help due to a lack of understanding or awareness of the disorder.Reference Coles, Ravid, Gibb, George-Denn, Bronstein and McLeod 9 Moreover, the stigma surrounding social anxiety may cause individuals to avoid revealing their condition, as it is often seen as embarrassing.Reference Anderson, Jeon, Blenner, Wiener and Hope 10 , Reference McDonagh, Lynch and Hennessy 11 Consequently, the full impact of SAD, especially among younger individuals, remains largely unrecognized and inadequately addressed in clinical practice and research.
Studies on the origins of SAD continue, and there has been significant progress in understanding its development. Evidence points to a strong familial link, with children of parents who have SAD being at a higher risk of developing the disorder themselves.Reference Joan, King and Ollendick 12 This familial trend may be influenced by both genetic factors and shared environmental influences. Although genetic factors contribute to the onset of SAD, researchers are still working to pinpoint the exact genes and biological mechanisms involved. Many studies have explored genes associated with serotonergic (5-HT) and dopaminergic pathways, but definitive genetic markers for SAD remain unidentified.Reference Knappe, Sasagawa, Creswell, Ranta, La Greca, Garcia-Lopez and Marttunen 13 While ample evidence highlights the importance of genetic factors, brain structures, and temperament in SAD development, further research is necessary to better understand how these elements interact.Reference Baba, Kloiber and Zai 14 , Reference Mizzi, Pedersen, Lorenzetti, Heinrichs and Labuschagne 15
Research exploring the link between neurobiological processes and anxiety disorders has indicated that various neuropeptides, neurohormones, and immune factors may play a role in the development of these conditions.Reference Stein, Westenberg and Liebowitz 16 , Reference Rauch, Shin and Wright 17 In addition to the commonly studied neurotransmitters, an increasing number of studies are investigating the potential involvement of the immune system and inflammatory molecules, yielding promising results. Contextually, several studies have highlighted the role of inflammation in neurotransmitter metabolism and neuroplasticity, suggesting that inflammation may not merely be a consequence but a primary factor contributing to the observed clinical severity of anxiety disorders.Reference Amitai, Kaffman and Kroizer 18 , Reference Ekinci and Ekinci 19 These findings have provided a new framework for understanding the molecular basis of psychiatric symptoms and introduced the possibility of a link between immune system alterations and the onset of psychiatric disorders.Reference Sperner-Unterweger 20 At the same time, emerging research has revealed that oxytocin (OXT) significantly affects mood, social cognition, stress reduction, and social interactionsReference Guyton, Hall, Guyton and Hall 21—such as gossiping behaviorReference Brondino, Fusar-Poli and Politi 22—with some suggesting that OXT may also play a role in the development of SAD in humans.Reference Oh, Kim and Ha 23 Despite the expanding body of research into the biological foundations of anxiety disorders, particularly SAD, the current findings remain inconclusive in forming a comprehensive biological theory.
One of the major clinical implications in the management of SAD is its comorbidity with numerous psychiatric disorders. Indeed, up to 69% of people with SAD will at some point in their lives have another significant comorbid disease.Reference Schneier, Johnson, Hornig, Liebowitz and Weissman 24In particular, social anxiety is frequently observed in individuals with Autism Spectrum Disorder (ASD), a neurodevelopmental condition characterized by deficits in social communication, presence of restricted and repetitive behaviors, and heightened or diminished sensory sensitivities. 1 Indeed, according to numerous studies, when assessed by medical professionals, 13–40% of individuals with autism fit the criteria for SAD.Reference Hofvander, Delorme and Chaste 25 –Reference Russell, Murphy and Wilson 28Furthermore, the scientific literature has suggested the presence of a genetic overlap between ASD and social anxiety, as evidenced by the higher prevalence of social anxiety among biological relatives of people with ASD.Reference Adams, Young, Simpson and Keen 29 , Reference Piven and Palmer 30This genetic link may indicate that the underlying mechanisms of ASD and SAD are similar, making it more difficult to distinguish between the two. Moreover, ASD and SAD exhibit remarkably similar traits. In particular, the convergence between the two disorders is especially noticeable in domains pertaining to social interaction and social skills.Reference Simonoff, Pickles, Charman, Chandler, Loucas and Baird 31 –Reference White, Ollendick and Bray 33 Indeed, while individuals with ASD and SAD may display overlapping social behaviors, these can arise from distinct underlying processes; for example, apparent social withdrawal in ASD may reflect atypical or unsuccessful attempts at interaction rather than a lack of social motivation, complicating differential diagnosis. 1Similarly, the ongoing challenges in interacting with others experienced by people with ASD may lead to the development of symptoms of social anxiety, especially in individuals who have greater awareness of their communication deficits.Reference Bejerot and Mörtberg 34 , Reference Salazar, Baird and Chandler 35
In this context, considering the lack of research on the biological mechanism underlying social anxiety and the high prevalence of SAD in subjects of the autism spectrum, the aim of our work was to investigate possible biochemical correlates of SAD in patients with ASD, their first-degree relatives, and healthy controls (HCs), mainly focusing on inflammatory markers, neuropeptides, and growth factors.
Materials and methods
Sample recruitment procedures
For this study, a total of 72 participants were recruited, including 24 adults with ASD, their first-degree relatives, named the BAP group, and 24 HCs. Blood samples were collected from all participants, and biochemical parameters were assessed using enzyme-linked immunosorbent assays (ELISA).
Participants in the ASD group were recruited from both inpatients and outpatients at the Psychiatric Department of the Azienda Ospedaliera Universitaria Pisana (AOUP) at the University of Pisa. In order to participate in the study, patients had to have a clinical diagnosis of ASD, be at least 18 years old but not older than 65, and have no language developmental alterations and no or only mild intellectual impairment. In particular, individuals with evident intellectual disability interfering with the ability to complete assessments were excluded based on clinical judgment; however, formal IQ testing was not systematically performed. In order to participate in the study, subjects had to have received an ASD diagnosis, which was then confirmed according to DSM-5 criteria through comprehensive clinical evaluation conducted by expert psychiatrists with specific experience in autism spectrum conditions. A parent or sibling was asked to participate in the recruitment process for each patient, forming the BAP group. Subjects belonging to the BAP group showed intermediate autistic features compared to the ASD and HC groups, as reported in previous studies.Reference Carpita, Nardi and Palego 36 The primary exclusion criteria for both groups were having a diagnosis of schizophrenia, neurodegenerative disease, or any other pertinent medical or neurological problem, as well as being unable to complete the examinations because of a language or intellectual difficulty. Substance use disorder, defined according to DSM-5 criteria and assessed through clinical evaluation, was an exclusion criterion. However, substance use behaviors reported in the SHY-SR questionnaire, referring to subclinical or occasional use and not necessarily indicating the presence of a diagnosable substance use disorder, were allowed. Furthermore, BAP participants who were less than 18 years old or who had received a DSM-5 diagnosis of ASD or another neurodevelopmental disorder were not included in the study. The recruitment of the HC group was carried out on a voluntary basis. Exclusion criteria for the HC group were the same as those for the ASD group; in addition, from the HC group were excluded individuals diagnosed with any mental disease based on the DSM-5 criteria. Data about differences in socio-demographic variables among groups and assessment of other diagnoses with the Structured Clinical Interview for DSM-5 (SCID-5) have been reported and discussed elsewhere.Reference Carpita, Nardi and Palego 36
All participants received comprehensive information about the study and had the opportunity to ask questions prior to providing their signed informed consent. The current study was carried out in accordance with the Helsinki Declaration, and all procedures were approved by the local ethics committee.
Each participant was asked to complete a questionnaire for the self-assessment report of social phobic symptoms, the Social-Phobia Spectrum—Self Report (SHY-SR), and had a blood sample drawn in order to perform the biochemical evaluation.
Biochemical procedures
For each participant, on the day of the psychometric evaluation, 20 ml of peripheral venous blood were collected after at least 12 hours of fasting. The samples were stored in three types of tubes:
-
– one containing tripotassium ethylenediaminetetraacetic acid (K3EDTA) in order to allow the separation of platelet-poor plasma (PPP) from platelets
-
– one containing lithium-heparin for the separation of plasma
-
– one without anticoagulant and with a coagulation activator for the separation of serum.
The peripheral blood was centrifuged at low speed (150× g) for 15 minutes at room temperature within 30 minutes of collection, in order to determine the precipitation of the corpuscular part (erythrocytes and leukocytes) according to the density gradient. The remaining supernatants, platelet-rich plasma (PRP), were divided into two 15 ml Falcon tubes and centrifuged at 1500× g for 15 minutes. Aliquots of PPP and platelet pellet were then obtained. All aliquots thus obtained, PRP, PPP, and serum, were then stored in high-quality, low protein binding Eppendorf Safe-Lock tubes, in a refrigerator at −80 °C, until the day of testing.
All biochemical parameters were measured with ELISA kit.
IL-6 and BDNF levels (plasma and intraplatelet) were measured in PPP with sandwich ELISA kits (Picokine IL-6 assay, Boster Biological Technology, Pleasanton, CA, USA and Biosensis, mature BDNF RapidTM, Thebarton, Australia) equipped with a 96-well microplate. These kits are equipped with a primary monoclonal antibody anti-IL-6 or anti-BDNF, a second biotinylated antibody and a streptavidin-biotin complex placed at the bottom of each well.
Levels of homocysteine (HCY), 5-HT, TRP and metabolites of the KYN pathway, such as quinolinic acid (QUIN) and Kynurenic acid (KYNA), were measured by indirect competitive ELISA kits (ImmuSmol, Bordeaux, France), in PPP, with the sole exception of KYNA, which was measured in serum, according to the indications provided by the assay.
The intraplatelet content used for the measurement of intraplatelet 5-HT and BDNF was obtained by homogenization and fractionation techniques.
The ELISA kits used were equipped with microtinylated plates pre-coated with the same analyte under investigation. To perform the competitive assay, the analyte and a specific antibody were added to each well. Subsequently, for the detection reaction, a second biotinylated antibody linked to horseradish peroxidase (HRP) was added, followed by the HRP substrate, 3,3′,5,5′-tetramethylbenzidine (TMB).
For HCY, in order to reduce the risk of matrix effects obtained by direct measurement, aliquots of PPP were incubated with an enzymatic reaction mixture containing the enzyme S-adenosyl-l-homocysteine hydrolase and its substrate adenosine/dithiothreitol (DTT). This allowed the transformation of HCY into S-adenosyl-homocysteine.
Regarding the measurement of intraplatelet levels of 5-HT and BDNF (ng/mL), in order to avoid any bias associated with individual differences in platelet count, we proceeded according to the Bradford method.Reference Bradford 37
The final step of the test was performed using the multi-scan spectrophotometer to read the absorbance of the samples at 450 nm, as indicated in the kit instructions. The calibration curve was created using a 4-parameter nonlinear regression equation using Graph-Pad Prism Software (version 8.0, San Diego, CA, USA). The analyte values obtained were interpolated in logarithmic form and subsequently transformed into exponential form and multiplied by the dilution factors to obtain the final concentration.
Psychometric instruments
SHY-SR
The SHY-SV is a self-report questionnaire designed to evaluate a wide variety of social phobia characteristics, including personality traits, from the most prevalent to the unusual and subtle ones. It is made of 168 dichotomous items organized in five domains that capture different manifestations of the condition and various scenarios in which symptoms may exacerbate such as Childhood and adolescence, Interpersonal sensitivity, Behavioral inhibition, Specific anxieties and phobic features and Substance Abuse. The questionnaire showed considerable internal consistency, convergent validity with other dimensional measures of social anxiety, and strong test–retest reliability.Reference Dell’Osso, Rucci and Cassano 38
Statistical analysis
We used non-parametric methods to elaborate our data because our sample did not pass the variance homoscedasticity or normality tests.
Firstly, aiming to assess significant differences in social anxiety symptoms among ASD, BAP, and HCs individuals, we carried out a Kruskal–Wallis test including all SHY-SR domains as well as its total. Additionally, we then performed a Chi-square analysis to compare the use of different substances of abuse in order to improve social interaction skills, assessed with items 163–168 of the SHY-SR, in all three groups. We then used Spearman’s correlation analyses in order to investigate the pattern of correlations between the scores reported on the SHY-SR domains and the biochemical variables.
Lastly, we performed a set of linear regression analyses. The first regression was performed using the SHY-SR total score as the dependent variable and platelet BDNF, KYNA, HCY, BDNF, IL-6, 5-HT, TRP, KYN, platelet 5-HT, and QUIN as independent variables, in order to investigate the presence of significant positive or negative association with higher SHY-SR score. Further regression analyses were performed with the same independent variables and the scores reported at each SHY-SR domain as dependent variables, in order to investigate the presence of significant positive or negative association with specific social phobic symptomatology.
Every statistical evaluation was performed with SPSS version 26.0.
Results
Results from the Kruskal-Wallis test, used to compare the scores obtained in all SHY-SR domains and total by the three groups are reported in Table 1. In particular, ASD subjects appeared to score significantly higher in all SHY-SR domains -with the only exception for Substance Abuse- as well as in its total score, compared to BAP and HCs, which in turn did not significantly differ from one another. ASD individuals also scored significantly higher than HCs in the Substance Abuse domain however, no significant differences were highlighted between BAP and ASD and BAP and HCs.
Comparison of the Scores Obtained in All SHY-SR Domains Between Groups
Table 1. Long description
The table consists of six columns: Domain, A S D (Mean plus or minus S D), B A P (Mean plus or minus S D), H Cs (Mean plus or minus S D), H, and p.
* Childhood/adolescence: A S D 6.32 plus or minus 3.34; B A P 2.82 plus or minus 2.56; H Cs 2.75 plus or minus 2.03; H 16.959; p less than .001.
* Interpersonal sensitivity: A S D 17.48 plus or minus 5.93; B A P 7.14 plus or minus 5.69; H Cs 4.42 plus or minus 4.60; H 32.288; p less than .001.
* Behavioral inhibition: A S D 10.36 plus or minus 4.63; B A P 2.27 plus or minus 2.16; H Cs 1.46 plus or minus 1.72; H 37.878; p less than .001.
* Specific anxiety/phobic features: A S D 51.50 plus or minus 20.28; B A P 16.91 plus or minus 14.92; H Cs 10.08 plus or minus 10.95; H 34.832; p less than .001.
* Substance Abuse: A S D 1.36 plus or minus 1.40; B A P 0.82 plus or minus 1.05; H Cs 0.46 plus or minus 0.83; H 7.179; p equals .028.
* S H Y dash S R total score: A S D 85.24 plus or minus 29.93; B A P 20.95 plus or minus 23.78; H Cs 19.17 plus or minus 17.30; H 34.386; p less than .001.
Footnotes indicate that A S D scores are significantly higher than B A P and H Cs for most domains, and significantly higher than H Cs for Substance Abuse.
* ASD > BAP, HCs; °ASD > HCs. Significant for p < .001.
Abbreviation: ASD: autism spectrum disorder; BAP: broad autism phenotype; HC: healthy controls.
Moreover, results from the Chi-Square analysis showed that the subjects belonging to the ASD group tended to make greater use of anxiolytics in order to improve their social interaction skills (see Table 2).
Comparison of the Use of Substance of Abuse to Improve Social Interaction Skills Between Diagnostic Groups
Table 2. Long description
The table contains seven columns: Substance, Response (No or Yes), A S D n (%), B A P n (%), H Cs n (%), Chi-square, and p-value.
* Tobacco: A S D Yes 9 (40.9%), B A P Yes 7 (31.8%), H Cs Yes 4 (16.7%). Chi-square 3.340, p .118.
* Alcohol: A S D Yes 5 (22.7%), B A P Yes 4 (18.2%), H Cs Yes 6 (25.0%). Chi-square .319, p .853.
* Anxiolytics: A S D Yes 8 (36.4%), B A P Yes 4 (18.2%), H Cs Yes 0 (0%). Chi-square 10.450, p .005. This result is statistically significant.
* Marijuana: A S D Yes 4 (18.2%), B A P Yes 1 (4.5%), H Cs Yes 1 (4.2%). Chi-square 3.542, p .170.
* Cocaine or amphetamines: A S D Yes 1 (4.5%), B A P Yes 0 (0%), H Cs Yes 0 (0%). Chi-square 2.122, p .346.
* Other drugs and medicines: A S D Yes 3 (13.6%), B A P Yes 2 (9.1%), H Cs Yes 0 (0%). Chi-square 3.277, p .194.
Abbreviations: A S D stands for autism spectrum disorder, B A P stands for broad autism phenotype, and H C stands for healthy controls.
* ASD, BAP > HCs; significant for p < .05.
Abbreviation: ASD: autism spectrum disorder; BAP: broad autism phenotype; HC: healthy controls.
Results from Spearman’s correlation highlighted a specific pattern of correlation between the biochemical variables and the SHY-SR domains (see Table 3). In particular, the Interpersonal sensitivity domain appeared to be negatively correlated with platelet BDNF, KYNA, 5-HT, and TRP and positively correlated with HCY and IL-6. Similarly, the SHY-SR Behavioral inhibition domain negatively correlated with platelet BDNF, KYNA, 5-HT, and TRP and positively correlated with HCY. The domain of Specific anxiety and phobic features negatively correlated with KYNA, TRP, and 5-HT and positively correlated with HCY and IL-6; while Substance abuse correlated positively with IL-6 and negatively with QUIN. Interestingly, the Childhood and Adolescence domain positively correlated only with HCY.
Spearman Correlation Coefficients Between SHY–SV Domain Scores and Biochemical Variables in the Overall Sample
Table 3. Long description
The table consists of 6 columns representing S H Y dash S R domains: Childhood/adolescence, Interpersonal sensitivity, Behavioral inhibition, Specific anxi./phobic feat., Substance abuse, and S H Y dash S R total score.
Biochemical variables (rows) and their correlations include:
* Platelet B D N F: Correlations range from minus .376 (Behavioral inhibition, p < .001) to .070 (Substance abuse).
* K Y N A: Significant negative correlations with Interpersonal sensitivity (minus .261), Behavioral inhibition (minus .303), Specific anxi./phobic feat. (minus .290), and Total score (minus .272).
* H C Y: Significant positive correlations across most domains, notably Behavioral inhibition (.492, p < .001) and Total score (.402, p < .001).
* B D N F: No significant correlations, values range from .034 to .151.
* I L dash 6: Significant positive correlations with Interpersonal sensitivity (.454, p < .001), Specific anxi./phobic feat. (.336), Substance abuse (.404), and Total score (.395).
* 5 dash H T: Significant negative correlations with Behavioral inhibition (minus .481, p < .001), Specific anxi./phobic feat. (minus .364, p < .001), and Total score (minus .339, p < .001).
* T R P: Significant negative correlations with Interpersonal sensitivity (minus .266), Behavioral inhibition (minus .258), and Specific anxi./phobic feat. (minus .246).
* K Y N: No significant correlations, values range from minus .202 to minus .001.
* Platelet 5 dash H T: No significant correlations, values range from minus .211 to minus .024.
* Q U I N: Only significant correlation is with Substance abuse (minus .256, p < .05).
*significant for p < .05; ** significant for p < .001.
Abbreviations: 5-HT: serotonin; BDNF: brain derived neurotrophic factor; HCY: homocysteine; IL-6; interleukin 6; KYN: kynurenine;KYNA: Kynurenic acid; QUIN: quinolinic acid; TRP: tryptophan.
Lastly, SHY-SR total score emerged as negatively correlated with platelet BDNF and 5-HT and positively with KYNA, HCY, and IL-6.
Results from the first linear regression analysis identified a significant positive association between IL-6 and higher SHY-SR scores, while 5-HT was found to be significantly negatively associated with these scores (see Table 4).
Linear Regression Analysis with SHY-SR Total Score as a Dependent Variable and Platelet BDNF, KYNA, HCY, BDNF, IL-6, 5-HT, TRP, KYN, Platelet 5-HT, and QUIN as Independent Variables in the Overall Sample
Table 4. Long description
A linear regression analysis table with S H Y-S R total score as the dependent variable. The model has an R square of 0.325 and an Adjusted R square of 0.214.
Columns from left to right: Variable, b (S E), B E T A, t, p, and C.I. 95% (Lower bound, Upper bound).
* constant: 28.568 (21.515), t = 1.328, p = .189, C.I. [-14.453, 71.590].
* Platelet B D N F: -0.728 (1.489), B E T A = -0.057, t = -0.489, p = .627, C.I. [-3.706, 2.250].
* K Y N A: -0.950 (0.591), B E T A = -0.202, t = -1.608, p = .113, C.I. [-2.132, 0.231].
* H C Y: 1.415 (0.743), B E T A = 0.220, t = 1.903, p = .062, C.I. [-0.072, 2.901].
* B D N F: 0.869 (0.757), B E T A = 0.137, t = 1.148, p = .256, C.I. [-0.645, 2.382].
* I L-6: 1.859 (0.754), B E T A = 0.296, t = 2.465, p = .017 (significant), C.I. [0.351, 3.368].
* 5-H T: -0.262 (0.125), B E T A = -0.242, t = -2.092, p = .041 (significant), C.I. [-0.513, -0.012].
* T R P: -0.187 (0.170), B E T A = -0.117, t = -1.103, p = .274, C.I. [-0.527, 0.152].
* K Y N: -0.009 (0.016), B E T A = -0.078, t = -0.588, p = .559, C.I. [-0.041, 0.022].
* Platelet 5-H T: -0.064 (0.308), B E T A = -0.024, t = -0.207, p = .837, C.I. [-0.680, 0.552].
* Q U I N: 0.069 (0.138), B E T A = 0.061, t = 0.501, p = .618, C.I. [-0.206, 0.344].
R square = 0.325; Adjusted R square = 0.214.
Abbreviations: 5-HT: serotonin; BDNF: brain derived neurotrophic factor; HCY: homocysteine; IL-6; interleukin 6; KYN: kynurenine; KYNA: Kynurenic acid; QUIN: quinolinic acid; TRP: tryptophan.
* significant for p < .05.
Lastly, results from the set of regression analyses performed with SHY-SR domain scores as dependent variables highlighted a significant positive association between IL-6 and SHY-SR Interpersonal sensitivity and Specific anxieties and phobic features domain scores. Similarly, KYNA was significantly positively associated with the Behavioral inhibition domain score, while HCY appeared to be significantly positively associated with the Childhood and adolescence and Specific anxieties and phobic features domain scores. Conversely, 5-HT emerged as negatively associated with both SHY-SR Behavioral inhibition and Specific anxieties and phobic features domain scores (see Table 5).
Linear Regression Analyses with SHY-SR Domains Score as a Dependent Variable and Biochemical Correlates as Independent Variables in the Overall Sample
Table 5. Long description
The table presents four linear regression models. The columns are: Variable, b (S E), BETA, t, p, and 95% C.I. (Lower bound, Upper bound).
1. Childhood/Adolescence Model:
- Constant: b (S E) 1.971 (1.844), t 1.069, p .289.
- H C Y: b (S E) .213 (.064), BETA .392, t 3.346, p .001 (significant).
- Model stats: R square 0.305, Adjusted R square 0.191.
2. Interpersonal sensitivity Model:
- Constant: b (S E) 10.040 (4.689), t 2.141, p .036 (significant).
- I L-6: b (S E) .369 (.164), BETA .283, t 2.243, p .029 (significant).
- Model stats: R square 0.257, Adjusted R square 0.135.
3. Behavioral inhibition Model:
- Constant: b (S E) 5.150 (2.914), t 1.768, p .082.
- K Y N A: b (S E) minus .166 (.080), BETA minus .259, t minus 2.080, p .042 (significant).
- 5-H T: b (S E) minus .043 (.017), BETA minus .293, t minus 2.552, p .013 (significant).
- Model stats: R square 0.335, Adjusted R square 0.226.
4. Specific anxieties and phobic features Model:
- Constant: b (S E) 20.531 (13.552), t 1.515, p .135.
- H C Y: b (S E) 1.104 (.468), BETA .265, t 2.358, p .022 (significant).
- I L-6: b (S E) 1.097 (.475), BETA .270, t 2.309, p .024 (significant).
- 5-H T: b (S E) minus .165 (.079), BETA minus .236, t minus 2.093, p .040 (significant).
- Model stats: R square 0.359, Adjusted R square 0.254.
Abbreviations: 5-H T is serotonin; H C Y is homocysteine; I L-6 is interleukin 6; K Y N A is Kynurenic acid.
* Significant for p < .05.
Abbreviations: 5-HT: serotonin; HCY: homocysteine; IL-6; interleukin 6; KYNA: Kynurenic acid.
Discussion
As expected, our results highlighted significantly higher scores of ASD subjects in the measures for SAD compared to the other two diagnostic groups. This finding is in line with the available literature that has long reported similarities and correlations between SAD and ASD.Reference Bellini 39 –Reference Spain, Happé and Johnston 44 Numerous studies have highlighted the high prevalence of social anxiety symptoms in individuals with ASD. In fact, estimates suggest that the prevalence of social anxiety in individuals on the spectrum can reach up to 50%, significantly higher than the 7–13% prevalence observed in the general population.Reference Capriola, Maddox and White 45 –Reference Spain, Sin, Linder, McMahon and Happé 48 This evidence can be attributed to the core challenges in acquiring and applying social skills, along with the generally weak social competence seen in ASD, which increase the likelihood of developing SAD.Reference Capriola, Maddox and White 45 A growing body of research offers several plausible explanations for why children with ASD or neurotypical children displaying elevated autism traits are at a greater risk of developing social phobic symptomatology. For example, children and adolescents with ASD often face bullying or mistreatment due to their social difficulties, leading to numerous unsuccessful social interactions with peers.Reference Sreckovic, Brunsting and Able 49 , Reference van Roekel, Scholte and Didden 50 Additionally, individuals with ASD, because of their social communication challenges, are more likely to experience persistent social rejection and failure in interactions, which can contribute to negative self-perceptions—especially among those more cognitively able and with greater self-awarenessReference Kerns, Kendall and Berry 51 –Reference White, Bray and Ollendick 53—and may increase the tendency to interpret social cues or situations as threatening or harmful.Reference BlöteAW, Van den Bos and Westenberg 54 , Reference Miers, Blöte, Bögels and Westenberg 55 Ultimately, these negative social experiences and stressful encounters often act as early triggers or direct contributors to the development of SAD.Reference Bellini 39 , Reference Rapee and Spence 56 Additionally, high levels of impairment in social communication and social motivation are identified as risk factors for anxiety, as noted by Briot et al.Reference Briot, Jean and Jouni 40 These factors, much like the ones previously mentioned, are associated with challenges in developing essential social skills,Reference Kim, Ahn, Lee, Ha and Cheon 43 which in turn contribute to higher levels of social anxiety in individuals with ASD. Interestingly, it has also been observed that anxiety symptoms can worsen social deficits in ASDReference Rapee and Spence 56 , Reference Factor, Ryan, Farley, Ollendick and Scarpa 57, leading to outcomes such as social withdrawal, impaired social communication, and heightened sensitivity to social situations. This evidence suggests that the relationship between social deficits and anxiety in ASD may be bidirectional: with social difficulties that can lead to anxiety, and anxiety symptoms that may further aggravate social impairments.
Interestingly, when addressing the misuse of substances in order to improve social interaction skills, the ASD group emerged to make significantly more use of benzodiazepines. This data is not only in line with the research that reports a doubled risk of substance misuse in the ASD community when compared to non-clinical populations,Reference Carpita, Cremone and Nardi 41 , Reference Butwicka, Långström and Larsson 58—although some studies reported instead that ASD subjects are less likely to indulge in substancesReference Weir, Allison and Baron-Cohen 59—but, given the higher prevalence of SAD in the ASD group, it is also coherent with the substantial research on the use of drugs to reduce social anxiety in clinical samples of SAD.Reference Buckner, Heimberg, Ecker and Vinci 60 , Reference Buckner, Morris, Abarno, Glover and Lewis 61 Indeed, it is commonly hypothesized that, in order to reduce stress, ease anxiety, improve social empathy, lessen sensory stimulation, and promote social connection, people with ASD may turn to sedatives.
Interestingly, our results highlighted a positive correlation between IL-6 levels and SHY-SR scores, and also how it emerged as positively associated with greater social phobic symptomatology. Although evidence on biochemical correlates is scant and sometimes conflicting, these data can be explained in light of the roles of these molecules in the inflammatory cascade and of the recent data on the correlation between inflammation and various psychiatric disorders, including anxiety and depression.Reference Harsanyi, Kupcova, Danisovic and Klein 62 , Reference Vogelzangs, Beekman, de Jonge and Penninx 63
In particular, IL-6 is a pro-inflammatory cytokine that plays a key role in the body’s immune and inflammatory responses. When the body experiences stress, including psychological stress, an inflammatory cascade is triggered, leading to an increase in cytokine levels, like IL-6. Moreover, chronic or severe stress, often seen in anxiety disorders, disrupts the regulation of the HPA axis, resulting in prolonged cortisol release and systemic inflammation. Elevated levels of IL-6 are commonly observed in individuals with anxiety and depression, and these cytokines are believed to impact neural circuits involved in mood regulation and fear processing, thereby contributing to anxiety symptoms. Specifically, heightened IL-6 has been associated with increased activity in stress-sensitive brain regions, such as the amygdala, which is implicated in fear and social anxiety.Reference Maydych 64 –Reference Roohi, Jaafari and Hashemian 66 Therefore, elevated IL-6 levels in individuals with social anxiety may reflect the complex interplay between chronic stress, immune activation, and neurobiological changes.
Furthermore, recent studies have highlighted a potential link between inflammation—mediated by cytokines such as IL-6—and alterations in 5-HT levels, particularly in conditions like depression and other neuropsychiatric disorders, which have been associated with increased inflammatory activity.Reference Harsanyi, Kupcova, Danisovic and Klein 62 , Reference Vogelzangs, Beekman, de Jonge and Penninx 63 In this context, IL-6 may influence 5-HT levels indirectly. Indeed, both acute and chronic inflammation activate the enzyme (IDO), which converts TRP into KYN. This shift leads to elevated KYN levels, thereby reducing the amount of TRP available for 5-HT synthesis. The increased KYN is subsequently broken down into QUIN, an N-methyl-d-aspartate (NMDA) receptor agonist that can contribute to glutamate excitotoxicity and oxidative stress.Reference Haroon, Raison and Miller 67 These cytokine-induced changes in the IDO/KYN pathway can lower 5-HT and dopamine levels, while simultaneously raising glutamate levels—imbalances that are linked to stress-related symptoms such as depressed mood and anhedonia.Reference Ravi, Miller and Michopoulos 65 Ultimately, changes in neurotransmitter synthesis, release, and reuptake may mediate the effects of central cytokines on neurotransmitter systems, eventually leading to decreased monoamine availability and elevated extrasynaptic glutamate, which may cause excitotoxic damage.Reference Ravi, Miller and Michopoulos 65
Furthermore, IL-6 was positively correlated with SHY-SR domains Interpersonal sensitivity, Specific Anxieties and Phobic Features, and Substance abuse, although the regression analysis confirmed its association only with the first two domains. These results align with the consistent literature regarding the link between real or perceived social stress and increased inflammatory responses,Reference Ehrlich, Miller, Rohleder and Adam 68 –Reference Robles 71 although to date the evidence remains controversial.Reference Chiang, Eisenberger, Seeman and Taylor 72 –Reference Moieni and Eisenberger 74 In particular, Feldman et al.Reference Feldman, Jolink and Alvarez 75 highlighted that, after an induced inflammatory response achieved after influenza vaccine administration, subjects with higher IL-6 levels misjudged the trustworthiness of neutral faces, leading the authors to suggest that inflammation may play a role in altering the prominence of emotional signals during social decision-making.Reference Feldman, Jolink and Alvarez 75 On the other hand, a recent review addressed the link between social interaction, psychotic disorders, and inflammation and reported how social experiences may influence inflammatory activity.Reference Faustmann, Kamp, Räuber, Dukart, Melzer and Schilbach 76 For example, loneliness was reported to be linked with higher systemic inflammation and impaired immune responses after a vaccine or an infection, while negative interpersonal interactions were found to be connected with higher levels of IL-6 and soluble tumor necrosis factor-receptor 2 (sTNF-R2) in oral mucosal fluids of humans. Conversely, according to data obtained from animal investigations, social support seems to reduce plasmatic levels of proinflammatory cytokines and to prevent their increase after an ischemic insult.Reference Faustmann, Kamp, Räuber, Dukart, Melzer and Schilbach 76 On the other hand, increased pro-inflammatory cytokines following an infection or other inflammatory stimuli (that is, a vaccine) could lead to specific affective, social, and behavioral changes, globally named “Sickness behavior.”Reference Dantzer and Kelley 77 , Reference Devlin, Smith and Bilbo 78 This set of symptoms, including increased body temperature, loss of appetite, fatigue, anhedonia, disturbed sleep, and social withdrawal, is considered a survival response to inflammatory challenges, but can become pathological when prolonged.Reference Dantzer and Kelley 77 , Reference Devlin, Smith and Bilbo 78 Although the underlying mechanisms are not completely clarified, IL-1𝛽 and TNF-𝛼 seem to contribute to sickness behavior through specific neuronal pathway regulation, while IL-6 and IFN were proposed to potentiate these behavioral effects.Reference McCusker and Kelley 79 In conclusion, it is conceivable that there is a bi-directional connection between social stress and inflammation, as well as between inflammation and sensitivity to social interactions, with IL-6 playing a key role in this context.Reference Faustmann, Kamp, Räuber, Dukart, Melzer and Schilbach 76
Regarding the relationship between IL-6 and Specific Anxieties and Phobic Features, while research on this specific topic is still limited and primarily based on animal models, our results are consistent with existing studies. Several key observations support the idea that peripheral IL-6 plays a distinct role in the biological response to fear. Recent studies have shown that peripheral IL-6 levels rise shortly after the retrieval of a phobic memory. Moreover, the extinction of fear responses following a phobic trigger is hindered by direct administration of certain cytokines, such as IL-6, to the amygdala prior to extinction training.Reference Bi, Shi, Yang, Wang and Qin 80 –Reference Young, Howell and Hopkins 82 Additionally, blocking IL-6 through pharmacological or genetic interventions during extinction training has been found to enhance the long-term effectiveness of extinction learning.Reference Young, Howell and Hopkins 82 Further evidence suggests that plasma IL-6 increases rapidly following re-exposure to a non-physiological aversive conditioned stimulus,Reference Zhou, Kusnecov, Shurin, DePaoli and Rabin 83 and similar quick increases in IL-6 levels have been observed in humans exposed to the threat of an impending electric shockReference Breznitz, Ben-Zur and Berzon 84 suggesting that the anticipation of harm triggers the IL-6 response. Supporting this notion, the extinction of the fear response has been shown to significantly reduce the IL-6 response to the conditioned stimulus.Reference Young, Howell and Hopkins 82 Moreover, changes in plasma IL-6 levels in response to the conditioned stimulus have been found to correlate with the intensity of fear-related behaviors elicited upon re-exposure to the stimulus.Reference Young, Howell and Hopkins 82
Furthermore, several studies have explored the relationship between inflammatory markers, such as cytokines, and the use of various psychotropic substances, with results that vary across different substances. Specifically, the findings regarding cannabis use remain inconsistent. For instance, research on tetrahydrocannabinol (THC) and other cannabinoids has highlighted their anti-inflammatory effects, which are mediated through multiple immune-related mechanisms. These include the suppression of inflammatory cytokines such as TNF-α, IL-6, IFN-γ, and IL-12.Reference Klein 85 –Reference Verhoeckx, Korthout and van Meeteren-Kreikamp 87 In addition, macrophages from marijuana smokers have shown reduced production of TNF-α, IL-6, and granulocyte-macrophage colony-stimulating factor when stimulated with lipopolysaccharide.Reference Baldwin, Tashkin, Buckley, Park, Dubinett and Roth 88 More broadly, cannabis use has been linked to lower serum levels of cytokines like IL-6,Reference Keen, Pereira and Latimer 89 –Reference Ribeiro, Castro and Dorneles 91 TNF-α,Reference Keen and Turner 92 , Reference Massi, Sacerdote and Ponti 93 and IL-2Reference Pacifici, Zuccaro and Farré 94 in various user populations. However, some studies have reported no significant changes or even slight increases in IL-6 and TNF-α levels.Reference Cloak, Alicata, Ernst and Chang 95 Interestingly, other research has observed higher levels of oxidative stress, total oxidant status, and proinflammatory cytokines such as IL-1β, IL-6, IL-8, and TNF-α in patients with cannabis use disorder.Reference Bayazit, Selek, Karababa, Cicek and Aksoy 96 Regarding alcohol abuse, the available evidence largely supports our findings. Studies have shown that binge drinking leads to elevated plasma levels of inflammatory cytokines,Reference González-Quintela, Dominguez-Santalla, Pérez, Vidal, Lojo and Barrio 97 with chronic alcohol consumption resulting in the release of IL-1β, IL-6, IL-12, and TNF-α by dendritic cells in patients who do not have liver disease.Reference Laso, Vaquero, Almeida, Marcos and Orfao 98 Post-mortem analyses have also revealed increased C-reactive protein (CRP) levels and elevated IL-6 and IL-10 in the femoral venous blood of individuals with alcoholic ketoacidosis.Reference Palmiere and Augsburger 99 Furthermore, in animal models, binge drinking and prolonged alcohol exposure have been associated with increased levels of IL-6, IL-1β, TNF-α, and IL-10.Reference Zhou, Jiang, Zhao, Zhen, Wang and Nanji 100 Chronic opioid use has similarly been shown to elevate levels of peripheral cytokines such as IL-6, IL-1β, TNF-α, and CRP.Reference Zhou, Jiang, Zhao, Zhen, Wang and Nanji 100 –Reference Mohamed and Mahmoud 103 The release of proinflammatory cytokines within the central nervous system has also been linked to behaviors associated with opioid addiction. In particular, increased expression of IL-6, TNF-α, and IL-1β has been observed in the nucleus accumbens of individuals with opioid addiction,Reference Chen, Tao and Chu 102 , Reference Zhang, Li and Li 104 as well as in the prefrontal cortex for IL-6 and IL-1β. Reference Chen, Tao and Chu 102 Lastly, in individuals with cocaine addiction, immune system activation has been documented, including significant increases in the proinflammatory cytokine IL-6, alongside a reduction in the anti-inflammatory cytokine IL-10.
The hypothesis suggesting the presence of an inflammatory state in individuals with SAD, which may eventually contribute to shifting TRP metabolism toward the KYN pathway at the expense of 5-HT synthesis, is further supported by our findings. Specifically, we observed a positive correlation between QUIN and a negative correlation between TRP and SHY-SR scores. In fact, if normally part of TRP metabolism is oriented toward the synthesis of 5-HT,Reference Richard, Dawes, Mathias, Acheson, Hill-Kapturczak and Dougherty 105 in conditions of acute or chronic inflammation, pro-inflammatory cytokines such as IL-6 can activate the enzyme IDO, which catalyzes the first step of the KYN pathway.Reference Fernstrom 106 , Reference Guillemin, Cullen and Lim 107 This pathway is also linked to chronic stress and pathological conditions such as anxiety, depression, and neuropsychiatric disorders. When IDO is activated, TRP is converted into KYN, which can follow different metabolic routes. One of these pathways produces KYNA, an NMDA receptor antagonist with neuroprotective properties, which, in SAD samples, was found to be reduced.Reference Savitz 108 Conversely, KYN can also lead to the production of QUIN, an NMDA receptor agonist that correlates positively with social phobic symptoms. QUIN is potentially neurotoxic, contributing to neuronal excitotoxicity, which can damage brain areas responsible for mood regulation and stress responses.Reference Guillemin 109 –Reference Williams, Zhang and Nance 111 Thus, the increased activity of IDO may divert TRP metabolism away from 5-HT synthesis, reducing 5-HT availability while simultaneously increasing the levels of toxic metabolites such as QUIN. This shift in metabolic pathways could contribute to the pathophysiology of social anxiety and related symptoms.
Similarly, our results showed a positive correlation between HCY levels and the SHY-SR total score, suggesting that higher HCY levels may be associated with greater symptomatology. HCY is a sulfur-containing amino acid that is produced as an intermediate during the metabolism of methionine. Elevated levels of HCY have been linked to inflammation and several neuropsychiatric disorders.Reference Chung, Chiou and Chen 112 –Reference Moradi, Lotfi, Armin, Clark, Askari and Rouhani 114 HCY has been studied in relation to anxiety disorders, although the body of research in this area is still limited compared to other fields. However, some studies, especially in animal models, have explored the association between elevated HCY and the onset or severity of anxiety disorders, suggesting that higher HCY levels may be linked to neurochemical dysfunction and systemic inflammation that may impact anxiety.Reference Dos Santos, Siebert, Bobermin, Quincozes-Santos and Wyse 115 –Reference Wyse, Sanches, Dos Santos, Siebert, Kolling and Netto 117 Although to date there is a paucity of studies with conflicting results and none with reference to HCY and SAD, it has been suggested that the methionine-HCY metabolic pathway plays an important role in the methylation reactions involved in the biosynthesis of monoamine neurotransmitters, particularly of 5-HT, dopamine, and norepinephrine, which are critical for mood regulation and anxiety control.Reference Folstein, Liu and Peter 118 , Reference Kim, Stewart, Kim, Yang, Shin and Yoon 119 In particular, it has been suggested that HCY may interfere with 5-HT production by affecting the methylation cycle, a biochemical pathway essential for neurotransmitter synthesis. Elevated HCY levels are associated with folate and vitamin B12 deficiencies, which serve as crucial cofactors in the conversion of HCY to methionine via the enzyme methionine synthase. This step is necessary for the production of S-adenosylmethionine (SAMe), a key methyl donor for 5-HT synthesis. When HCY accumulates, these cofactors are less available, hindering 5-HT production and contributing to mood and anxiety disorders.Reference Bhatia and Singh 120 Moreover, elevated HCY may promote oxidative stress and inflammation, which can negatively affect brain regions like the amygdala, implicated in anxiety and fear processing.Reference Chen, Dong and Zhao 121 Ultimately, it is conceivable that chronic stress induced by exposure to social situations activates the HPA axis, increasing the secretion of cortisol, the increase of which has been correlated with elevated HCY levels.Reference Cascalheira, Parreira, Viegas, Faria and Domingues 122 In turn, elevated HCY exacerbates oxidative stress, further stimulating HPA axis activity and sustaining the stress response in a vicious cycle.Reference Kobayashi, Machida and Takahashi 123 , Reference Tyagi, Sedoris, Steed, Ovechkin, Moshal and Tyagi 124
Moreover, HCY appeared to be positively correlated with SHY-SR Childhood and Adolescence, Interpersonal sensitivity, Behavioral inhibition, and Specific Anxieties and Phobic Features domains, being also significantly associated with the first and last domains. Alterations in HCY levels and consequently vitamin B12 deficiencies in early childhood have been recognized as an important cause of neurodevelopmental delay and regression.Reference Hasbaoui, Mebrouk, Saghir, Yajouri, Abilkassem and Agadr 125 Furthermore, HCY can disrupt DNA methylation, an essential epigenetic mechanism that regulates gene expression during brain development and function.Reference Li, Dang and Yan 126 Elevated HCY levels are associated with alterations in neurotransmitter systems, such as glutamate and gamma-aminobutyric acid (GABA), both of which are crucial for brain function and have been hypothesized to contribute to some of the development of autistic-like symptoms such as social withdrawal in animal models.Reference Wang, Wang and Zhao 127 Moreover, HCY levels have been associated with social isolation and depression in elderly mice.Reference Wei, Huang and Zeng 128 In particular, it emerged that elevated HCY levels induced by a high methionine diet mimicked the depressive-like behaviors and that social isolation influenced HCY levels. The authors therefore hypothesized a pivotal role of HCY in driving social isolation-induced depressive-like behaviors. However, it should be noted that there are several clinical differences between the SHY-SR Interpersonal sensitivity domain and the above-mentioned “models,” leading to a word of caution in the interpretation of these results. Regarding behavioral inhibition, to date elevated HCY levels have been associated with decreased cognitive function, poor memory, and impaired executive function, which also includes behavioral inhibition. This is particularly true for individuals with neurodegenerative diseases, such as Alzheimer’s disease, or those with mild cognitive impairment.Reference Luzzi, Papiri and Viticchi 129 , Reference Troen and Rosenberg 130 Moreover, as previously mentioned, HCY may interfere with neurotransmitter systems, including those involving dopamine, 5-HT, and glutamate, all of which play a role in mood regulation and cognitive control. Disruptions in these systems could potentially affect behavioral inhibition.Reference Bhatia and Singh 120 Lastly, although still scant, some evidence is highlighting how elevated HCY levels may be implicated in anxiety and fear processing. Indeed, oxidative stress and increased inflammation, caused by elevated HCY levels, can negatively affect brain regions like the amygdala, which is involved in the processing of phobic and anxious stimuli.Reference Chen, Dong and Zhao 121
Consistent with previously discussed data, our results showed a negative correlation between 5-HT plasmatic levels and SHY-SR scores. Results of regression analyses also suggest that higher 5-HT plasmatic levels were associated with less social phobic symptomatology. Despite consistent evidence linking functional deficits in 5-HT neurotransmission within emotion-regulating brain circuits to the pathophysiology of psychiatric disorders like anxiety and depression,Reference Belmaker and Agam 131 direct evidence of reduced 5-HT levels in the synaptic clefts of affected individuals remains elusive. This is primarily due to the absence of techniques capable of directly measuring synaptic 5-HT levels in the brain. As a result, researchers have long investigated peripheral markers, particularly platelet abnormalities, as potential reflections of central 5-HT dysfunction. In this context, platelet parameters—such as 5-HT content, 5-HT uptake activity, and SERT binding—have been proposed as valuable tools for elucidating the mechanisms of antidepressant action and predicting clinical responses to treatment.Reference Axelson, Perel and Birmaher 132 , Reference Franke, Schewe, Uebelhack and Müller-Oerlinghausen 133 These peripheral measures offer an indirect but promising approach to understanding serotonergic dysfunction in psychiatric disorders. In particular, studies examining platelet 5-HT content and SERT-mediated 5-HT uptake reported that SAD patients showed higher levels of platelet 5-HT, and how the same was significantly reduced after SSRI treatment.Reference Zhuang, Xu, Fang, Xu, Xue and Hong 134 Given that platelets are widely regarded as a useful peripheral model for assessing 5-HT levels and SERT activity in neuropsychiatric research, our study aimed to evaluate both platelet and plasma 5-HT levels. Interestingly, while we observed a reduction in plasma 5-HT levels, platelet 5-HT levels did not show significant differences compared to those found in HCs.
The evidence of a negative correlation between plasmatic 5-HT levels and SAD symptomatology is in line with the previously discussed data on the influence of IL-6 and HCY levels on the synthesis of monoamines, and in particular 5-HT. In fact, HCY homocysteine may interfere with 5-HT production by affecting the methylation cycle, while the inflammatory state associated with high IL-6 levels may shift the TRP and KYN pathway toward the production of QUIN at the expense of 5-HT synthesis.
5-HT plays a crucial role in various biological functions, including learning, memory, mood regulation, appetite, sleep, reproduction, circadian rhythms, thermoregulation, pain perception, and social behavior.Reference Kriegebaum, Gutknecht, Schmitt, Lesch and Reif 135 Although the findings on this topic are sometimes inconsistent, several studies have observed reduced levels of 5-HT in the peripheral blood of individuals with depression or anxiety.Reference Maurer-Spurej, Pittendreigh and Solomons 136 –Reference Quintana 138 However, the precise mechanisms by which peripheral 5-HT might influence brain activity remain unclear, especially considering that it cannot cross the blood–brain barrier in adults.Reference Berger, Gray and Roth 139
Additionally, 5-HT is thought to play an important role in regulating social behavior.Reference Crockett, Clark and Robbins 140 –Reference Kiser, Steemers, Branchi and Homberg 142 Research indicates that in both primates and humans, 5-HT levels are positively associated with prosocial behaviors such as caregiving, cooperation, and affiliation, and negatively associated with antisocial behaviors like aggression and social withdrawal.Reference Higley, King, Hasert, Champoux, Suomi and Linnoila 143 , Reference Raleigh, Brammer, Yuwiler, Flannery, McGuire and Geller 144 Some authors have hypothesized that 5-HT may influence these social behaviors by modulating neuroplasticity, thus affecting the brain response to environmental changesReference Kiser, Steemers, Branchi and Homberg 142 , Reference Belsky, Jonassaint, Pluess, Stanton, Brummett and Williams 145 –Reference Uher 148. Moreover, the interaction between 5-HT and the environment appears to be bidirectional. The 5-HT system, responsible for mediating responses to the environment, must adapt when changes occur, particularly in social contexts. This reciprocal interaction is particularly important in the social domain, since 5-HT influences sensitivity and subsequent response to social cues and, conversely, social behavior, or the general social environment, influences the functioning of the 5-HT system. Given that deficits in social behavior are characteristic of several psychiatric disorders—including anxiety disorders like SAD, depression, autism, and schizophrenia—understanding 5-HT’s role in social behavior and the dynamic interplay between the two is a critical but currently unmet need for advancing treatment options.
Furthermore, 5-HT, KYNA, and TRP levels appeared to be negatively correlated with SHY-SR Specific Anxieties and Phobic Features, Behavioral inhibition, and Interpersonal sensitivity, with 5-HT also emerging as negatively associated with the first two domains, while KYNA being negatively associated exclusively with the SHY-SR Behavioral inhibition domain. In this framework, the role of inflammation in social interactions could help explain our result. Research on the effect of 5-HT levels and alterations in the KYN pathway and interpersonal sensitivity are still scant, although some promising results are emerging. For instance, some clinical trials reported a positive influence of a TRP-enriched diet on social cognition,Reference Reuter, Zamoscik and Plieger 149 , Reference Zamoscik, Schmidt and Bravo 150 while gestational TRP fluctuations were associated with neuroembryogenesis and psychosocial development in humans and viviparous animals.Reference Huang, Feng and Cheng 151 Moreover, Kanen et al.Reference Kanen, Arntz and Yellowlees 152 investigated the influence of 5-HT on emotional reactions to social conflict in healthy volunteers, using a computerized task involving unjust harm and reported how different personalities, combined with fluctuations of 5-HT, resulted in diverse emotional phenotypes. Supporting these results, a recent study conducted by Peters et al.Reference Peters, Yilmaz, Li and Balbuena 153 investigated the role of SSRIs in reducing the outcome of interpersonal sensitivity using a specific factor from the Symptom Checklist. The authors collected data regarding 1709 patients from three randomized, double-blind, placebo-controlled trials on fluoxetine and paroxetine in subjects suffering from acute major depressive disorder. Both medications were more efficacious in reducing interpersonal sensitivity in comparison to the placebo. The effect of these drugs was independent of the depression improvement and was not related to the baseline interpersonal sensitivity. This important finding led the authors to suggest a pivotal role of the biochemical cascade of 5-HT in modulating the specific aspect of interpersonal sensitivity among mood disorders and other psychiatric illnesses, and its potential as a therapeutic target.Reference Peters, Yilmaz, Li and Balbuena 153 Conversely, the negative correlation found between plasmatic KYNA levels and Interpersonal sensitivity is more controversial. Indeed, to date, specific literature arising from similar clinical contexts is still lacking, and very few studies are targeted toward the analysis of such psychopathological dimensions. Furthermore, Huang and colleagues reported increased KYNA levels correlated with reduced social cognition in schizophrenic subjects, while others reported normal or reduced plasma levels of KYNAReference Peters, Yilmaz, Li and Balbuena 153 –Reference Plitman, Iwata and Caravaggio 155. In particular, a recent study reported that a reduction in KYNA levels was significantly associated with depressive symptoms. Specifically, a one-unit increase in KYNA was associated with a 0.21-unit decrease in the Depression Anxiety and Stress Scale-21 depression score, supporting the hypothesis of KYNA’s role as an endogenous glutamatergic inhibitor in regulating depression severity.Reference Francis, Stevenson and Tan 156 In conclusion, the data now available are quite heterogeneous and inconclusive, and differences in the results may be flawed by the absence of a cerebral fluid dosage of metabolites, leading to a partial interpretation of convergences and divergences.
Research on the role of TRP metabolism and its associated pathways, particularly the KYN and 5-HT systems, and behavioral inhibition remains limited and somewhat inconsistent, with most studies conducted on animal models. It has been suggested that 5-HT plays a role in behavioral inhibition by limiting or preventing actions in anticipation of aversive outcomes.Reference Gray 157 –Reference Soubrié 159 There is evidence supporting the involvement of 5-HT in the learning of Pavlovian predictions related to future negative outcomes, which leads to inhibitory responses. While this suggests a connection, causal links between these processes remain less well established.Reference Dayan and Huys 160 Additionally, some studies have identified a causal relationship between reduced 5-HT levels and impairments in behavioral inhibition, which in turn may contribute to increased aggression and impulsivity. For instance, pharmacologically reducing 5-HT levels in rats has been shown to enhance aggressive behaviors.Reference Di Chiara, Camba and Spano 161 , Reference Vergnes, Depaulis and Boehrer 162 Conversely, increasing 5-HT transmission—through interventions like SSRIs or genetic deletions of the 5-HT transporter—has been associated with reduced aggression.Reference Fuller 163 , Reference Homberg, Pattij and Janssen 164 Overall, diminished 5-HT function seems to impair the processing of aversive stimuli in animal studies,Reference Graeff 165 but this effect appears to be reversed in humans, where 5-HT reduction tends to increase sensitivity to aversive outcomes.Reference Cools, Roberts and Robbins 166 A key distinction between these studies is that animal research typically uses behavioral metrics (e.g., lever pressing) to assess aversive prediction, while human studies often rely on cognitive measures (e.g., accuracy in decision-making tasks). If 5-HT generally facilitates the avoidance of aversive stimuli (e.g., through punishment-induced inhibition), then a decrease in 5-HT would logically reduce this avoidance behavior. This has been confirmed in animal models, where global 5-HT depletion in the forebrain of rats led to decreased avoidance of aversive stimuli.Reference Tye, Everitt and Iversen 167 Human research has focused more on nonreinforced motor inhibition, particularly studies like that of Young et al.,Reference Young, Smith, Pihl and Ervin 168 who used acute TRP depletion to lower 5-HT levels in healthy human participants. Their findings suggested that 5-HT plays a significant role in behavioral inhibition, especially when aversive outcomes are anticipated (such as in punishment-induced inhibition), rather than influencing motor response inhibition more generally, and that TRP, by being a precursor to 5-HT, plays an indirect but crucial role in behavioral inhibition.
Moreover, 5-HT is now recognized as playing a role in the regulation of fear and anxiety, although its exact role may be complex and context-dependentReference Burghardt, Bush, McEwen and LeDoux 169 , Reference Deakin and Graeff 170 and the hypotheses currently available are concordant with our results. In particular, 5-HT may regulate anxiety and fear reactions through modulation of the amygdala, a brain region critically involved in the processing of fear and emotional responses.Reference Parent, Descarries and Beaudet 171 , Reference Ravinder, Burghardt, Brodsky, Bauer and Chattarji 172 Low levels of 5-HT are often associated with increased emotional reactivity, including increased fear and anxiety.Reference Lowry, Johnson, Hay-Schmidt, Mikkelsen and Shekhar 173 Conversely, 5-HT signaling may help dampen excessive activity in the amygdala, reducing the intensity of fear responses. It may also play a role in the regulation of the prefrontal cortex, which is involved in the regulation and inhibition of fear responses. Research also suggests that 5-HT may play a role in facilitating fear extinction, which is important for adapting to new experiences and reducing maladaptive fear responses.Reference Tortora, Hadipour, Battaglia, Falzone, Avenanti and Vicario 174 , Reference Vicario and Martino 175 In this context, TRP, being a precursor of 5-HT, acts indirectly on the regulatory circuits of anxiety and fear. Furthermore, KYNA, being a non-competitive antagonist of the NMDA receptor involved in glutamatergic signaling, can modulate the excitability of neurons involved in fear processing, including those in the amygdala and prefrontal cortex. This may influence both the acquisition and extinction of fear memories. In particular, high levels of KYNA can dampen NMDA receptor activity, potentially reducing the emotional intensity of fear responses.Reference Battaglia, Di Fazio and Battaglia 176
Finally, our findings revealed a negative correlation between platelet BDNF levels and SHY-SR scores. BDNF is the most abundant member of the neurotrophin family and plays a key role in the growth and survival of neurons.Reference Leibrock, Lottspeich and Hohn 177 It is well-established that BDNF is essential for the formation, branching, and strengthening of synaptic connections.Reference Binder and Scharfman 178 , Reference Huang and Reichardt 179 BDNF is present not only within the central nervous system but also in peripheral tissues, and much research has shown that the levels of BDNF found in circulation are reflective of its concentrations in the brain.
To date, only a limited number of studies have investigated BDNF levels in individuals with SAD, and, to the best of our knowledge, this is the first study to examine its intraplatelet concentrations.Reference Carlino, Francavilla and Baj 180 , Reference Molendijk, Bus and Spinhoven 181 Although existing research on BDNF and SAD remains inconclusive, the topic continues to attract significant interest as a potential biomarker for both neurological and psychiatric conditions. This growing interest is fueled by emerging evidence from animal models, where altered BDNF levels are associated with anxiety-like behaviors, possibly linked to increased spine density in the amygdala.Reference Govindarajan, Rao and Nair 182 Overall, previous studies assessing BDNF levels in SAD patients have not reported significant differences compared to the general population.Reference Carlino, Francavilla and Baj 180 However, these findings should be interpreted cautiously, given several limitations, including the fact that even though BDNF is mostly stored in platelets and released after degranulationReference Govindarajan, Rao and Nair 182 , Reference Fujimura, Altar and Chen 183 no prior studies have specifically addressed intraplatelet BDNF levels.
Limits
Our findings must be considered in light of several important limitations. First, the relatively small sample size restricts the generalizability of our results, making it difficult to extend these findings to larger, more diverse populations. Additionally, the limited sample size prevented us from adequately accounting for potential confounding variables, such as the interaction between ASD and SAD diagnoses. This interaction could influence the outcomes and warrant further exploration.
Another limitation is the reliance on self-report questionnaires, which inherently carry the risk of response bias. Participants may either underreport or exaggerate their symptoms, leading to a skewed understanding of the true nature of their experiences. Moreover, the cross-sectional design of our study precludes any conclusions about the directionality or causality of the relationships between the variables examined. Without longitudinal data, we cannot ascertain whether the observed associations are causal or merely reflective of co-occurring conditions. In addition, for the purpose of this work, we did not take into account the psychiatric comorbidities in the ASD and BAP groups due to the limited sample size, which prevented us from performing further data stratifications. Given the well-established association between mood disorders and inflammatory alterations, the presence of comorbid conditions may have influenced the observed biochemical findings and should be considered as a potential confounding factor.
In light of these limitations, future research with larger, more representative samples and a longitudinal design is needed to further elucidate the prevalence, underlying mechanisms, and biochemical correlates of SAD within the context of autism spectrum conditions. Such studies will be crucial in advancing our understanding of this complex relationship and in developing more targeted interventions.
Conclusion
Despite the aforementioned limitations, our study represents one of the few to date that investigates the biochemical correlates of SAD in a sample with ASD. Notably, it is the only study that integrates the analysis of both inflammatory markers and key metabolites from the KYN and 5-HT pathways. Additionally, our study is the first to examine intraplatelet levels of both 5-HT and BDNF in this context, providing novel insights into the biochemical landscape of SAD in ASD.
Consistent with existing literature, our findings suggest a heightened presence of social phobic symptoms among individuals with autism spectrum conditions. This appears to be accompanied by a greater tendency to rely on benzodiazepines as a coping mechanism for addressing social deficits. From a biochemical standpoint, our results indicate elevated levels of HCY and IL-6, as well as reduced levels of KYNA and 5-HT, further underscoring the potential dysregulation of neurochemical and inflammatory systems. These biochemical markers were also found to correlate with specific psychopathological dimensions, supporting the hypothesis that the inflammatory cascade might play a critical role in the maintenance and exacerbation of SAD symptoms within the ASD population.
Together, our findings point to the importance of considering both inflammatory and neurochemical factors in understanding the complex interplay between ASD and SAD. Future research, ideally with larger and more diverse samples, will be essential in clarifying the underlying mechanisms of this relationship and in informing targeted therapeutic strategies that address both the biological and behavioral aspects of these co-occurring conditions.
Acknowledgements
Not applicable.
Author contribution
Conceptualization: B.N., S.P., A.S., G.G., L.D., B.C.; Data curation: B.N., V.T., C.P., M.P., G.M.; Investigation: B.N., L.P., L.P., C.B., l.b., B.C.; Writing - original draft: B.N., C.B., V.T., C.P., M.P.; Writing - review & editing: B.N., V.T., C.P., M.P.; Formal analysis: L.P., L.P., G.M., B.C.; Methodology: L.P., L.P., G.M., S.P., A.S., l.b., G.G., B.C.; Supervision: L.P., L.P., G.M., S.P., A.S., l.b., G.G., L.D., B.C.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Disclosures
The authors do not have competing interests to disclose.
Ethical statement
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.
Open access
