Letter to the Editor | Published:

A canine chromosome 7 locus confers compulsive disorder susceptibility

Molecular Psychiatry volume 15, pages 810 (2010) | Download Citation


Human and canine compulsive disorders (CCDs) manifest as time-consuming, repetitive behaviors causing distress and functional impairment. CCDs derive from normal behaviors such as grooming (acral lick dermatitis), predatory behavior (tail chasing, fly snapping), eating/suckling (pica and flank sucking (FS)/blanket sucking (BS)) or locomotion (pacing/circling). They commonly appear between pre-pubescence and early social maturity1 and may be precipitated by anxiety or stress. However, the behavior soon becomes ‘fixed’ and is shown even in the absence of obvious stressors. CCD has compelling parallels with human obsessive–compulsive disorder (OCD).2, 3 OCD, characterized by recurring obsessions and/or compulsions, affects millions of individuals.4 Here, we describe the novel association of Doberman pincher CCD with a chromosome 7 locus containing CDH2, an attractive candidate.

Genome-wide association analyses were performed using DNA from 92 rigorously phenotyped Doberman pinscher FS/BS cases and 68 controls5 obtained under the Institute of Animal Care Use Committee protocol consent. The median onset age was 4.8 months for BS and 8.5 months for FS.6 DNA was genotyped on the Affymetrix v2 canine single-nucleotide polymorphism (SNP) array.5 After quality filtering, genome-wide association of the resulting 14 700 SNPs to CCD was performed using PLINK. Only three chromosome 7 SNPs (between 61.83 to 63.87 Mb) had P<0.1 after permutation testing; the most significant SNP was at 63.867472 Mb (Praw=7.6 × 10−7 and Pgenome=0.013, Figure 1a,b). Weaker associations on chromosomes 18, 33, 34 and 38 failed to reach genome-wide significance.

Figure 1
Figure 1

Elevated S100B serum concentrations (a) and an increased C-peptide/glucose ratio (b), indicating insulin resistance, were schizophrenia related. Increased BMI was primarily a consequence of antipsychotic medication and not a source of elevated S100B (c). Data are given as means with 95% confidence intervals. Only P-values of significant group differences are displayed.

Fine-mapping using 84 SNPs across 1.7 Mb surrounding the chromosome 7 peak, including nine SNPs from the genome-wide analysis, was performed on an increased sample number (94 affected; 73 controls). After analyzing 63 SNPs passing quality filters, the association peak remained at 63.867472 Mb (Praw=5.5 × 10−6). Multi-SNP haplotype testing revealed an extended region of moderate association spanning roughly 400 kb (63.8–64.2 Mb), likely reflecting the long haplotypes found within dog breeds. The most significantly associated SNP is located within CDH2 (Figure 1c). CDH2 is widely expressed, mediating synaptic activity-regulated neuronal adhesion.7 Dogs showing multiple compulsive behaviors have a higher frequency of the risk allele than dogs with a less severe phenotype (60 and 43%, respectively, compared with 22% in unaffected dogs; Figure 1d).

The highly significant association of CCD with the CDH2 region on chromosome 7 is the first genetic locus identified for any animal compulsive disorder, and raises the intriguing possibility that CDH2 and other neuronal adhesion proteins are involved in human compulsive behaviors. A genetic association of cadherins with autism spectrum disorder, which often includes repetitive and compulsive behaviors, has also recently been reported.8 As little is known about the underlying molecular mechanisms for compulsive behaviors, this discovery could provide a better understanding of disease biology and facilitate development of genetic tests, enabling earlier interventions and even treatment or prevention of compulsive disorders in at-risk canines and humans.

Supporting material can be found online (Supplementary Material).

Conflict of interest

The authors declare no conflict of interest.


  1. 1.

    , . JAVMA 2002; 221: 1445–1452.

  2. 2.

    , , , , . In: Hollander E, Stein DJ (eds). Obsessive Compulsive Disorders. Marcel Dekker, Inc: New York, 1997, pp 99–141.

  3. 3.

    , . JAVMA 1998; 212: 1252–1257.

  4. 4.

    , , , . Arch Gen Psychiatry 1988; 45: 1094–1099.

  5. 5.

    , , . JAVMA 2007; 231: 907–912.

  6. 6.

    , . Nat Rev Genet 2008; 9: 713–725.

  7. 7.

    , , , , , et al. Neuron 2000; 25: 93–107.

  8. 8.

    , , , , , et al. Nature 2009; 459: 528–533.

Download references

Author information

Author notes

    • E K Karlsson
    •  & A Moon-Fanelli

    These authors contributed equally to this work.


  1. Cummings School of Veterinary Medicine, Tufts University, North Grafton, MA, USA

    • N H Dodman
    •  & A Moon-Fanelli
  2. Broad Institute of MIT and Harvard, Cambridge, MA, USA

    • E K Karlsson
    • , M Perloski
    •  & K Lindblad-Toh
  3. FAS Center for Systems Biology, Harvard University, Cambridge, MA, USA

    • E K Karlsson
  4. University of Massachusetts Medical School, Worcester, MA, USA

    • M Galdzicka
    •  & E I Ginns
  5. Tufts University School of Medicine, Boston, MA, USA

    • L Shuster
  6. Department of Medical Biochemistry and Microbiology, Uppsala University, Uppsala, Sweden

    • K Lindblad-Toh


  1. Search for N H Dodman in:

  2. Search for E K Karlsson in:

  3. Search for A Moon-Fanelli in:

  4. Search for M Galdzicka in:

  5. Search for M Perloski in:

  6. Search for L Shuster in:

  7. Search for K Lindblad-Toh in:

  8. Search for E I Ginns in:

Corresponding author

Correspondence to N H Dodman.

Supplementary information

Word documents

  1. 1.

    Supplementary Material

About this article

Publication history




Supplementary Information accompanies the paper on the Molecular Psychiatry website (http://www.nature.com/mp)

Further reading