Starvation and the Clinicopathologic Abnormalities Associated with Starved Dogs: A Review of 152 Cases
The objectives of this retrospective study were to identify the most common clinicopathologic abnormalities in starved dogs, assess the time required for those abnormalities to resolve, and determine whether clinicopathologic abnormalities recorded at time of intake to the hospital influenced time to regain weight. Records of 152 very underweight or emaciated dogs seized by the American Society for the Prevention of Cruelty to Animals (ASPCA) Humane Law Enforcement (HLE) division were reviewed. Dogs were classified as emaciated if the admission body weight was estimated to be ≥ 30% below the anticipated ideal body weight and classified as very underweight if the admission weight was estimated to be 20–29% below the anticipated ideal body weight. An initial minimum database was obtained on each animal, and when possible, clinicopathologic abnormalities were serially assessed. The most common initial abnormalities, present in ≥ 25% of dogs, were hypoalbuminemia, thrombocytosis, anemia, elevated blood urea nitrogen (BUN), elevated BUN/creatinine ratio, and hypocalcemia. Mean time to gain 20% of admission body weight was similar for the abnormalities studied. Although there was some evidence that dogs with anemia and/or hypoalbuminemia required more days to gain weight, future studies are required for confirmation.
Introduction
Intentional starvation is a common form of animal abuse, and emaciated body condition was one of the most common reasons for a Humane Law Enforcement (HLE) officer to seize dogs in New York, NY during the time period studied. The American Society for the Prevention of Cruelty to Animals (ASPCA) HLE officers are New York State Peace officers with the authority to investigate animal cruelty, seize abused animals, and make arrests to prosecute animal abusers.
There is a paucity of published information in the current veterinary literature about intentionally starved animals. Early studies on the effects of starvation in dogs were conducted in controlled laboratory environments, and food was purposely and systematically withheld by the investigators.1–3 Clinicians involved in real life starvation cases need to know what abnormalities to expect and how long those abnormalities may take to resolve.
The primary goals of this retrospective study were to identify and analyze the most common clinicopathologic abnormalities in starved animals. Time required to gain 20% of their intake body weight was assessed, but because most animals had more than one clinicopathologic abnormality it was not possible to accurately determine how any individual abnormality influenced time to weight gain.
Materials and Methods
Medical records of 152 dogs seized by the ASPCA’s HLE division from January 2007 to May 2009 were reviewed. Criteria for inclusion in the study were veterinarian-assessed body condition score at admission of 4–5/5 based on a modified Tufts Animal Care and Condition scale (Figures 1A–D, Table 1), ≥ 6 mo of age, no apparent systemic illness (based on initial physical exam and blood work performed at admission) that could have caused the observed degree of emaciation, and an adequate medical record for evaluation. A body condition score of 4–5/5 on the modified TACC scale corresponds to an animal estimated at intake to weigh at least 20% less than its ideal body weight.4
Citation: Journal of the American Animal Hospital Association 49, 2; 10.5326/JAAHA-MS-5762
A physical exam and blood work were performed on each animal at intake and abnormalities were recorded. Dog breed and estimated age were recorded. An identical series of tests was performed on each animal, including a complete blood cell count, serum biochemical analysis, thyroxine level, urinalysis, heartworm antigen test, fecal floatation and analysis for Giardia spp. using an enzyme-linked immunosorbent assay. Due to the well-documented breed association with the disease, Pit bull-type dogs were tested for Babesia spp. by polymerase chain reaction.5–7 Blood work and fecal analyses were performed at outside laboratoriesa,b. The timing and type of subsequent laboratory tests were at the discretion of the attending clinician.
Dogs evaluated as clinically dehydrated were administered IV fluids, and dogs with evidence of infection (most commonly skin infections and decubital ulcers) were prescribed antibiotics. Further treatments were at the discretion of the attending clinician.
Each dog was weighed at intake and serially thereafter (a minimum of q 7 days). All dogs were offered a commercially available veterinary therapeutic dietc to meet daily energy requirement (DER). In intact adult dogs, DER was calculated using the following equation: resting energy requirement (RER) × 1.8. In castrated and spayed dogs, DER was calculated using the following equation: RER × 1.6. The RER was calculated using the following equation: body weight at intake (kg)0.75 × 70.8 This approach was instituted in an attempt to prevent refeeding syndrome. After 10–14 days, the dogs were transitioned to a commercially available adult maintenance dietd to meet 150% of the calculated DER at time of intake until adequate weight was gained (based on the assessment of the attending clinician). Food intake was neither measured nor recorded during either period.
Results
The mean and median ages of dogs included in this study (based on estimated ages at intake) were 3.5 yr and 2 yr, respectively (range, 6 mo to 15 yr). There were 83 male dogs (13 castrated) and 69 female dogs (1 spayed). Twenty-seven percent of the dogs had positive fecal tests at admission. The types of parasites found were Trichuris vulpis (n = 13), Ancylostoma caninum (n = 11), Giardia spp. (n = 10), Toxocara canis (n = 9), Isospora spp. (n = 3), and Capillaria spp. (n = 1). Overall, 53.9% of the dogs were Pit bull-type dogs.
The mean time required for all dogs to gain 20% of the body weight measured at intake was 20.7 days (median, 19 days; range, 1–56 days). Seventeen dogs failed to gain 20% of the body weight measured at intake within the study period.
Clinicopathologic abnormalities affecting at least 15% of the population (in order of frequency) were hypoalbuminemia, thrombocytosis, elevated blood urea nitrogen (BUN)/creatinine ratio, elevated BUN, anemia, hypocalcemia, hyperglobulinemia, monocytosis, decreased albumin/globulin ratio, elevated γ-glutamyltransferase, neutrophilia, hypomagnesemia, decreased total thyroxine, elevated alanine aminotransferase, and elevated aspartate aminotransferase.
Details regarding the number of dogs affected, number of dogs experiencing and not experiencing resolution of the abnormalities, and time to resolution for abnormalities affecting at least 25% of the population have been summarized in Table 2. The groups of dogs described in Table 2 are not mutually exclusive, and any given dog could have had more than one clinicopathologic abnormality concurrently. Of the 152 dogs studied, 31 dogs had none of the most common clinicopathologic abnormalities listed above, 29 dogs had 1 abnormality, 35 dogs had 2 abnormalities, 34 dogs had 3 abnormalities, 15 dogs had 4 abnormalities, 5 dogs had 5 abnormalities, and 3 dogs had all 6 of the most common abnormalities affecting at least 25% of the population.
An individual dog could have had > 1 abnormality concurrently (i.e., anemia and hypoalbuminemia in the same animal), and the mean days to recovery could also have been affected by other abnormalities present.
Total calcium was measured.
BUN, blood urea nitrogen.
Mean time to gain 20% of the body weight measured at admission was similar for the common abnormalities studied (i.e., those affecting at least 25% of the population), ranging from 18.7–23.5 days.
Fifty-five dogs had an elevated BUN/creatinine ratio at time of admission, but only 53 had a urine specific gravity measured concurrently. Of those 53 dogs, 39 (73.6%) had urine specific gravities ≥ 1.035. Of the 55 dogs with elevations in BUN/creatinine ratio, 16 (29.1%) dogs did not have the ratios rechecked, and 6 (10.9%) dogs did not have resolution of their elevated BUN/creatinine ratio within the study period. Seventeen of those 22 dogs (77.3%) had a urine specific gravity ≥ 1.035.
Forty-seven dogs had elevated BUN levels at intake, but only 45 had urine specific gravities assessed concurrently. Of the latter 45 dogs, 77.8% (35 of 45) had urine specific gravities ≥ 1.035. Of the total population with elevated BUNs, 14 dogs (29.8%) did not have their abnormal BUN levels rechecked within the time frame studied, and in three dogs (6.4%), the elevated BUN did not resolve. Of this total of 17 dogs, 15 (88.2%) had a urine specific gravity value ≥ 1.035.
Of the 47 dogs with anemia at presentation, 32 (68.1%) had nonregenerative anemia, 14 (29.8%) had regenerative anemia, and 1 (2.1%) had a reticulocyte count that was not reported. Classification of anemia was determined by reticulocyte count, with a reticulocyte count > 60,000/uL used as evidence of regeneration. Anemia did not resolve in 10 (21.3%) dogs, and 3 of those dogs tested positive for Babesia gibsoni.
Thirty-eight dogs presented with hypocalcemia. Of those, 31 (81.6%) had concurrent hypoalbuminemia.
Discussion
Hypoalbuminemia was one of the two most common clinicopathologic abnormalities identified in this population of dogs. Albumin concentrations decrease with decreased synthesis, increased loss, and hemodilution.9 Decreased albumin synthesis occurs in inflammatory conditions by two major mechanisms. First, albumin is a negative acute phase protein and concentrations are decreased due to decreased production by the liver. Second, other groups of proteins (positive acute phase proteins, such as globulins) are increased with inflammation. In the face of a pathologic increase in positive acute phase proteins, albumin levels can fall to preserve normal oncotic pressure.9 The dogs in this study had numerous reasons for inflammation (e.g., skin, gastrointestinal tract, infectious disease). It is also important to note that hyperglobulinemia was the seventh most common clinicopathologic abnormality noted on the blood panels of the included dogs, experienced by 23.6% of the population (data not shown).
In cases of starvation, the rate of protein catabolism exceeds the rate of production.9 Body protein is the last tissue pool to be broken down in a cachectic state, so low albumin is expected only after stores of fat and muscle have been depleted, as was the case in the dogs in this study.9 The dogs in this study also had numerous possible causes of protein loss, including gastrointestinal loss (e.g., vomiting, diarrhea, chronic bleeding), exudative skin disease, chronic wounds, and ecto- and endoparasites.
It is likely that the percentage of dogs with true hypoalbuminemia was underestimated due to the percentage of dogs with concurrent dehydration and hemoconcentration.
In a previous study, a statistically significant decrease in total protein and albumin concentrations was reported in dogs starved in a research setting for 21 days.1 Interestingly, those dogs’ albumin concentrations still remained within normal range at the end of the 3 wk experiment. It is possible that a dog must be starved for > 21 days for albumin concentrations to fall below the reference range in the absence of concurrent causes of protein loss.
Thrombocytosis was the other most common clinicopathologic abnormality noted at admission in this population. Physiologic thrombocytosis, also called redistribution thrombocytosis, occurs with epinephrine release secondary to stress and/or excitement. Platelets are redistributed from the lungs and released from the spleen.9 Reactive thrombocytosis results in a mild to moderate increase in platelets, and the health of the animal generally is not affected. Reactive thrombocytosis has been associated with blood loss (especially chronic), iron deficiency, and a wide variety of infectious and noninfectious inflammatory conditions (secondary to an increase in the cytokine interleukin-6).10–12 The platelet increases in this study likely were both reactive and physiologic in nature. Chronic inflammatory skin diseases were commonly reported, and most dogs were housed in unsanitary conditions and required multiple baths to remove urine and fecal stains and odors from their coats. Included dogs also had numerous risk factors for chronic blood loss and possible iron deficiency, including intestinal parasitism, chronic foreign bodies (presumably from eating whatever substances were available to them), stress-induced and/or mechanical gastric ulceration. Several dogs also had confirmed infectious disease, including Dirofilaria immitis, Babesia gibsoni, and Salmonella enterica (data not shown).
Elevated BUN/creatinine ratio and elevated BUN were the next most common abnormalities in this study. Prerenal azotemia secondary to dehydration is an obvious explanation for those findings, given the well-concentrated urine (specific gravities were ≥ 1.035) of the majority of the dogs with these blood work changes. Affected dogs also had significant decreases in muscle mass, likely resulting in decreased blood creatinine concentrations. Gastrointestinal bleeding and increased proteolysis can also cause an increased BUN and BUN/creatinine ratio.9
Anemia was a common abnormality in this study, and the majority of cases were nonregenerative. As with hypoalbuminemia, it is likely that the percentage of affected animals and the magnitude of the anemia was underestimated due to the number of dogs that presented dehydrated and hemoconcentrated. Anemia of chronic disease, the most common cause of nonregenerative anemia in domestic mammals, is caused by inflammation and is likely responsible for a significant number of the nonregenerative anemias seen in this population.9 In anemia of chronic disease, red blood cell lifespan is decreased, there is impairment of iron mobilization and utilization, and red blood cell production is decreased.9 In addition, bone marrow suppression was previously reported in a group of dogs starved in a laboratory setting for 20 days.3
Nutrient deficiencies, such as iron, copper, folate, and cobalamin deficiencies, can also result in nonregenerative anemia. In this study, iron depletion secondary to chronic gastrointestinal bleeding is a possibility.9
Further, blood loss anemia caused by ecto- and endoparasitism, gastric ulcerations, and chronic wounds are a potential cause of regenerative anemia.9 Hemolytic anemia cannot be excluded as a cause of regenerative anemia in this study given that four dogs tested positive for Babesia gibsoni. Given their poor living conditions and immunocompromised states, other infectious diseases that could cause hemolysis are possible. In addition, oxidative damage to red blood cells from ingestion of inappropriate items is a possible cause of hemolysis.9
Finally, low total calcium was noted on initial blood work in 25% of the dogs in this study. The majority of affected dogs (31 of 38) had concurrent hypoalbuminemia, offering an obvious explanation for the finding. Other possibilities in this population include hypovitaminosis D and pancreatitis.9 Ionized calcium concentrations were not measured.
Although there are numerous causes for anemia and hypoalbuminemia, the most likely causes in the dogs included in this study (i.e., anemia of chronic disease and hypoalbuminemia secondary to inflammation and pathologic breakdown of body protein) are associated with more chronic debilitation. It is reasonable to hypothesize that animals that have sustained a longer period of starvation, possibly evidenced by low albumin and/or low hematocrits, could take longer to regain weight. When dogs with anemia and/or hypoalbuminemia were compared with the population as a whole, there was some evidence that they did in fact require more time to regain adequate weight. Specifically, anemic dogs took 23.5 days to regain weight and hypoalbuminemic dogs took 22.2 days. However, due to the fact that 121 of the 152 (79.6%) dogs included in this study had multiple multiple clinicopathologic abnormalities, it was impossible to accurately determine how any individual clinicopathologic abnormality may have influenced time to regain weight.
Limitations of this study are those inherent to retrospective studies. The data presented cannot be used to establish a cause and effect relationship; only associations can be asserted. A timetable for retesting clincopathologic abnormalities was not established in advance of patients’ admission to the hospital. Some abnormalities were deemed insignificant and were never re-evaluated. Some abnormalities were re-evaluated after longer time intervals, and the exact period of time when an abnormality resolved could not be definitively established. Similarly, the exact day when a dog had regained 20% of intake body weight could not be definitively determined because dogs were generally weighed q 7 days after initial stabilization. Finally, several veterinarians were involved in the care of the patients, which likely affected some of the subjective data points in the study, such as body condition scoring.
Conclusion
Clinicians involved in starvation cases should be aware of common clinicopathologic abnormalities in this population of dogs as well as possible causes (Table 3). Dogs suspected to be the victims of starvation need to be assessed thoroughly and carefully to monitor their recoveries and to convey to authorities the extent of their neglect. Suggested initial work-up for canine starvation cases can be found in Table 4. Future studies are warranted to more accurately assess how clinicopathologic abnormalities, specifically the combination and severity of common clinicopathologic abnormalities, may affect time to regain weight.
BUN, blood urea nitrogen.
All must be well documented and complete.
EDTA, ethylenediaminetetraacetic acid; ELISA, enzyme-linked immunosorbent assay; PCR, polymerase chain reaction; PCV, packed cell volume; RER, resting energy requirement; TS, total solids.
A: Lateral view of a dog showing a body condition score of 5/5 on the modified Tufts Animal Care and Condition scale. Note the obvious loss of muscle mass, prominence of skeletal structures, severe abdominal tuck, and extreme hourglass shape. At time of hospital admission, this dog weighed 11.3 kg. B: Dorsoventral view of the same dog at time of hospital admission. C: Photograph of the same dog shown in panels A and B 30 days after hospital admission, weighing 18.2 kg. D: Dorsoventral view of the same dog 30 days after hospital admission.
Contributor Notes
