ULTRASONOGRAPHIC ADRENAL-GLAND MEASUREMENTS IN DOGS WITHOUT EVIDENCE OF ADRENAL DISEASE

The purpose of this retrospective study was to determine measurements of adrenal glands from longitudinal sonograms, in a large population o f dogs and to correlate these measurements to age, sex and descriptors of body size. Dogs were selected from the clinic population presented for routine abdominal ultrasonography between September, 1991 and Mar ch, 1994. Dogs with elevated serum cholesterol or alkaline phosphatase levels, polyuria/polydipsia, and/or clinical diagnoses of adrenal pat hology were excluded. Dogs with ultrasound-documented abnormalities (m ass lesions, abnormal architecture) of either adrenal were not conside red. Age, sex, weight and breed were recorded, and the body surface ar ea of each dog was calculated. Adrenal length and caudal polar width w ere determined from longitudinal, two-dimensional ultrasound images. A drenal measurements were compared with body size measurements and age using least squares linear regression analysis and the correlation coe fficient (r) and coefficient of determination (r(2)) calculated. One h undred and ninety three dogs were included in the study, with a weight and body surface area range of 1.8-72 kg and 0.2-1.8 m(2), respective ly. The left adrenal gland (n = 182) length range was from 10.7 to 50. 2 mm, and the range of the caudal polar widths was 1.9 to 12.4 mm. Rig ht adrenal gland (n = 85) length range was from 10 to 39.3 mm, and the range of the caudal polar widths was from 3.1 to 12 mm. In dogs where both adrenal gland lengths were measured (n = 74), the right adrenal gland length was less than that of the left in 46 dogs, equal to the l eft in one dog, and greater than the length of the left in 27 dogs. Th e strongest linear association was noted between the left adrenal glan d length and body weight (kg, r = 0.71, p < 0.0001) or body surface ar ea (m(2), r = 0.71, p < 0.0001). Similar significant association was n oted between the right adrenal gland length and body weight (kg, r = 0 .69, p < 0.0001). A significant positive association was also noted be tween age and left adrenal gland length, (r = 0.25, p = 0.009). The su mmation of all four adrenal measurement values (left length and width, right length and width) did slightly improve the correlation (r = 0.7 4, p < 0.0001) when compared with body weight (kg). There was not a si gnificant difference in the adrenal measurements with regard to sex. W ith regard to the correlation coefficient values, there was no advanta ge noted to calculating the body surface area from body weight. The ca uses for the low degrees of the correlation between adrenal size and t he variables investigated in this study are unknown; measurement error , effects of non-adrenal illness on adrenal size, and a non-linear or complex linear relationship between adrenal gland size (as measured by longitudinal parameters) and descriptors of body size are among possi ble explanations.