Introduction
Hereditary spastic paraplegia (HSP) is a heterogeneous group of hereditary neurodegenerative disorders mainly characterized by spasticity and weakness, predominantly affecting the lower limbs. Reference Harding1 Patients with additional manifestations are classified as the complicated form of HSP in contrast to the pure form of the disease. HSP has several inheritance patterns, including autosomal dominant (AD-HSP), autosomal recessive (AR-HSP), X-linked, and mitochondrial/maternal inheritance. De novo mutations can also happen in HSP. AD-HSP patients usually present a pure form of the disease, while AR-HSP cases often show a complicated form. Reference Klebe, Stevanin and Depienne2 The most frequent mutated gene among all patients and specially AD-HSP ones is the SPAST gene. It usually results in a pure form of HSP and comprises approximately 60% of AD-HSPs (accounting for 17–79% of all AD-HSP cases), a third of all HSP-affected patients, and 15% of sporadic cases. Reference Erfanian Omidvar, Torkamandi and Rezaei3,Reference Shribman, Reid, Crosby, Houlden and Warner4 The SPAST gene encodes for an enzyme called spastin, an AAA protein (ATPases associated with a variety of cellular Activities). Reference Errico, Ballabio and Rugarli5 Heterozygous mutations in SPAST cause the SPG4 type of HSP. Up to now, more than 900 mutations throughout SPAST have been reported in the Human Gene Mutation Database (HGMD professional 2020.4), indicating a high level of allelic heterogeneity. The most common type of SPAST mutations are point mutations (75–80%), and the frequency of large rearrangements such as exon deletions or duplications in SPAST has been estimated as 20%. Reference Beetz, Nygren and Schickel6,Reference Kadnikova, Rudenskaya, Stepanova, Sermyagina and Ryzhkova7 It is suggested based on type and where the mutation occurred, pathology and phenotypes might vary among SPG4 patients. So, a high level of both inter and intrafamilial clinical heterogeneity are observed in SPG4 patients Reference Parodi, Fenu and Barbier8 : scrutinized assessment of patients show differences in age at onset (AAO; ranging from early childhood to the seventh decade), disease progression pace, and clinical variability. In addition to variable expressivity in SPG4 patients, some studies indicate incomplete penetrance in these patients. Reference Parodi, Fenu and Barbier8 Thus, the role of environmental and other genetic factors should be considered for explaining this AAO and clinical heterogeneity. Reference Parodi, Fenu and Barbier8–Reference Fink10 Such variability could be due to (i) the genetic modifiers; for example, the presence of the c.131C>T:p. Ser44Leu variant in the patients who carry a SPAST mutation is associated with a lower AAO and more severe disease, Reference Parodi, Fenu and Barbier8 (ii) the type and location of SPAST mutations, and (iii) sex. A bimodal pattern of distribution of AAO has been reported in SPG4 patients that is related to the type of mutations; patients who carry the missense mutations in SPAST manifest a significantly lower AAO (often <10 years and located in the first peak) than those with truncating mutations (the interval of the second and fifth decade of life and located in the second peak). Reference Solowska and Baas11 There is no informative correlation between the type of mutation and severity of symptoms except for intellectual disability in some patients who carried the missense mutations Reference Loureiro, Brandao and Ruano12 and psychiatric disorders among a few patients carrying a loss of function mutation in the SPAST gene. Reference Chelban, Tucci and Lynch13 Also, a sex-linked penetrance has been reported in SPG4 patients; a lower penetrance was observed in females. Reference Solowska and Baas11
Anticipation is a biological event, defined by a progressive decline in AAO, and in most cases, escalation in disease severity in successive generations in a pedigree is also apparent. Reference Paulson14 It is typically related to the context of dynamic mutations and detected in several diseases including Huntington’s disease (HD), Reference Ranen, Stine and Abbott15 myotonic dystrophy (MD), Reference Redman, Fenwick, Fu, Pizzuti and Caskey16 amyotrophic lateral sclerosis (ALS with repeat expansion, RE, mutations in C9orf72), Reference Paulson14 and spinocerebellar ataxia (SCA). Reference Imbert, Saudou and Yvert17 It is noteworthy that this phenomenon has been reported in other neurological and non-neurological disorders without dynamic mutations, like some types of cancers, including breast or ovarian cancer, Reference El Tannouri, Albuisson, Jonveaux and Luporsi18,Reference McInnis19 Lynch syndrome, Reference Bozzao, Lastella and Stella20 Crohn’s disease, Reference Bayless, Picco and LaBuda21 ALS, Reference Alavi, Nafissi and Rohani22,Reference Iwai, Yamamoto, Yoshihara and Sobue23 hypertrophic cardiomyopathies, Reference Guo, Fan and Wang24 and a few subtypes of HSP such as SPG4. Reference Bönsch, Schwindt and Navratil25–Reference Reddy, Seltzer and Grewal34 However, the presence of anticipation in SPG4 cases is still doubtful, and the presence of anticipation may be due to diagnostic sensitivity of these families. So, further studies of large pedigrees with SPAST mutations and maybe re-analysis of previous reported SPG4 families are needed to confirm whether anticipation exists in the SPG4 families. Confirmation of anticipation in these diseases and identification of their related factors may reveal the pathophysiology of those and help the genetic counseling. The importance of analyzing genetic anticipation is that clinicians to decide from what age they should follow presymptomatic carriers. Reference Van Mossevelde, van der Zee and Gijselinck35
Herein we assessed this phenomenon in the identified Iranian SPG4 families and compared it to other reported families. Actually, this phenomenon might be more common than expected in these families, but due to inadequate knowledge and the small size of families, it is often overlooked or underestimated in clinical practice.
Materials and Methods
This research was performed in accordance to the Declaration of Helsinki and with the approval of the ethics board of the University of Social Welfare and Rehabilitation Sciences (USWR). All patients and their family members were informed of the nature of the research and the consent form signed.
Subjects
Fourteen unrelated Iranian families affected with AD-HSP referred to Genetics Research Center (GRC) at the USWR for genetic analysis. All affected and unaffected members of these families were recruited when available. Clinical data of unavailable affected members were also collected by in depth interviewing family members or based on their previous medical records.
Genetic Analysis
Whole-Exome Sequencing (WES)
DNA was extracted from whole blood using the salting-out protocol. DNAs of probands were whole-exome sequenced and subsequently analyzed to detect candidate disease-causing variants. Detail of analysis has been presented in the Supplementary text A.
In silico Analysis
To detect the pathogenic nature of variants in the corresponding proteins, several in silico bioinformatics tools were used (Supplementary text B). For the clinical interpretation of variants, the American College of Medical Genetics (ACMG) criteria were used. Reference Richards, Aziz and Bale36 ACMG criteria were checked in InterVar; http://wintervar.wglab.org/ and Varsome; https://varsome.com/.
Co-segregation Analysis of Candidate Variants
Amplification of exons 5 and 13 of the SPAST gene that carried the candidate variants c.C782A:p. Ser261*, c.806dupA:p. Tyr269_Ser270delins*, respectively, in the probands SPG4-A-IV5, SPG4-C-II3, and c.G1496A:p. Arg499His in the probands SPG4-D-III4 and SPG4-E-III6 was done by polymerase chain reaction (PCR). The PCR products were sequenced using the Sanger method (Big Dye kit and the Prism 3130 sequencer; Applied Biosystems, Foster City, CA, USA). Sequences were checked by Sequencher 4.1 software. Sequence variants were assessed with the reference sequences available at NCBI: NC_000002.11, NM_014946.4, and NP_055761.2 and confirmed in the probands. Thereafter, to perform a co-segregation analysis of the candidate variants in affected families based on disease-status, direct sequencing of corresponding exons was performed in the parents, siblings, and other family members.
Co segregation analysis was also done for families with KIF5A (c.A758T:p. Lys253Met), ATL1 (c.C715T:p. Arg239Cys), and MFN2 (c.G380A:p. Gly127Asp) variants. Assessing of these sequence variants was done using the reference sequences available at NCBI: NC_000012.11, NM_004984.4, and NP_004975.2 for KIF5A, NC_000014.8, NM_015915.5, and NP_056999.2 for ATL1, and NC_000001.10, NM_014874.4, and NP_055689.1 for MFN2.
Multiplex Ligation-Dependent Probe Amplification (MLPA)
MLPA was performed in seven remaining families (undiagnosed families by WES), including family SPG4-B who presented anticipation. MLPA was carried out using the SALSA® MLPA® Probemix P165-C3 HSP mix-1 (MRC-Holland, Amsterdam, The Netherlands), according to the manufacturer’s protocols detecting large deletions or duplications in the SPAST and ATL1 genes. Data were analyzed by Coffalyser.NET software. The SALSA MLPA Probemix P165-C3 contains 47 probes with amplification products between 130 nucleotides and 481 nucleotides. It includes 16 probes for the ATL1 gene and 20 probes for the SPAST gene. In addition, 11 reference probes are included that detect autosomal chromosomal locations.
Evidence for Possible Anticipation
AAO of affected parents and their affected offsprings was considered and compared to detect probable anticipation in all families, especially in five Iranian SPG4 families (families without mutation in the SPAST gene did not present anticipation, so their anticipation analysis was not included in this study). Clinical manifestations of 39 SPG4 patients (excluding asymptomatic individuals of SPG4-A, SPG4-C, and SPG4-E families, Table 1) from these families were recorded and used to assess anticipation (Table 2). AAO and present age of some previously reported SPG4 families were also collected and overviewed to confirm potential anticipation in these families (Supplementary Table S1).
Clinical findings of the Iranian families harboring the mutations in the SPAST gene
NR: not reported; F: female; M: male; LL: lower limb; -: negative; +: positive; Asymp: asymptomatic, ?: Undetermined.
All variants have been reported on the basis of NM_014946.
Probands have been shown in bold.
* Asterisks show the people in each pedigree that were genetically tested.
# AAO and clinical manifestations were determined based on their previous medical records.
& AAO and presence of spasticity were recorded based on interview with their relatives.
Trend of alternation of age at onset in successive generations of Iranian HSP families who carried SPAST mutations
AAO: age at onset; NM: normal-mutant.
Statistical Analysis
We used the “t test” to evaluate the correlation between two variables, i.e., AAO in parents and children. The correlations were mentioned as statistically significant when p-values were ≤0.05. Statistical calculations were done using online software (https://www.socscistatistics.com/tests/studentttest/default2.aspx).
Results
Clinical Features
Pedigrees of the five SPG4 families were shown in Figure 1A–E, and relevant information of them was presented in Table 1. Patients (total 39) showed intra- and inter familial clinical variability. The female to male ratio was 23:16. The mean AAO of all SPG4 patients was 24.46 ± 21.6 years, and the range of AAO was 2–70 years. The averages of AAO in females and males were 26.95 ± 21.1 and 20.87 ± 21.7, respectively, and the difference between them was not significant (p-value = 0.199878). The total average of the present age was 32.05 ± 19.2 with the range of 4–80 years (Tables 1 and 2).
The Iranian SPG4 pedigrees: (A) SPG4-A, (B) SPG4-B, (C) SPG4-C, (D) SPG4-D, (E) SPG4-E. The present age and genotypes of the candidate variants for each family are shown when individuals were assessed. Arrows show probands. Unfilled circles and squares, normal individuals; black circles and squares indicate SPG4 patients. Gray circles indicate asymptomatic individuals with heterozygous genotype. Grey circles with “?” indicate asymptomatic individuals, who have died but, we can consider as heterozygotes due to their heterozygous affected offspring and/or sib. Abbreviations: M, mutated allele; N, normal allele.
All patients had spasticity of lower limbs. In 20 out of 33 patients (60.6%; clinical data for some patients were not complete), knee jerk was brisk. Babinski sign was observed in 33 out of 34 patients (97%). Impaired vibration sense in feet, urinary dysfunction, and intellectual disability were presented in 13/33 (39.4%), 3/35 (8.5%), and 2/37 (5.4%), respectively, (Figure 2). Pes cavus, epilepsy, and dysarthria were observed in 9/34 (26.5%) and 7/35 (20%), and 13/33 (39.4%) patients, respectively. Hearing impairment and scoliosis were detected in two patients (Table 1 and Figure 2).
Clinical data of SPG4 families. Different clinical features of the 39 SPG4 patients have been shown. Data were not available for all patients.
Clinical manifestations of other AD-HSP carrying mutations in KIF5A, ATL1, and MFN2 were shown briefly in Supplementary Table S2.
Results of WES
Assessment of the WES data confirmed the high quality of sequencing in all probands.
Filtering of WES data revealed three candidate variants in the SPAST gene in four families (Supplementary Table S3): variants c.C782A:p. Ser261* in SPG4-A, c.806dupA:p. Tyr269_Ser270delins* in SPG4-C, and c.G1496A:p. Arg499His in SPG4-D and SPG4-E (Table 1). These variants were screened in the family members and co-segregated with the disease status (Figure 1). The c.G1496A:p. Arg499His variant was predicted as a deleterious/damaging variant using 9/11 in silico software. It had a CADD_phred score equal to 34 and was predicted as a "likely pathogenic" variant by ACMG criteria. The c.C782A:p. Ser261* variant was anticipated as "pathogenic" in the InterVar (ACMG criteria). Its CADD_phred score was 37. It was predicted as a "damaging" variant by Mutation Taster, FATHMM, DANN, fitCons and Eigen in silico software. C.806dupA:p. Tyr269_Ser270delins* was also predicted as the "pathogenic" by ACMG criteria.
Filtering of WES also detected three pathogenic/likely pathogenic candidate variants in the KIF5A; c.A758T:p. Lys253Met, ATL1; c.C715T:p. Arg239Cys, and MFN2; c.G380A:p. Gly127Asp genes among remaining HSP families (Supplementary Table S2) that were co segregated with the disease status.
Results of MLPA
MLPA revealed a heterozygous deletion of exon 17 in the SPAST gene only in family SPG4-B (Supplementary Figure S1). We could not determine the boundaries of this deletion.
There was no deletion/duplication in the ATL1 gene (both genes were in one probe mix).
Results of Anticipation in Iranian Families
In this study, we investigated 39 cases in five multi generation, multi affected SPG4 families for anticipation. The averages of AAO in different generations showed the progressive decline in AAO in later generations of all pedigrees (Table 2). In family SPG4-A, the mean of AAO decreased from 47 to 23 years within two generations (individuals I2, II3, and II4 died before presenting their symptoms and II2 and III11 were asymptomatic in ages 75 and 55 years, respectively; so clearly, the mean AAO in the second generation was more than 61 years), while in family SPG4-B, the mean of AAO declined from ∼60 to 2.5 years during four generations. Also, in families SPG4-C and SPG4-D, the averages of AAO diminished from ∼60 to 27 and ∼50 to 3 years within three generations, respectively. In the last family, SPG4-E, the mean of AAO decreased from 68 to 6.8 years within two generations. These differences were significant between parents and their children in all five families (p-value <0.05). But the families with mutations in other genes or genetically undiagnosed families in this study did not present any significant anticipation.
Discussion
In this study, WES detected the disease-causing variants in seven out of 14 HSP families (50%). Clinical evaluation of them showed that only all SPG4 families presented a putative genetic anticipation. Among the seven remaining families, anticipation was observed only in the family SPG4-B who presented the large deletion of exon 17 of SPAST by MLPA as we expected. The deletion has been previously reported in several SPG4 families and suggested that the high concentration of Alu elements in the intronic and flanking regions of the SPAST gene likely facilitates the occurrence of intragenic Alu-mediated rearrangements especially in the final exon, exon 17. Reference Boone, Liu and Zhang37 Genetic anticipation was not observed in other subtypes of HSP with or without genetic diagnosis in our study (data have not been included).
Anticipation usually resulted from the dynamic mutations/repeat expansions. Reference Ranen, Stine and Abbott15–Reference Imbert, Saudou and Yvert17 Furthermore, it has also been reported occasionally in several diseases with static mutations. For instance, a few missense and splice-site mutations in the SOD1 gene, which causes ALS, already have been associated with anticipation. Reference Alavi, Nafissi and Rohani22,Reference Iwai, Yamamoto, Yoshihara and Sobue23,Reference Kim, Kim, Lee, Gwoun, Kim and Lee38 Or, anticipation has been considered with several mutations in the SPAST gene. Before discovering SPAST as a SPG4-causing gene, anticipation has been mentioned in a few families linked to SPAST locus, Reference Bruyn, Van Deutekom, Frants and Padberg39–Reference Thurmon, He, Haskell, Thorpe, Thurmon and Rosen43 but there was insufficient clinical evidence to support anticipation in SPG4 families. Bruyn et al. reported anticipation for the first time in a large Dutch family with clinically probable Strumpell’s disease with autosomal dominant inheritance. They observed that the AAO in offsprings declined in the successive generations from 45.8 to 9.6 years and suggested possible anticipation Reference Bruyn, Van Deutekom, Frants and Padberg39 (Supplementary Table S1). After identifying the SPAST gene, anticipation has been noted in the small number of families affected by SPG4 (Supplementary Table S1). Reference Bönsch, Schwindt and Navratil25–Reference Reddy, Seltzer and Grewal34,Reference Bruyn, Van Deutekom, Frants and Padberg39–Reference Thurmon, He, Haskell, Thorpe, Thurmon and Rosen43 All these families were diagnosed with static mutations in the SPAST gene. Neither nature of the mutation nor position of the alteration changed among generations. However, high variability in AAO in the SPG4 patients in different generations of the families suggests that additional genetic or environmental modifiers also contribute to their phenotypes. Reference Morita, Ho, Hosler, McKenna-Yasek and Brown28 Reddy et al. found a deletion mutation in SPAST, and they concluded it might be dynamic and variability in the size of the deletion could account for anticipation. Reference Reddy, Seltzer and Grewal34 Nevertheless, this hypothesis does not seem to be correct because this mutation did not alter transmission from one generation to another, and therefore, it cannot be dynamic. Also, anticipation was seen in some other SPG4 families with missense mutations. So, it seems that other molecular mechanisms or environmental factors may play roles in the development of anticipation in SPG4 families. Reference Reddy, Seltzer and Grewal34 Considering the same living environment for different generations, it looks unlikely that environmental factors play an important role in anticipation in these families. Reference Solowska and Baas11 Potentially, unknown intracellular factors like telomere position effect (TPE) or miscellaneous modifiers like sex and hormones or additional genetic modifiers are involved in such cases. Reference Parodi, Fenu and Barbier8 It is also suggested that some genes such as SPAST contain a few microsatellites in introns that are often highly mutable, and their copy numbers vary in different generations. These copy number variations may affect gene expression and thus AAO of disease. Reference Reddy, Seltzer and Grewal34 If this hypothesis is correct, the lack of anticipation in some families could be justified. However, the correctness of this hypothesis needs to be proven.
Although many SPG4 families have been reported worldwide, in many of them, anticipation was not reported, which may be due to (1) the lack of anticipation may be due to the above hypothesis, (2) the lack of attention (Supplementary Table S4), Reference Mitne-Neto, Kok and Beetz44–Reference Nielsen, Johnsen and Koefoed47 (3) the cohort studies to detect genotype–phenotype correlation have not investigated anticipation, Reference Kadnikova, Rudenskaya, Stepanova, Sermyagina and Ryzhkova7,Reference Chelban, Tucci and Lynch13 (4) the small number of affected individuals in these families and the lack of access to several generations of patients Reference Basri, Yabe and Soma48 or (5) the lack of access to clinical data of the earlier generations. Regarding the last one, the determination of AAO of the older affected individuals from the first affected generations in each family can be associated with recollection/ascertainment bias. Reference Minikel, Zerr and Collins49 On the other hand, the presence of disease may not be recognized at all when there is very late onset and possibility of misdiagnosis with primary lateral sclerosis (PLS). Reference Almomen, Martens and Quadir50 These limitations can affect the prediction of anticipation. Thus, its confirmation needs to focus on a large number of SPG4 families or maybe apply of objective rating scales such as Spastic Paraplegia Rating Scale (SPRS) Reference Schüle, Holland-Letz and Klimpe51 and Expanded Disability Status Scale (EDSS) Reference Guthrie, Pfeffer and Bailie52 or even designing of a novel rating scales to better compare phenotypes between generations. All in all, the existence of anticipation in SPG4 has remained controversial; however, it seems that anticipation can be relatively more common in SPG4 than it appears (Supplementary Table S4).
The literature search also showed anticipation might rarely be observed in some other subtypes of AD-HSP, including a few SPG3A (with mutations in ATL1) Reference Ming53 and SPG31 (with mutations in REEP1) Reference Kamada, Kawarai and Miyamoto54 families. Although anticipation may be observed in other subtypes of HSP, it seems SPAST will be the first candidate gene in families who present autosomal dominant inheritance, the pure form of disease as well as anticipation. This seems to be a powerful genotype-phenotype correlation.
Altogether, further research about SPAST mutations in the affected families should be validated experimentally for possible anticipation and finding its causes. Confirmation of anticipation can be valuable for genetic counseling in presymptomatic carriers and eventually their follow-up. Furthermore, providing more information about the underlying mechanisms has resulted in a less obscure understanding of the pathophysiology of SPG4. Reference Parodi, Fenu and Barbier8,Reference Solowska and Baas11
Clinical manifestations of SPG4 families in this study revealed intra- and interfamilial variability with differences in AAO and disease severity: intellectual disability was observed only in two affected individuals of the family SPG4-D but not in other affected individuals of the family or the patients of the SPG4-E family with the same mutation. These results confirm intra and inter-familial heterogeneity among SPG4 families (Table 1). Despite these heterogeneities, interestingly, the AAO between affected individuals in a given generation especially between siblings was less variable than between different generations (Table 2). Parodi et al. reported such correlation and suggested that other genetic factors may significantly impact the AAO. Reference Parodi, Fenu and Barbier8
Our results also showed another genotype–phenotype correlation: missense mutations seem to be associated with a more severe form of the disease. Intellectual disability, dysarthria, pes cavus, and epilepsy were observed only in two families with a missense mutation in the AAA ATPase cassette, and none of the SPG4 patients with truncating mutations developed these symptoms. Although these manifestations are not common in SPG4 patients, those have been previously reported in other SPG4 families. Reference de Bot, van den Elzen and Mensenkamp55,Reference Lu, Li and Dong56 Similar results were reported by Parodi et al. in 2018. They reported an increasing severity in patients harboring a missense mutation and mentioned that intellectual disability was significantly more frequent among SPG4 patients with missense mutations. Reference Parodi, Fenu and Barbier8 It is suggested that these interfamilial phenotypic variabilities among SPG4 patients might be due to type and location of mutations. Missense mutations that are often located in AAA-ATPase domain (such as p. Arg499His) result in weaken the microtubule severing activity of the protein or sometimes lack of affinity to ATP, suggesting a loss-of-function mechanism. Reference Bürger, Fonknechten and Hoeltzenbein57,Reference Evans, Gomes, Reisenweber, Gundersen and Lauring58 Whereas missense mutations in other domains of spastin may alter the endosomal tubule fusion function. Reference Allison, Edgar and Reid59,Reference Boutry, Morais and Stevanin60 or may increase the lifespan of the mutant protein, leading to toxic accumulation and interaction with the wild-type SPAST proteins, highlighting a negative dominant mechanism. Reference Al Mutairi, Alfadhel and Nashabat61 Nonsense mutations (such as p. Ser261*), resulting in truncated protein, can trigger nonsense-mediated decay machinery to cleanse the cell from such abnormalities, but given the rapid pileup, slight increase in stability, and the less susceptibility to degradation, neurotoxicity of truncated SPAST subunits results in defective outgrowth of neurites. Reference Bürger, Fonknechten and Hoeltzenbein57,Reference Qiang, Piermarini and Muralidharan62,Reference Solowska, Rao and Bass63 This phenomenon can be a gain-of-function that occurs in even very low doses of truncated spastin protein, which plays an essential role in axonal transportation. It has been suggested that this accumulation is progressive in nature, which is in concert with that progressive nature of SPG4, as well. Reference Solowska, Rao and Bass63
Briefly, although we observed a presumptive anticipation among SPG4 families, several researchers have suggested the earlier AAO might be linked to a potential ascertainment bias. Reference Chelban, Tucci and Lynch13,Reference Reddy, Seltzer and Grewal34 Due to the lack of knowledge and attention of these families regarding this disease, diagnostic sensitivity might have been enhanced in subsequent generations. Also, the lack of access to details of clinical data and assessment of the precise AAO, especially in the earlier generations, and the limited number of studied families can affect the prediction of anticipation. However, considering the observation of anticipation in a large number of previous reported SPG4 patients (Supplementary Table S1), as well as a significant difference in the AAO of the disease (>50 years), cannot be accidental.
Acknowledgements
The authors acknowledge the Iran National Institute for Medical Research Development (NIMAD; 963846) and the Genetic Research Center of the University of Social Welfare and Rehabilitation Sciences for funding the research and thank the patients and their family members for participating in the study.
Funding
The National Institute for Medical Research Development (NIMAD; grant number 963846) and the Genetic Research Center of the University of Social Welfare and Rehabilitation Sciences have granted this project.
Conflicts of Interest
All authors claim absence of financial interests and absence of conflicts of interest. All authors read and approved the final version of the manuscript.
Statement of Authorship
-
1) Research project: A. Conception, B. Organization, C. Execution;
-
2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique;
-
3) Manuscript: A. Writing of the first draft, B. Review and Critique.
Seyyed-Saleh Hashemi: 1C, 2B, and 3B
Reza Hajati: 1C, 2B, and 3B
Atefeh Davarzani: 1C, 2B, and 3B
Mohammad Rohani: 1A, 3A, and 3B
Mohammad Masoud Rahimi Bidgoli: 1C, 2B, 3B
Fardad DanaeeFard: 1C, 2B, and 3B
Farzad Fatehi: 1A, 3B
Ariana Kariminejad: 1A, 3B
Hossein Najmabadi: 1A, 3B
Shahriar Nafissi: 1A, 3B
Afagh Alavi: 1A, 1B, 2A, 2B, 3A, and 3B
Supplementary Material
To view supplementary material for this article, please visit https://doi.org/10.1017/cjn.2021.188
