BDNF, proBDNF and IGF-1 serum levels in naïve and medicated subjects with autism

Abstract

Brain-derived neurotrophic factor (BDNF) and insulin-like growth factor 1 (IGF-1) promote the development and maintenance of neural circuits. Alterations in these factors might contribute to autism spectrum disorder (ASD). We asked whether serum BDNF, proBDNF, and IGF-1 levels are altered in an ASD population compared to controls. We measured serum BDNF, proBDNF, and IGF-1 immunoreactive protein in boys and girls aged 5-15 years old with mild to moderate ASD and non-autistic controls by ELISA. IGF-1 was increased in ASD serum compared to controls and was correlated with age and with CARS scores. Serum BDNF levels did not differ between groups, however, proBDNF serum levels were decreased in subjects with ASD compared to non-autistic controls. Medicated, but not unmedicated, ASD subjects exhibited lower serum proBDNF levels compared to controls, while neither IGF-1 nor BDNF levels differed between treatment groups. These data support the involvement of proBDNF and IGF-1 in the pathogenesis and treatment of autism.

Figure 1

Serum protein levels measured by ELISA. (a) IGF-1, (b) total BDNF, (c) proBDNF and (d) the ratio of proBDNF/total BDNF immunoreactive protein levels in serum of controls (n = 29, or n = 28 where indicated) and subjects with autism (n = 22). 2-tailed Mann–Whitney test, IGF-1 *p = 0.037, total BDNF p = 0.350, ns not significant; proBDNF **p = 0.005, controls n = 28; proBDNF/total BDNF *p = 0.024, controls n = 28. Data are shown as mean ± SEM.

Figure 2

Serum BDNF and IGF-1 protein levels in unmedicated and medicated subjects. BDNF and IGF-1 immunoreactive protein levels in serum of unmedicated (n = 8) and medicated (n = 14) subjects with autism and in non-autistic controls (n = 29 or n = 28 where indicated). Kruskal–Wallis test, (a) total BDNF, p = 0.073; (b) proBDNF **p = 0.004 followed by Dunn’s test unmedicated vs. medicated p = 0.187, unmedicated vs. controls p > 0.999, medicated vs. controls **p = 0.003, controls n = 28; (c) proBDNF/total BDNF p = 0.070, controls n = 28; (d) IGF-1 p = 0.108. Data are shown as mean ± SEM.

Figure 3

Correlations between serum IGF-1 and BDNF immunoreactivity and age. Spearman’s correlations: (a) IGF-1 correlation of all samples with age, r_s = 0.345, p = 0.013, n = 51. (b) IGF-1 correlation of each group separately with age (control group r_s = 0.207, p = 0.282, n = 29; autism group r_s = 0.429, p = 0.046, n = 22). (c) Total BDNF correlation of all samples with age, r_s = − 0.233, p = 0.101, n = 51. (d) Total BDNF and proBDNF correlations with age in the control group (total BDNF r_s = − 0.436, p = 0.018, n = 29; proBDNF r_s = 0.516, p = 0.005, n = 28). (e) Total BDNF and proBDNF correlations with age in the autism group (total BDNF r_s = − 0.010, p = 0.966, n = 22; proBDNF r_s = − 0.205, p = 0.359, n = 22. (f) Correlation of proBDNF/total BDNF in each group separately with age (control group r_s = 0.469, p = 0.012, n = 28; autism group r_s = − 0.075, p = 0.739, n = 22). Open triangles, controls; filled circles, autism; open diamonds, total BDNF; closed diamonds, proBDNF.

Figure 4

Correlations between Childhood Autism Rating Scale (CARS) scores and IGF-1 and proBDNF/BDNF. (a) IGF-1 immunoreactivity, Spearman’s correlation r_s = 0.486, p = 0.022 and (b) ratio proBDNF/BDNF, Spearman’s correlation r_s = − 0.407, p = 0.060. n = 22.