Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Prevalence, comorbidity, and breed differences in canine anxiety in 13,700 Finnish pet dogs


Behaviour problems and anxieties in dogs decrease their quality of life and may lead to relinquishment or euthanasia. Considering the large number of pet dogs and the commonness of these problematic behaviours, a better understanding of the epidemiology and related molecular and environmental factors is needed. We have here studied the prevalence, comorbidity, and breed specificity of seven canine anxiety-like traits: noise sensitivity, fearfulness, fear of surfaces and heights, inattention/impulsivity, compulsion, separation related behaviour and aggression with an online behaviour questionnaire answered by dog owners. Our results show that noise sensitivity is the most common anxiety-related trait with a prevalence of 32% in 13,700 Finnish pet dogs. Due to the high prevalence of noise sensitivity and fear, they were the most common comorbidities. However, when comparing the relative risk, the largest risk ratios were seen between hyperactivity/inattention, separation related behaviour and compulsion, and between fear and aggression. Furthermore, dog breeds showed large differences in prevalence of all anxiety-related traits, suggesting a strong genetic contribution. As a result, selective breeding focusing on behaviour may reduce the prevalence of canine anxieties. Anxious animals may suffer from chronic stress and thus, modified breeding policies could improve the welfare of our companion dogs.


Problematic behaviours can be a threat to dog welfare. Anxious dogs may be more vulnerable to diseases and show decreased lifespan1. Satisfaction with the dog’s behaviour may increase attachment to the dog2 and problematic behaviours, especially aggressiveness, destructiveness, fearfulness and hyperactivity are a common reason for relinquishment to shelters3,4. Problematic behaviours can even lead to euthanasia at a young age5,6. Behaviour problems, especially aggressiveness, may be public health concerns7. Some of these behaviour problems have been suggested to be analogous, or possibly even homologous to human anxiety disorders8, and the study of these spontaneous behaviour problems arising in a shared environment with people may reveal important biological factors underlying many psychiatric conditions. For example, canine compulsive disorder resembles human OCD on both phenotypic and neurochemical level8.

Behaviour problems are common in our companion canines9. The most common reported behaviour problems include excessive barking, inappropriate elimination, destructiveness, aggression and fearfulness6,10,11,12,13. The prevalence of noise sensitivity in previous studies has varied between 20% and 50%14,15,16,17,18. Around 20–25% of dogs show fearfulness of strangers, dogs or situations15,16 and separation anxiety occurs in 14–20% of dogs15,16,19. Furthermore, comorbidity between noise sensitivity and separation anxiety has also been observed15,20.

Behaviour has a major genetic component21,22,23,24,25,26,27 and many traits are both phenotypically and genetically correlated22. For example, relatives of compulsive dogs are often also affected28. Some genomic areas and loci are associated with problematic behaviour, including compulsion29, fear and noise sensitivity30. Furthermore, problematic behaviour may be influenced by many environmental factors, including, for example, maternal care, owner experience, training and exercise31,32,33. Behaviours are complex traits affected by several genes with small effects, multiple environmental factors varying in effect, and intricate interactions between them34.

Many epidemiological cross-sectional studies of behaviour are hindered by a small sample size and breed coverage or limited number of traits for a comprehensive overview. For example, our earlier anxiety survey included only three traits and reached ~3,300 participants16. Therefore, we developed here a more comprehensive owner-completed online questionnaire to collect a large data set of companion dogs in the study of canine anxiety. We expanded the study from three to seven anxiety-related traits and acquired four times as many records to assess the prevalence of different anxiety-like behaviour problems in a large home living population of companion canines, comorbidity between the anxiety-related traits, and trait-specific behavioural variation between breeds, which could indicate genetic inheritance.



We examined the epidemiology of seven anxiety-like traits in dogs: noise sensitivity, fear, fear of surfaces, inattention/impulsivity, compulsive behaviour, aggression and separation related behaviour with a comprehensive owner-answered online questionnaire. We collected 13,715 responses in 264 dog breeds. In total, 51.5% of the dogs were female and the age of the dogs varied between 10 weeks and 17 years 10 months (mean 4.7 years). We received more than 200 responses from mixed breed dogs and from 14 breeds: Bernese Mountain Dog, Border Collie, Finnish Lapponian Dog, German Shepherd Dog, Labrador Retriever, Lagotto Romagnolo, Lapponian Herder, Miniature Schnauzer, Rough Collie, Shetland Sheepdog, Smooth Collie, Soft-Coated Wheaten Terrier (labelled Wheaten Terrier), Spanish Water Dog and Staffordshire Bull Terrier (labelled Staff. Bull Terrier). These breeds and mixed breed dogs made up 35% of all dogs in the data. The sex ratio did not significantly differ between these breeds (χ2 = 4.17, DF = 14, P = 0.994), but the mean age differed (Kruskal-Wallis χ2 = 77.53, DF = 14, P < 0.001): Bernese Mountain Dogs (\(\bar{{\rm{x}}}\)  = 3.8 years, SD = 2.7) were, on average, younger and Miniature Schnauzers (\(\bar{{\rm{x}}}\) = 5.3 years, SD = 3.4) and Wheaten Terriers (\(\bar{{\rm{x}}}\) = 5.7 years, SD = 3.7) older than other breeds.


In total, 72.5% of dogs had some kind of highly problematic behaviour. Noise sensitivity was the most common anxiety trait with 32% of dogs being highly fearful of at least one noise (Fig. 1a, Supplementary Table S1). Fear was the second most common trait with a prevalence of 29%. Separation related behaviour and aggression were the most uncommon traits with prevalences of 5% and 14%, respectively.

Figure 1
figure 1

Prevalence of the traits (a) and subtraits (b) in a sample of 13715 dogs in 264 breeds.

The prevalence of the subtraits varied (Fig. 1b, Supplementary Table S1). In noise sensitivity, fear of fireworks was the most common subtrait with a prevalence of 26%. When comparing subtraits of fear, fear of other dogs was the most common. Aggression toward human family members was slightly more common than aggression toward strangers. From compulsion subtraits, self-biting was the most commonly reported.


Many dogs exhibited comorbidities between different anxiety-related traits (Fig. 2a). Most common comorbidity was fear, especially in aggressive and hyperactive/impulsive dogs, and the second most common was noise sensitivity, especially in fearful dogs. However, a somewhat different pattern emerged when comparing the relative risks (Fig. 2b, Supplementary Table S2). Dogs displaying separation related behaviour were 4.1 times more often hyperactive/impulsive and 3.4 times more often inattentive than dogs not displaying separation related behaviour. Similar comorbidities were seen between compulsion and hyperactivity/inattention, and compulsion and separation related behaviour. Finally, aggressive dogs were 3.2 times more often fearful, and dogs showing separation related behaviour were 2.8 times more likely fearful.

Figure 2
figure 2

Comorbidity heat maps. Proportion of dogs with pairs of diagnoses using the disorder listed in the column header as the denominator (a) and relative risk (b). (a) Adapted from Goldstein-Piekarski et al.72.

Comorbidity between subtraits was also common but showed high variability (Supplementary Table S3). Of the highly noise sensitive dogs, 53% displayed noise sensitivity toward more than one target. Similarly, 38% of the highly fearful dogs were highly fearful toward more than one subtrait. However, only 9% of aggressive dogs showed aggression toward both family members and strangers.

Age and sex differences

Male dogs were more often aggressive and hyperactive/impulsive, but female dogs were more often fearful (Fig. 3b–d, Supplementary Fig. S1, Supplementary Table S4). Separation related behaviour was slightly more common in male dogs. The prevalence of noise sensitivity increased with age, especially fear of thunder (Fig. 3a, Supplementary Fig. S1, Supplementary Table S5). Similarly, fear of surfaces and heights increased with age, whereas hyperactivity/impulsivity and tail chasing decreased. Other traits did not show clear linear changes.

Figure 3
figure 3

Prevalences of fear of thunder (a), aggression toward family members (b), hyperactivity/impulsivity (c) and fear of strangers (d) for both sexes and six age groups.

Breed specificity

Large breed differences were observed in all behavioural traits (Fig. 4, Supplementary Fig. S2, Supplementary Table S6). For example, 10.6% of Miniature Schnauzers were aggressive toward strangers, whereas only 0.4% of Labrador Retrievers showed aggression (Fig. 4). Similarly, 9.5% of Staffordshire Bull Terriers were reported to display tail chasing, but none of the Lagotto Romagnolo dogs chased their tails.

Figure 4
figure 4

Breed differences in fear of thunder (a), fear of strangers (b), fear of surfaces and heights (c), hyperactivity/impulsivity (d), inattention (e), aggression toward strangers (f), tail chasing (g), fly snapping/light chasing (h) and vocalization/salivation/panting alone (i). For other breed-wise differences, see Supplementary Fig. S2.

When focusing on individual breeds, these differences in prevalences resulted in breed-specific patterns (Fig. 5). Border Collies displayed a very high prevalence of compulsive staring and fly snapping, but only moderate levels of other traits (Fig. 5). In contrast, Miniature Schnauzers displayed high levels of aggression and social fear (fear of strangers and other dogs), but few stereotypies. Lagotto Romagnolos displayed high levels of noise sensitivity, social fear, and aggression. Staffordshire Bull Terriers showed high levels of compulsive behaviour, hyperactivity, and inattention. For more breeds, see Supplementary Fig. S3.

Figure 5
figure 5

Radar chart representation of the behaviour of dog breeds: Border Collie (a), Miniature Schnauzer (b), Lagotto Romagnolo (c) and Staffordshire Bull Terrier (d). Colors represent the larger traits. Clockwise from top: blue – noise sensitivity, lime green – fear, violet – fear of surfaces and heights, orange – aggression, pine green – hyperactivity/inattention, purple – separation related behaviour, yellow – compulsive behaviour. Radar charts for other breeds in Supplementary Fig. S3. The minimum in all traits is the breed-wise minimum prevalence and the maximum in all traits is the breed-wise maximum prevalence. For the minimum and maximum prevalences, see Supplementary Table S7.


We have here developed a comprehensive behavioural survey and applied a citizen science approach to collect a large canine behavioural dataset of over 260 breeds of dogs. Our data replicates results from some earlier studies and the study also shows a number of novel insights, including prevalences, comorbidities and breed differences of traits not described before. These results improve the overall understanding of problematic behaviour.

Based on the results here and in previous studies, noise sensitivity stands out as the most common canine anxiety with a prevalence of 32% in this study. Earlier, the prevalence has varied between 20% and 50%9,12,14,15,16,17,18. Based on our study and previous studies14,17,18, the most common noise sensitivity is the fear of fireworks. Fear was the second most common canine anxiety, with a prevalence of 29%. Specifically, 17% of dogs showed fear of other dogs, 15% fear of strangers and 11% fear of novel situations. Prevalence of total fearfulness12,16 and prevalence of fear subtraits9,15,16,35 were quite similar in previous studies as well. Fear of surfaces and heights appears to be highly prevalent in our study population, as 23.5% of dog owners reported that their dogs were highly fearful of different surfaces and heights.

Based on our results, every fifth dog displays high levels of inattention and 15% high levels of hyperactivity/impulsivity. Excessive activity has been reported in 12% to 34% of dogs9,12. Compulsive behaviour patterns were observed in 16% of the dogs, agreeing with a previous study9. Based on our results and previous studies9,28, self-mutilation is the most common compulsive behaviour. Self-mutilation may be a compulsion, but it may also be caused by allergies, ectoparasites or other skin problems, possibly explaining the high prevalence of the subtrait.

The prevalence of aggression was 14%, with both aggression toward human family members and toward strangers occurring in 6% of dogs. In our previous study, 16% of dogs had at least once displayed aggressive behaviours toward family members and 45% toward strangers16. In previous studies, the prevalence of aggression and its subtraits has varied between 2% and 30%9,12,15. Studies focusing on referrals to veterinary/behaviour clinics often report aggression as the most common behaviour problem10,11,13, possibly because owners find aggressiveness more problematic than, for example, fear of fireworks. Separation related behaviours were only displayed by 6% of the dogs. Previously, the prevalence of separation anxiety has been 2–3 times higher than in this present study9,15,16,19, possibly because we only included dogs with high frequencies of separation related behaviour. Taken together, our results are surprisingly similar compared to previous studies, even though the populations studied and the criteria for anxiety-related traits differ in every study.

Male and female dogs displayed differences in the prevalence of behaviour problems. Male dogs had a higher prevalence of aggressiveness, separation related behaviour, inattention and hyperactivity/impulsivity. In contrast, female dogs had a higher prevalence of fearfulness. Noise sensitivity, fear of surfaces and compulsive behaviour occurred independent of sex. Previous studies have shown similar sex differences in aggression9,13,36,37,38, fearfulness13,32,39,40,41, separation related behaviour36 and compulsion33. Reported sex differences in noise sensitivity are less clear (more common in females17,32, more common in males14, no difference between sexes18).

We detected differences between age groups in the prevalence of most behaviour problems. Younger dogs had a higher prevalence of destroy/urinate when alone, inattention, hyperactivity/impulsivity, tail chasing and self-biting. Older dogs had a higher prevalence of aggression, noise sensitivity and fear of surfaces. Fearfulness was most common in dogs aged 4–8 years. Furthermore, vocalize/salivate/pant when alone and other compulsive behaviours (beside the aforementioned tail chasing and self-biting) occurred independent of age. Previous studies are in agreement at least in noise sensitivity14,17,18,32, inattention42,43, hyperactivity/impulsivity12,42,43,44 and aggression12,37. Tail chasing, destroying and urinating indoors are typical behaviours for puppies, and this likely explains the age differences in these subtraits.

We observed large behavioural differences between breeds. Noise sensitivity was the most common in Lagotto Romagnolo, Wheaten Terrier and mixed breed dogs. Previous studies have also ranked these breeds high in noise sensitivity14,16,17,18. Among some other breeds, Miniature Schnauzers and Staffordshire Bull Terriers were, in contrast, noise sensitive less often, as also observed in previous studies10,17. Fear was most common in Spanish Water Dogs, Shetland Sheepdogs and mixed breeds. In contrast, Labrador Retrievers were seldomly fearful. These results are in agreement with previous studies, ranking mixed breed dogs high in fearfulness9,13,35,40 and Labrador Retrievers and Staffordshire Bull Terriers low in fearfulness35,41,45,46. Fear of surfaces and heights was the most often observed in Rough Collie and mixed breed dogs.

Inattention was most often reported in mixed breed dogs, Finnish Lapponian Dogs and Lapponian Herders, and rarely reported in Spanish Water Dogs and Border Collies. Although Finnish Lapponian Dogs and Lapponian Herders have not been studied before, our results agree with previous results43. As the dog breeds showing high prevalence of inattention are breeds that are often regarded as “hard to train”, owners may rate dogs not easily motivated by food or petting as inattentive. Hyperactivity and impulsivity were the most common in mixed breed dogs, German Shepherds, Spanish Water Dogs and Staffordshire Bull Terriers, and the least common in Rough Collies and Miniature Schnauzers. Similar breed differences were observed in a previous study44. Furthermore, herding dogs (including, for example, Border Collie and German Shepherd) and terriers (including, for example, Staffordshire Bull Terriers) have ranked high in extraversion and, in contrast, toy dogs low in extraversion47. In our study, Labrador Retrievers and Rough Collies had a low prevalence of hyperactivity/impulsivity, but otherwise our results match previous studies. However, it seems that classification into traditional and genetic breed groups poorly reflect behavioural differences48, possibly explaining the differences in these results. Compulsive behaviour was most often reported by owners of German Shepherds, mixed breed dogs and Staffordshire Bull Terriers. However, the breed differences varied highly between different compulsions. For example, Staffordshire Bull Terriers had a high prevalence of tail chasing, with nearly 10% of them chasing their tails. In contrast, light chasing and staring were the most often observed in Border Collies. Interestingly, Border Collies were bred to herd livestock by staring at them intensely, and even though this method of herding can be perfected by training, the behaviour itself seems to be innate49. Pacing and excessive drinking were often performed by mixed breed dogs and German Shepherds. In a previous study, German shepherds had high odds of being presented to a behaviour clinic for obsessive behaviour10.

Mixed breed dogs and Miniature Schnauzers had the highest prevalence of aggression, whereas Labrador Retrievers had the lowest prevalence of aggression. Aggression toward strangers was most prevalent in Miniature Schnauzers, mixed breed dogs, German Shepherd Dogs and Spanish Water Dogs, and least prevalent in Labrador Retrievers. Aggression toward human family members was most common in Miniature Schnauzers and Lagotto Romagnolos. Our results agree with previous studies both in total aggression9,13 and in the subtraits, aggression toward strangers36,37,38,45 and aggression towards family members16,37. Separation related behaviour was most common in mixed breed dogs and Wheaten Terriers. Specifically, mixed breed dogs were likely to destroy, urinate or defecate when left alone, whereas Wheaten Terriers were likely to vocalize, salivate or pant. Based on our results and a previous study36, mixed breed dogs may be more prone to show separation related behaviour. It is possible that the high prevalence of separation distress and other anxieties in the mixed breed dogs is caused by a poor early life environment and adverse experiences in life, as many mixed breed dogs in our data are likely rescues.

Within-trait comorbidity was common in noise sensitivity and fear: 53% of dogs that were fearful of one noise were fearful of several noises, and 38% of fearful dogs were fearful of more than one target. This result was also discovered in our previous study16. Based on previous studies, noise sensitivity is often generalised and displayed toward several different noises9,14,17,18. We discovered that dogs were seldomly aggressive toward both family members and strangers, as reported before in some dog breeds50. In contrast, one previous study did report a significant comorbidity between stranger-directed and owner-directed aggression13. However, it seems that aggression toward strangers and family members are genetically distinct traits51.

We discovered that noise sensitivity and fear were the most common comorbidities, likely due to their high prevalence in our study population. However, when comparing the risk ratios in comorbid traits, the largest risk ratios were seen between separation related behaviour, hyperactivity/impulsivity, inattention and compulsive behaviour, and between fear and aggression. Fearful dogs were 3.2 times more often aggressive than non-fearful dogs, a relationship found in previous studies as well9,13,16,38. This indicates that aggression is commonly motivated by fear. The connection between impulsivity, compulsive behaviour and separation related behaviour is an interesting finding that demands further research. One previous study discovered that excitable dogs had 9.8 times higher odds of separation distress52 and another study discovered a connection between compulsive behaviour and hyperactivity9. Intriguingly, impulsivity and compulsion are related constructs, as both are proposed to be caused by a failure of response control and mediated by basal ganglia53. We observed many trait connections detected in previous studies as well, including comorbidity between fear and noise sensitivity14,16,17, between fear and separation related behaviour16, between separation related behaviour and aggression13,16 and between fear and compulsive behaviour9,33. However, previous studies have detected a comorbidity between separation anxiety and noise sensitivity9,13,15,17,19,20. We indeed discovered that separation related behaviour was 1.4 times more prevalent in noise sensitive dogs. However, the opposite was not true, as dogs showing separation related behaviour were not fearful of noises more often than dogs not showing separation related behaviour. Furthermore, we discovered a positive connection between compulsive behaviour and aggression, contrasting with the results of our previous study33.

This study has limitations. Although the fear section of the questionnaire was validated and the test-retest reliability of the fear and noise sensitivity sections was good54 and that the results we have obtained from the data collected with it32,33,55 replicate many previous results, the psychometric properties of the rest of the questionnaire have not been formally evaluated. Future studies should aim to assess the reliability and validity of these additional components. Secondly, the categorisation into low, moderate, and high categories was mostly based on the frequency of signs and not the severity, except in aggression, separation anxiety, and impulsivity/inattention. Thus, in the high groups, the severity of the symptoms can be variable. Thirdly, our sample is a self-selected convenience sample, and may not be representative of the overall Finnish dog population. Although the most common breeds in our sample are also common in Finland56, the representativeness of our sample is still unknown. We are currently working on a separate study to understand the participant profiles and details of the sample demographics.

Our findings on breed differences indicate that canine anxieties likely have a genetic basis. In previous studies, many behavioural traits have been indeed shown to have small to moderate heritabilities22,26,57 and recently we mapped two loci for generalized fear and noise sensitivity30. Therefore, it could be possible to decrease the prevalence of canine anxieties by selecting non-anxious animals for breeding. Our results also show that these canine anxieties are phenotypically correlated. Some of these traits, like many behaviour traits21,22, may also be genetically correlated, and therefore selection for one trait may influence other traits as well. Interestingly, a genomic region associated with noise sensitivity in German Shepherd Dogs30 contains the oxytocin receptor gene (OXTR). The gene is associated with social behaviour58 (but see a contrasting study with a smaller sample size59) and most likely has been under strong selection during domestication. This could explain the high prevalence of noise sensitivity in many study populations12,14,15,16,17,18 and could also indicate that breeding efforts to reduce the prevalence of noise sensitivity may prove difficult.

Based on our results, canine anxieties and behaviour problems are common across breeds. There are around 77 million dogs in the United States60 and 85 million in Europe61, and therefore these behaviour problems can affect millions of animals. As anxiety can impair welfare1 and problematic behaviour may be an indication of poor welfare62, efforts should be made to decrease the prevalence of these canine anxieties. Breeding policies may help to improve dog welfare, as could changes in the living environment14,19,32,33,37. Our ongoing efforts aim to identify environmental and genetic risk factors behind these canine anxiety-related traits using the large survey data collected here.



A comprehensive online questionnaire was designed to collect extensive information about behaviour and background information of dogs (Supplementary information). The questionnaire consisted of a background section and sections focusing on seven canine anxiety-related traits: noise sensitivity, fear, fear of surfaces and heights, impulsivity/inattention, compulsive behaviour, aggression and separation related behaviour. These traits consisted of several subtraits (Fig. 1b). The fear part of this questionnaire was previously shown to have good validity and both the fear and noise sensitivity sections had good test-retest reliability54. A questionnaire can be a good method for collecting data, since the reliability of questionnaires has been good in behavioural science and the answers are strongly linked to the behaviour of the animals54,63,64,65,66,67,68. Furthermore, as the owners have a long history with the dogs, it is possible to study traits that would be difficult to study using other methods (for example, behaviour test batteries), including compulsive behaviour69.


Owners were mainly recruited from a social media (Facebook) channel. Furthermore, dog breed organisations advertised the questionnaire on their own websites and Facebook pages. In total, we received questionnaire responses of 13,715 dogs in 264 breeds. To ensure reliable prevalence estimates in breed comparisons, we calculated a minimum sample size per breed with the Epitools sample size calculator70. For this calculation, we used the average prevalence of behaviour traits (19%), the average number of registrations per dog breed within the last 10 years (1500), a desired precision of 0.05 and a confidence level of 0.95. This calculation resulted in a minimum sample size of 200 individuals per breed. We obtained more than 200 answers from 14 breeds and mixed breed dogs (N = 418): Finnish Lapponian Dog (N = 538), Labrador Retriever (N = 465), German Shepherd Dog (N = 461), Shetland Sheepdog (N = 411), Soft-coated Wheaten Terrier (N = 351), Lapponian Herder (N = 294), Border Collie (N = 268), Miniature Schnauzer (N = 255), Spanish Water Dog (N = 251), Rough Collie (N = 248), Lagotto Romagnolo (N = 248), Bernese Mountain Dog (N = 209), Smooth Collie (N = 203) and Staffordshire Bull Terrier (N = 200).

Informed consent was obtained from all participants. Participants agreed that all questionnaire answers could be used for research. We emphasized that all data will be handled strictly confidentially, and that individual dogs and owners cannot be recognized from the published results.

Subject categorisation

Based on the questionnaire scores, dogs were categorised into low, moderate, and high groups for each subtrait (Fig. 1b, Supplementary Table S1) depending on the frequency of the behaviour. The subtraits were, in turn, combined to form trait groups. Low trait groups consisted of dogs that fell into low groups in all subtraits, and high trait groups consisted of dogs that had a high score in at least one subtrait. The categorisation for different subtraits is explained below.

Noise sensitivity

Noise sensitivity consisted of three subtraits: fear of thunder, fireworks, and gunshot. The respondents were first asked whether their dog showed fear toward these targets, and secondly, how often their dog shows fear, from rarely (0–20% of the time) to always (100% of the time). Dogs belonging to low group did not show fear toward these targets and the owners did not report a frequency for noise sensitivity. Dogs belonging to the high group showed fear at least often (40–60% of the time).


The fear section of the questionnaire consisted of three fear subtraits: fear of strangers, other dogs, and novel situations. Dogs that showed fear at least often (40–60% of the time) constituted the high group and dogs that never showed fear constituted the low group in all subtraits. Furthermore, low fear dogs never barked or growled at strangers or other dogs. Behaviour tests have previously been conducted to validate fear of strangers and fear of novel situations phenotypes54.

Fear of surfaces and heights

In this section, owners were asked whether their dog had difficulties walking on different surfaces, including on a metal grid, on shiny floors, or moving from one surface to another. Moreover, owners were asked whether their dog had difficulties in high places: climbing stairs where you can or cannot see between steps, walking next to glass railings, climbing metal stairs, and walking over narrow bridges. In all questions, these answers were scored between 0 (never) and 4 (always). Dogs were categorised as high fear of surfaces if at least one of the answers was 3 (often) or 4, and low group consisted of dogs that never had difficulties in any of these tasks.


Dog owners were asked to rate their dog’s behaviour on a scale of 1 (never) to 4 (very often) in 13 questions, developed by Vas and colleagues42 and designed to measure inattention, hyperactivity, and impulsivity. A principal component analysis with a promax rotation was run to divide the questions into two components, inattention and hyperactivity/impulsivity (Supplementary Table S8), and component scores were calculated for each dog. One question (Item 11) was removed from the analysis, as it loaded equally on both components. The cut-off between low and moderate category was set at the first dog having a rating of 3 (often) in the item that had the highest loading (Item 2 in inattention, Item 5 in hyperactivity/impulsivity). The cut-off between the moderate and high category was set at the first dog having a rating of 4 (very often) in the highest loading item and the same rating on one additional item.

Compulsive behaviour

Compulsive behaviour part of the questionnaire measured the occurrence of tail chasing, fly snapping/light chasing, surface licking, pacing, staring, excessive drinking, and self-biting. Tail chasing, fly snapping/light chasing, surface licking, pacing, and staring were scored on a scale of 0 (I’ve never noticed this behaviour) to 6 (several times per day). The dogs scoring from 4 (every other day-weekly) to 6 were placed in the high group, and dogs scoring 0 or 1 (a few times during the dog’s lifetime) were placed in the low group. Furthermore, owners were asked to estimate the time dog spent drinking or near the water bowl from less than 5 minutes to 1 hour or more. Dogs that spent less than 5 minutes near the water bowl formed the low group in excessive drinking, and dogs spending more than 15 minutes near it formed the high group. Finally, the prevalence of self-biting was rated from 0 (never) to 3 (several hours per day). Dogs scoring 2 (almost every day) or 3 constituted the high group whereas dogs scoring 0 constituted the low group.


The aggression trait in the questionnaire consisted of two subtraits: aggression toward strangers and toward family members. The respondents were asked to score the likelihood of their dog growling and trying to snap/bite from 1 (never) to 5 (always or almost always) when a stranger tries to pet the dog in its home or outside and when the owner handles the dog or tries to take a resource (food, bone, or toy) from the dog. If the dog tried to snap or bite at least sometimes (3) or it growled at least often (4), it was categorised as a high aggression dog in each subtrait. The dogs that never showed aggression in any of these situations constituted the low group.

Separation related behaviour

Separation related behaviour was divided into two parts: destroy/urinate alone and vocalize/salivate/pant alone. Dog owners were asked whether their dog exhibited separation anxiety and how often did the dog perform the aforementioned behaviours from 0 (never) to 4 (very often). High group dogs were reported to exhibit separation anxiety and performed at least one of these behaviours often (3) or very often. With the subtrait vocalize/salivate/pant, high group dogs were reported to salivate or pant at least often (3–4), or barked at least often (3–4) and salivated or panted at least rarely (1–4). Low group dogs did not show separation anxiety and never performed these behaviours.

Statistical analyses

To estimate the prevalence of high anxiety dogs, the number of dogs in high group was divided by the sum of dogs in all trait groups, hence obtaining the percentage of dogs in the high group for each subtrait and trait in the population.

The same prevalence calculation was performed individually for different breeds (sample size > 200) and for defined age groups and both sexes. The dogs were divided into six age groups: less than 2 years, 2–4 years, 4–6 years, 6–8 years, 8–10 years, and more than 10 years. Chi-squared tests were run in R71 for breed, sex, and age group differences in all traits and subtraits to reveal the significant associations.

Comorbidities were calculated between all pairs of behaviour problems. In the pairs comorbidity/diagnosis adapted from Goldstein-Piekarsi et al.72, the number of dogs belonging to the high group in both of the traits was divided by the number of all dogs in the high group in the denominator trait, hence obtaining the percentage of dogs in each denominator trait that also had the comorbid trait. The relative risk was calculated by firstly calculating the same percentage and secondly calculating the percentage of dogs that belonged to the low group in the denominator trait but in the high group in the other trait. Thirdly, the first percentage was divided by the second. Two-proportions z-tests were run in R71 for all risk ratios to see whether any two traits have statistically significant comorbidity.

All P-values were corrected for false discovery rate (FDR) to decrease the probability of type I error. The significance cut-off P-value was set at P < 0.05.

Data availability

The anonymised data is available as a supplementary datasheet.


  1. 1.

    Dreschel, N. A. The effects of fear and anxiety on health and lifespan in pet dogs. Appl. Anim. Behav. Sci. 125, 157–162 (2010).

    Article  Google Scholar 

  2. 2.

    Serpell, J. A. Evidence for association between pet behavior and owner attachement levels. Appl. Anim. Behav. Sci. 47, 49–60 (1996).

    Article  Google Scholar 

  3. 3.

    New, J. C. et al. Characteristics of Shelter-Relinquished Animals and Their Owners Compared With Animals and Their Owners in U.S. Pet-Owning Households. J. Appl. Anim. Welf. Sci. 3, 179–201 (2000).

    Article  Google Scholar 

  4. 4.

    Patronek, G. J., Glickman, L. T., Beck, A. M., McCabe, G. P. & Ecker, C. Risk factors for relinquishment of dogs to an animal shelter. J. Am. Vet. Med. Assoc. 209, 572–581 (1996).

    CAS  PubMed  PubMed Central  Google Scholar 

  5. 5.

    O’Neill, D. G., Church, D. B., McGreevy, P. D., Thomson, P. C. & Brodbelt, D. C. Longevity and mortality of owned dogs in England. Vet. J. 198, 638–643 (2013).

    Article  PubMed  PubMed Central  Google Scholar 

  6. 6.

    Fatjó, J., Ruiz-de-la-Torre, J. L. & Manteca, X. The epidemiology of behavioural problems in dogs and cats: a survey of veterinary practitioners. Anim. Welf. 15, 179–185 (2006).

    Google Scholar 

  7. 7.

    Gilchrist, J., Sacks, J. J., White, D. & Kresnow, M.-J. Dog bites: still a problem? Inj. Prev. 14, 296–301 (2008).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. 8.

    Overall, K. L. Natural animal models of human psychiatric conditions: assessment of mechanism and validity. Prog. Neuropsychopharmacol. Biol. Psychiatry 24, 727–776 (2000).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. 9.

    Dinwoodie, I. R., Dwyer, B., Zottola, V., Gleason, D. & Dodman, N. H. Demographics and comorbidity of behavior problems in dogs. J. Vet. Behav. 32, 62–71 (2019).

    Article  Google Scholar 

  10. 10.

    Col, R., Day, C. & Phillips, C. J. C. An epidemiological analysis of dog behavior problems presented to an Australian behavior clinic, with associated risk factors. J. Vet. Behav. 15, 1–11 (2016).

    Article  Google Scholar 

  11. 11.

    Hsu, Y. & Serpell, J. A. Development and validation of a questionnaire for measuring behavior and temperament traits in pet dogs. J. Am. Vet. Med. Assoc. 223, 1293–1300 (2003).

    Article  Google Scholar 

  12. 12.

    Khoshnegah, J., Azizzadeh, M. & Mahmoodi Gharaie, A. Risk factors for the development of behavior problems in a population of Iranian domestic dogs: Results of a pilot survey. Appl. Anim. Behav. Sci. 131, 123–130 (2011).

    Article  Google Scholar 

  13. 13.

    Bamberger, M. & Houpt, K. A. Signalment factors, comorbidity, and trends in behavior diagnoses in dogs: 1,644 cases (1991–2001). J. Am. Vet. Med. Assoc. 229, 1591–1601 (2006).

    Article  Google Scholar 

  14. 14.

    Blackwell, E. J., Bradshaw, J. W. S. & Casey, R. A. Fear responses to noises in domestic dogs: Prevalence, risk factors and co-occurrence with other fear related behaviour. Appl. Anim. Behav. Sci. 145, 15–25 (2013).

    Article  Google Scholar 

  15. 15.

    Martínez, Á. G., Santamarina Pernas, G., Diéguez Casalta, F. J., Suárez Rey, M. L. & De la Cruz Palomino, L. F. Risk factors associated with behavioral problems in dogs. J. Vet. Behav. Clin. Appl. Res. 6, 225–231 (2011).

    Article  Google Scholar 

  16. 16.

    Tiira, K., Sulkama, S. & Lohi, H. Prevalence, comorbidity, and behavioral variation in canine anxiety. J. Vet. Behav. Clin. Appl. Res. 16, 36–44 (2016).

    Article  Google Scholar 

  17. 17.

    Storengen, L. M. & Lingaas, F. Noise sensitivity in 17 dog breeds: Prevalence, breed risk and correlation with fear in other situations. Appl. Anim. Behav. Sci. 171, 152–160 (2015).

    Article  Google Scholar 

  18. 18.

    Blackwell, E., Casey, R. & Bradshaw, J. Firework fears and phobias in the domestic dog. RSPCA Rep. (2005).

  19. 19.

    Sherman, B. L. & Mills, D. S. Canine Anxieties and Phobias: An Update on Separation Anxiety and Noise Aversions. Vet. Clin. North Am. - Small Anim. Pract. 38, 1081–1106 (2008).

    Article  PubMed  PubMed Central  Google Scholar 

  20. 20.

    Overall, K. L., Dunham, A. E. & Frank, D. Frequency of nonspecific clinical signs in dogs with separation anxiety, thunderstorm phobia, and noise phobia, alone or in combination. J. Am. Vet. Med. Assoc. 219, 467–473 (2001).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. 21.

    Arvelius, P., Eken Asp, H., Fikse, W. F., Strandberg, E. & Nilsson, K. Genetic analysis of a temperament test as a tool to select against everyday life fearfulness in Rough Collie. J. Anim. Sci. 92, 4843–4855 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. 22.

    Ruefenacht, S., Gebhardt-Henrich, S., Miyake, T. & Gaillard, C. A behaviour test on German Shepherd dogs: Heritability of seven different traits. Appl. Anim. Behav. Sci. 79, 113–132 (2002).

    Article  Google Scholar 

  23. 23.

    Ilska, J. et al. Genetic Characterization of Dog Personality Traits. Genetics 206, 1101–1111 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  24. 24.

    Goddard, M. E. & Beilharz, R. G. Genetics of traits which determine the suitability of dogs as guide-dogs for the blind. Appl. Anim. Ethol. 9, 299–315 (1983).

    Article  Google Scholar 

  25. 25.

    Wilsson, E. & Sundgren, P.-E. Behaviour test for eight-week old puppies—heritabilities of tested behaviour traits and its correspondence to later behaviour. Appl. Anim. Behav. Sci. 58, 151–162 (1998).

    Article  Google Scholar 

  26. 26.

    MacLean, E. L. & Snyder-Mackler, N., vonHoldt, B. M. & Serpell, J. A. Highly heritable and functionally relevant breed differences in dog behaviour. Proc. R. Soc. B Biol. Sci. 286, 20190716 (2019).

  27. 27.

    Hradecká, L., Bartoš, L., Svobodová, I. & Sales, J. Heritability of behavioural traits in domestic dogs: A meta-analysis. Appl. Anim. Behav. Sci. 170, 1–13 (2015).

    Article  Google Scholar 

  28. 28.

    Overall, K. L. & Dunham, A. E. Clinical features and outcome in dogs and cats with obsessive-compulsive disorder: 126 cases (1989-2000). J. Am. Vet. Med. Assoc. 221, 1445–1452 (2002).

    Article  PubMed  PubMed Central  Google Scholar 

  29. 29.

    Dodman, N. H. et al. A canine chromosome 7 locus confers compulsive disorder susceptibility. Mol. Psychiatry 15, 8–10 (2010).

    Article  CAS  Google Scholar 

  30. 30.

    Sarviaho, R. et al. Two novel genomic regions associated with fearfulness in dogs overlap human neuropsychiatric loci. Transl. Psychiatry 9, 18 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. 31.

    Bennett, P. C. & Rohlf, V. I. Owner-companion dog interactions: Relationships between demographic variables, potentially problematic behaviours, training engagement and shared activities. Appl. Anim. Behav. Sci. 102, 65–84 (2007).

    Article  Google Scholar 

  32. 32.

    Tiira, K. & Lohi, H. Early life experiences and exercise associate with canine anxieties. PLoS One 10, e0141907 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. 33.

    Tiira, K. et al. Environmental Effects on Compulsive Tail Chasing in Dogs. PLoS One 7, e41684 (2012).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  34. 34.

    Overall, K. L., Tiira, K., Broach, D. & Bryant, D. Genetics and behavior: A guide for practitioners. Vet. Clin. North Am. - Small Anim. Pract. 44, 483–505 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  35. 35.

    Edwards, P. T. et al. Investigating risk factors that predict a dog’s fear during veterinary consultations. PLoS One 14, e0215416 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. 36.

    Takeuchi, Y., Ogata, N., Houpt, K. A. & Scarlett, J. M. Differences in background and outcome of three behavior problems of dogs. Appl. Anim. Behav. Sci. 70, 297–308 (2001).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. 37.

    Hsu, Y. & Sun, L. Factors associated with aggressive responses in pet dogs. Appl. Anim. Behav. Sci. 123, 108–123 (2010).

    Article  Google Scholar 

  38. 38.

    Flint, H. E., Coe, J. B., Serpell, J. A., Pearl, D. L. & Niel, L. Risk factors associated with stranger-directed aggression in domestic dogs. Appl. Anim. Behav. Sci. 197, 45–54 (2017).

    Article  Google Scholar 

  39. 39.

    Kubinyi, E., Turcsán, B. & Miklósi, Á. Dog and owner demographic characteristics and dog personality trait associations. Behav. Processes 81, 392–401 (2009).

    Article  PubMed  PubMed Central  Google Scholar 

  40. 40.

    Temesi, A., Turcsán, B. & Miklósi, Á. Measuring fear in dogs by questionnaires: An exploratory study toward a standardized inventory. Appl. Anim. Behav. Sci. 161, 121–130 (2014).

    Article  Google Scholar 

  41. 41.

    Starling, M. J., Branson, N., Thomson, P. C. & McGreevy, P. D. ‘Boldness’ in the domestic dog differs among breeds and breed groups. Behav. Processes 97, 53–62 (2013).

    Article  PubMed  PubMed Central  Google Scholar 

  42. 42.

    Vas, J., Topál, J., Péch, É. & Miklósi, Á. Measuring attention deficit and activity in dogs: A new application and validation of a human ADHD questionnaire. Appl. Anim. Behav. Sci. 103, 105–117 (2007).

    Article  Google Scholar 

  43. 43.

    Lit, L., Schweitzer, J. B., Iosif, A.-M. & Oberbauer, A. M. Owner reports of attention, activity, and impulsivity in dogs: a replication study. Behav. Brain Funct. 6, 1 (2010).

    Article  PubMed  PubMed Central  Google Scholar 

  44. 44.

    Wright, H. F., Mills, D. S. & Pollux, P. M. J. Development and Validation of a Psychometric Tool for Assessing Impulsivity in the Domestic Dog (Canis familiaris). Int. J. Comp. Psychol. 24, 210–225 (2011).

    Google Scholar 

  45. 45.

    Serpell, J. A. & Duffy, D. L. Aspects of Juvenile and Adolescent Environment Predict Aggression and Fear in 12-Month-Old Guide Dogs. Front. Vet. Sci. 3, 49 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  46. 46.

    Svartberg, K. Breed-typical behaviour in dogs—Historical remnants or recent constructs? Appl. Anim. Behav. Sci. 96, 293–313 (2006).

    Article  Google Scholar 

  47. 47.

    Ley, J. M., Bennett, P. C. & Coleman, G. J. A refinement and validation of the Monash Canine Personality Questionnaire (MCPQ). Appl. Anim. Behav. Sci. 116, 220–227 (2009).

    Article  Google Scholar 

  48. 48.

    Turcsán, B., Kubinyi, E. & Miklósi, Á. Trainability and boldness traits differ between dog breed clusters based on conventional breed categories and genetic relatedness. Appl. Anim. Behav. Sci. 132, 61–70 (2011).

    Article  Google Scholar 

  49. 49.

    McConnell, P. B. & Baylis, J. R. Interspecific Communication in Cooperative Herding: Acoustic and Visual Signals from Human Shepherds and Herding Dogs. Ethology 67, 302–328 (1985).

    Google Scholar 

  50. 50.

    Duffy, D. L., Hsu, Y. & Serpell, J. A. Breed differences in canine aggression. Appl. Anim. Behav. Sci. 114, 441–460 (2008).

    Article  Google Scholar 

  51. 51.

    Zapata, I., Serpell, J. A. & Alvarez, C. E. Genetic mapping of canine fear and aggression. BMC Genomics 17, 572 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  52. 52.

    McGreevy, P. D. & Masters, A. M. Risk factors for separation-related distress and feed-related aggression in dogs: Additional findings from a survey of Australian dog owners. Appl. Anim. Behav. Sci. 109, 320–328 (2008).

    Article  Google Scholar 

  53. 53.

    Dalley, J. W., Everitt, B. J. & Robbins, T. W. Impulsivity, Compulsivity, and Top-Down Cognitive Control. Neuron 69, 680–694 (2011).

    Article  CAS  Google Scholar 

  54. 54.

    Tiira, K. & Lohi, H. Reliability and validity of a questionnaire survey in canine anxiety research. Appl. Anim. Behav. Sci. 155, 82–92 (2014).

    Article  Google Scholar 

  55. 55.

    Puurunen, J. et al. Physiological and environmental factors associated with social fear in pet dogs. Manuscript submitted. (2020).

  56. 56.

    Finnish Kennel Club. Breeding database: Registrations. Available at:, (Accessed: 28th January 2020)

  57. 57.

    van der Waaij, E. H., Wilsson, E. & Strandberg, E. Genetic analysis of results of a Swedish behavior test on German Shepherd Dogs and Labrador Retrievers. J. Anim. Sci. 86, 2853–2861 (2008).

    Article  CAS  Google Scholar 

  58. 58.

    Kis, A. et al. Oxytocin Receptor Gene Polymorphisms Are Associated with Human Directed Social Behavior in Dogs (Canis familiaris). PLoS One 9, e83993 (2014).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  59. 59.

    Ottenheimer-Carrier, L., Ricketts, C. J., Perry, E. A., Anderson, R. E. & Walsh, C. J. Owner-reported personality assessments are associated with breed groups but not with oxytocin receptor gene polymorphisms in domestic dogs (Canis familiaris). J. Vet. Behav. 18, 62–68 (2017).

    Article  Google Scholar 

  60. 60.

    AVMA. 2017-2018 U.S. Pet Ownership & Demographics Sourcebook.

  61. 61.

    FEDIAF. Facts & Figures. (2018).

  62. 62.

    Sonntag, Q. & Overall, K. L. Key determinants of dog and cat welfare: behaviour, breeding and household lifestyle. Rev. Sci. Tech. 33, 213–220 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  63. 63.

    Feaver, J., Mendl, M. & Bateson, P. A method for rating the individual distinctiveness of domestic cats. Anim. Behav. 34, 1016–1025 (1986).

    Article  Google Scholar 

  64. 64.

    Momozawa, Y. et al. Assessment of equine temperament by a questionnaire survey to caretakers and evaluation of its reliability by simultaneous behavior test. Appl. Anim. Behav. Sci. 84, 127–138 (2003).

    Article  Google Scholar 

  65. 65.

    Duffy, D. L. & Serpell, J. A. Predictive validity of a method for evaluating temperament in young guide and service dogs. Appl. Anim. Behav. Sci. 138, 99–109 (2012).

    Article  Google Scholar 

  66. 66.

    Mirkó, E., Dóka, A. & Miklósi, Á. Association between subjective rating and behaviour coding and the role of experience in making video assessments on the personality of the domestic dog (Canis familiaris). Appl. Anim. Behav. Sci. 149, 45–54 (2013).

    Article  Google Scholar 

  67. 67.

    Wilsson, E. & Sinn, D. L. Are there differences between behavioral measurement methods? A comparison of the predictive validity of two ratings methods in a working dog program. Appl. Anim. Behav. Sci. 141, 158–172 (2012).

    Article  Google Scholar 

  68. 68.

    Gosling, S. D., Kwan, V. S. Y. & John, O. P. A dog’s got personality: a cross-species comparative approach to personality judgments in dogs and humans. J. Pers. Soc. Psychol. 85, 1161–1169 (2003).

    Article  Google Scholar 

  69. 69.

    Bennett, S. L., Litster, A., Weng, H. Y., Walker, S. L. & Luescher, A. U. Investigating behavior assessment instruments to predict aggression in dogs. Appl. Anim. Behav. Sci. 141, 139–148 (2012).

    Article  Google Scholar 

  70. 70.

    Ausvet. Epitools - Sample size to estimate a proportion or apparent prevalence with specified precision. Available at:, (Accessed: 8th January 2020)

  71. 71.

    R Core Team. R: A language and environment for statistical computing. (2016).

  72. 72.

    Goldstein-Piekarski, A. N., Williams, L. M. & Humphreys, K. A trans-diagnostic review of anxiety disorder comorbidity and the impact of multiple exclusion criteria on studying clinical outcomes in anxiety disorders. Transl. Psychiatry 6, e847 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references


We thank all the dog owners who participated in the study. This study was partially funded by the Academy of Finland (308887), the ERCStG (260997), ERA-NET NEURON and the Jane and Aatos Erkko Foundation. H.L. is a member of HiLIFE. M.S., S.M., E.H. & H.L. are members of the Helsinki One Health.

Author information




M.S., S.S. & H.L. designed the study. S.S., K.T., & H.L. developed the questionnaire. S.S. & H.L. piloted and edited the questionnaire and collected the data. M.S., S.S., & S.M. handled and processed the data with help from J.P. & E.H. using the scripts planned by M.S., S.S., H.L. and developed by C.A. M.S. performed the statistical analyses. Manuscript was planned by M.S., S.S. and H.L., drafted by M.S., and revised and accepted by all.

Corresponding author

Correspondence to Hannes Lohi.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Salonen, M., Sulkama, S., Mikkola, S. et al. Prevalence, comorbidity, and breed differences in canine anxiety in 13,700 Finnish pet dogs. Sci Rep 10, 2962 (2020).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing