The effect of a hiding box on stress levels and body weight in Dutch shelter cats; a randomized controlled trial

While staying in an animal shelter, cats may suffer from chronic stress which impairs their health and welfare. Providing opportunities to hide can significantly reduce behavioural stress in cats, but confirmation with physical parameters is needed. Therefore, the aim of this study was to determine the effect of a hiding box on behavioural stress levels (scored by means of the Cat-Stress-Score) and a physical parameter, namely body weight, during the first 12 days in quarantine for cats newly arrived cats at a Dutch animal shelter. Twenty-three cats between 1 and 10 years of age were randomly divided between the experimental (N = 12) and control group (N = 11) with and without a hiding box. Stress levels were assessed on days 1, 2, 3, 5, 7, 9 and 12 according to the non-invasive Cat-Stress-Score (CSS). Body weights were measured on days 0, 7 and 12. Finally, adoption rates and length of stay (LOS) were determined. Major findings of the study are: (1) the mean Cat-Stress-Score decreased with time for all cats, but cats with a hiding box showed a significant faster decrease in the CSS, reaching a lower CSS-steady state seven days earlier than the control group; (2) nearly all cats in both groups lost significant body weight during the first two weeks; (3) hiding boxes did not significantly influence weight loss; (4) no differences were found in the adoption rates and the LOS between both groups. Hiding enrichment reduces behavioural stress in shelter cats during quarantine situations and can therefore be a relatively simple aid to shelter adaptation. It offers no prevention however against feline weight loss, which indicates a serious health risk for shelter cats.

This statement is required for submission and will appear in the published article if the submission is accepted. Please make sure it is accurate.

Unfunded studies
Enter: The author(s) received no specific funding for this work. The authors have declared that no competing interests exist. NO    While staying in an animal shelter, cats may suffer from chronic stress which impairs their health and 27 welfare. Offering hiding opportunities can significantly reduce behavioural stress in cats, but 28 confirmation with physical parameters is needed. Therefore, the aim of this study was to determine the 29 effect of a hiding box on behavioural stress levels (scored by means of the Cat-Stress-Score) and a 30 physical parameter as body weight in newly arrived cats in a Dutch animal shelter during the first 12 31 days in quarantine situations.

32
Twenty three cats between 1 and 10 years of age were randomly divided between the experimental (N 33 = 12) and control group (N = 11) with and without a hiding box. Stress levels were assessed on days 1, 34 2, 3, 5, 7, 9 and 12 according to the non-invasive Cat-Stress-Score (CSS). Body weights were measured 35 on days 0, 7 and 12. Finally, adoption rates and length of stay (LOS) were determined.

36
Major findings of the study are: (1) the mean Cat-Stress-Score decreased with time for all cats, but cats 37 with a hiding box however showed a significant faster decrease in the CSS, reaching a lower CSS-steady 38 state seven days earlier than the control group; (2) nearly all cats in both groups lost significant body 39 weight during the first two weeks; (3) hiding boxes did not significantly influence weight loss; (4) no 40 differences were found in the adoption rates and the LOS between both groups.

41
Hiding enrichment reduces behavioural stress in shelter cats during quarantine situations and can 42 therefore be a relatively simple aid to shelter adaptation. It offers no prevention however against feline 43 weight loss, which indicates a serious health risk for shelter cats.  [1,5,6,8,9], and chronic stress can therefore harm a cat's health as well [5,7,10,11].

54
When in a state of stress, the majority of cats will stop eating. Tanaka

58
Several studies show that stressed cats display increased alert resting behaviour behind their litter box 59 in an environment without hiding opportunities [10,16,17]. This is interpreted as alternative hiding 60 behaviour for it offers some concealment [10,16]. Real concealment can be offered by providing a hiding 61 box to shelter cats. A study of Kry and Casey [17]  from stress than cats without a hiding box [10]. The present study was designed with more frequent CSS 69 scoring between day 5 and 12, to gain more insight in the feline recovery to stress and to relate these 70 behavioural stress levels to a physical parameter as body weight.

-4 -
The primary aim of this study was to determine the effect of a hiding box on behavioural stress levels 72 and on body weight of newly arrived cats in a Dutch animal shelter during the first 12 days in quarantine.

73
The additional aim was to compare the Length of Stay (LOS) of cats in both study groups. It was 74 hypothesized that a hiding box would significantly reduce stress levels of newly arrived cats compared 75 to the non-hiding box group, reflected in a lower CSS, less weight loss and a shorter LOS. 102 Inclusion criteria for this study were based on breed (European shorthair cats), health status and age 103 (between 1 and 10 years of age). When new cats showed no clinical signs of illness, obvious heat, 104 pregnancy or signs of nursing during the physical examination at intake, they were included in this 105 study. As it is not generally accepted practice in Dutch animal shelters to screen apparently healthy cats 106 through diagnostic testing (e.g. FIV/FeLV) at shelter intake, apart from the physical examination, no 107 additional information was available on the feline health status of the cats in this study.

108
All cats were observed for at least 12 days after intake.

109
During the study two cats participating in this study, left the shelter before their last observation day: 110 from the Hiding box group one cat went to a foster home, from the Control group one cat was released 111 within a trap-neuter-relaese (TNR) program. Data of both cats were excluded from this study. Two other 112 cats were not included in data for the length of stay, but were included in data for the Cat-Stress-Score, 113 body weigth and the adoption rate. After the 12 days observation period, one of these cats (nr. 8, control 114 group) proved to be infected with FeLV and was euthanized a few days after the quarantine period of 115 14 days, while another cat (nr. 19, control group), because of its semi feral behaviour, was also released 116 through the TNR program. Because shelters often take in these non-clinical but infected cats and stray 117 cats being poorly socialized (and even rehome them), this study has included these two cats in three of 118 the four measured parameters.

119
The reason for selecting cats between 1 and 10 years of age was to avoid inclusion of juveniles with 120 related specific behaviour and elderly cats with increased chances for subclinical disease which might 121 influence their behaviour [18,20]. As previous studies [17,18] found no gender related significant 122 differences in stress behaviour, both male and female cats were included in the present study. The 23 123 cats were randomly assigned to one of the two groups with and without access to a hiding box. Cats were given an habituation period of 24 hours after shelter intake (= Day 0), before behavioural 150 assessment was performed [17]. Behavioural data were collected on days 1, 2, 3, 5, 7, 9 and 12 between 151 12:30 and 5:15 PM, during which interactions with caretakers were avoided.
-7 -Each cat was observed for 20 minutes per day by using video-recording. Outside the cage a video camera 153 (H.264 DVR) was mounted on a tripod at cage height. For new observations the combination camera-154 tripod had to be readjusted to the new cat cage. Video recordings were viewed in real-time in an adjacent 155 room and stored for subsequent analysis (Fig 1). Only one camera was used for recording.

226
The cats in this study are presented in the appendix with their ID, experimental group, gender, age, 227 bodyweight at intake (kg) and the quarantine wards they went after intake.  For the comparison of both experimental groups, the absolute body weight was used. At intake the 255 control cats were on average 300 grams heavier than those in the experimental group. This initial weight 256 difference between both groups reduced to 210 grams at Day 7 and Day 12: cats in the control group 257 lost overall 7.7% of their initial body weight, while cats with a hiding box lost 6.3% of their initial body 258 weight during those 12 days ( Table 2). The initial weight and weight reduction between the groups 259 however proved not to be significant.

271
The individual proportional decrease in body weight is visualized in Fig 3. All cats except one lost 272 weight during both weeks. When weight loss at Day 12 was calculated as a percentage of initial body 273 weight at intake, it was found that 7 of the 23 (35%) cats lost ≤ 5% of their body weight, whereas 15 of 274 the 23 (65%) cats lost 5% or more of their weight. The maximum body weight loss was found in cat nr.

282
Of the 23 shelter cats in this study, 21 were rehomed after the observation period was finished. In the 283 control group 9 out of 11 cats were adopted (82%), in the experimental group 12 out of 12 (100%). No 284 significant difference was found in the adoption rate between the two groups (p = 0.55).

285
As we defined LOS as the number of days between the shelter intake of a cat and its day of adoption, 2 286 cats were not included in this data set, for they were not adopted.

287
The mean LOS for the control group (n = 9) was 24.

323
When otherwise healthy cats loose weight unintentionally, it is a dramatic indicator of a health risk.

324
Weight loss can be caused by insufficient nutrional management (the shelter offers inadequate quantity 325 and/or quality of food) and also by a decrease in feline appetite by a physical stress response. Although 326 food intake was not registered in the present study, it was observed that some cats were completely 327 anorectic, especially during the first days. For the shelter this was the reason to standardize the feeding

337
The effect of stress however on the body weight of shelter cats was first shown by Tanaka et al. [12], 338 who found a negative correlation between food intake and stress scores of cats. The conclusion was that 339 cats, admitted to an animal shelter were likely to lose weight while in the shelter. These results are 340 consistent with our findings, indicating that a decrease of feline appetite caused by a physical stress 341 response, is most likely responsible for the weight loss.

342
Although the provided commercial food in this study was of a high quality, there is less understanding 343 of the role of palatability of food for shelter cats in relation to weight loss. The only cat in this study 344 which gained weight, received medication for cystitis (meloxicam 0.05 mg/kg) mixed with canned food.

345
This gives an indication of the importance of palatability of food for shelter cats.

346
Although the analysis of the effect of time and the presence of a hiding box on the body weight suggested 347 that there was a difference between the two groups in body weight losses, as cats with hiding boxes 348 showed approximately 40 grams less weight loss in comparison with the control group, this difference 349 was not significant. For the individual cat, however, this could be biologically relevant, for weight loss 350 due to feline anorexia has a serious impact on a cat's health, increasing the risks of hepatic steatosis 351 [5,14,15]. A significant difference in body weight between cats with and without hiding facilities, 352 however, was not identified. Therefore more research is necessary to monitor these cats for a longer 353 period of time, to register the process of adaptation to the new environment in correlation to the weight 354 losses and to experiment with ways to prevent or reduce body weight losses in shelter cats (i.e. highly 355 valued palatable food items might overcome the fear motivation and stimulate consumption despite the 356 challenging environment).

357
Apart from stress, progressive weight loss can also be a sign of serious medical problems [32]. One of 358 the cats from the control group showed a weight loss close to 20% in 12 days and was eventually -15 -diagnosed with FeLV. Shelters could use weight loss during quarantine time as an early warning sign 360 for serious declines in physical conditions, but this asks for a daily weighing as a standard procedure.

362
Adoption rates and length of stay (LOS) 363 Sometimes shelter staff expresses their worries about using hiding boxes, for boxes might decrease the 364 visibility of cats to potential adopters and therefore slow down adoption rates (personal 365 communications). Kry and Casey however showed that providing shelter cats with hiding enrichment, 366 did not decrease the likelihood of those cats being adopted [17]. Also our study did not reveal differences