The study of the HTT gene in a cohort of 69 patients with bipolar disorder did not identify any subjects in a pre-motor stage of HD, in fact no patient was a carrier of pathological allele. The IA was detected in 10.14% of cases. IA-carriers were older as compare with non- IA carriers at disease onset. Although psychiatric symptoms may often precede the motor phase of HD, there are few previous studies evaluating the frequency of the pathological expansion in the HTT gene in psychiatric patients: two studies in patients affected by major depression [
15,
16] and one in a Brazilian cohort affected by bipolar disorder [
17]. Major depression was frequent as pre-motor phase of HD [
15], while Ramos and colleagues [
17], in line with our results, did not identify any association between the pathological HTT allele and bipolar disorder. The frequency of IA-carriers in our psychiatric cohort was almost the double than that detected in the local reference population. Data from the literature has already shown a higher risk of behavioral disturbances in subjects carrying IA [
9,
16], such as apathy, obsessive disorder, anxiety, depression, suicidal ideation. However, the mechanism by which the IA of HTT gene is associated with psychiatric manifestations remains unclear [
9]: one hypothesis is through mechanisms unrelated to those leading to HD pathology [
9], such as interaction with other genes [
9]. In that case, the difference in the genetic structure between races could explained the lack of association between IA and bipolar disorder in the Brazilian population [
17,
18]. The analysis of the distribution of the HTT normal alleles revealed a statistically significant difference between bipolar patients and controls, with shorter normal alleles among patients. These data support the hypothesis of a non-linear association between the CAG repeat length and the risk of bipolar disorder, which means that both the normal alleles in the shorter range and the IA could increase the risk of developing bipolar disorder. This non-linear association has been already described between the HTT CAG length and the risk of developing major depression [
16], and between the HTT size and the intelligence in the general population [
19]. Therefore, it seems that the increasing size of non-pathological HTT gene confers an advantage until the values that determine pathology. In fact, the increasing number of HTT triplets is part of the evolutionary process [
20] and the variation of the number of repeats across the normal range influences brain’s structure in healthy subjects [
21]: longer non-pathological alleles have been found associated with an increasing volume of basal ganglia [
21]. Studies on HD patients [
22,
23,
24] demonstrated that the increasing size of normal allele could be protective by mitigating the effect of the pathological one and delaying disease onset. Interestingly, in the present cohort we found that bipolar patients carrying IA were older at disease onset as compare with non-IA carriers. Basal ganglia are involved in the pathophysiology of bipolar disorder [
25,
26], so, IA, via its effect on basal ganglia [
21], could have a protective role and delay the onset of bipolar symptoms. Thus, the IA of HTT could act on the development of bipolar disorder with two different mechanisms: one that increases the risk of the disease and the other that postpones its onset. Another hypothesis could be that when the development of bipolar disorder is driven by the IA of HTT the age at onset is later than when it is caused by other genetic factors.