Multifrequency bioimpedance analysis (MF-BIA) measures electrical resistance, reactance, and phase angle across multiple frequencies to assess body fluid distribution and cellular electrical properties. At low frequencies, current predominantly traverses the extracellular space and thus low-frequency resistance reflects extracellular fluid status, whereas higher frequencies probe both extracellular and intracellular compartments. MF-BIA is widely used in human medicine and single-frequency BIA has been applied in dogs, but fundamental data on MF-BIA measurement characteristics and the effects of electrode type, electrode placement, and body position in dogs are limited.
The investigators aimed to (i) characterize trunk MF-BIA measurements in anesthetized dogs and quantify repeatability and inter-individual variability across body positions and electrode placement sites, and (ii) evaluate clip electrodes as a non-invasive attachment method compared with needle electrodes, which are commonly used in experimental settings. The primary hypothesis was that clip electrodes would show acceptable repeatability and generally good agreement with needle electrodes under controlled anesthetized conditions, while electrode placement and body position would contribute to between-condition differences.
Seven intact adult Beagle dogs (4 males, 3 females), aged 7–9 years and weighing 8.9–15.5 kg (mean 12.9 ± 2.2 kg), were studied under an approved protocol. Dogs were judged clinically healthy by physical exam and routine laboratory testing, with minor abnormalities in a few animals reported in the manuscript. Anesthesia was induced with sevoflurane and maintained at 2.5% in oxygen. Rocuronium was administered (0.5 mg/kg IV then 0.5 mg/kg/h CRI) to minimise movement; controlled ventilation targeted end-tidal CO2 of 35–40 mmHg. Lactated Ringer’s solution (3 mL/kg/h) was infused. Measurements were started after stabilization, and dogs recovered uneventfully after reversal with sugammadex.
Trunk MF-BIA was performed using a tetrapolar bioimpedance spectrometer (InBody M20) to record resistance, reactance, and phase angle at 5, 50, and 250 kHz. Two electrode types were compared: noninvasive clip electrodes (with conductive ECG cream applied) and 25-gauge needle electrodes inserted subcutaneously to 1 cm. Measurements were obtained in sternal and left lateral recumbency on a non-conductive silicone mat; a cotton towel minimized limb contact in lateral recumbency.
For each electrode type and body position, three electrode paths were measured: right elbow to right stifle (ESt), first thoracic spinous process to right stifle (TSt), and vertex of the head to sacral spinous process (HSc). The fixed measurement sequence was: sternal with clips, lateral with clips, sternal with needles, lateral with needles. After each position change, a 15‑minute stabilization period was allowed. For each path and condition, 150 repeated measurements were recorded at 50 ms intervals; the mean of these 150 values was used for analysis.
Cole–Cole modelling was applied to derive R0 (resistance at zero frequency), Rinf (resistance at infinite frequency), Xmax (maximum reactance), and PAmax (maximum phase angle). Measurements failing the Cole–Cole criterion (center resistance outside the range between R at 5 and 250 kHz) were excluded. No randomization or blinding was performed because all dogs underwent a fixed measurement sequence.
Repeatability was assessed by coefficient of variation (CV) within each set of 150 repeats. Inter-individual variability used CV across dogs for each condition. Agreement between electrode types was evaluated with Bland–Altman analysis (bias and 95% limits of agreement) and correlation analysis. Linear mixed-effects models tested fixed effects of electrode type and body position with dog as a random effect. A P value <0.05 was considered significant.
A total of 1,800 measurements were completed per dog (150 repeats × 3 paths × 4 conditions). Measurements were feasible under all tested conditions. Repeatability within individual measurement series was low (stable), and inter-individual variability across dogs was generally in the 10–15% range. However, reactance and phase angle measured at 5 kHz showed higher inter-individual variability compared with other parameters.
Agreement analysis showed small mean biases between clip and needle electrodes for most parameters across frequencies and electrode paths. Despite small mean biases, Bland–Altman limits of agreement varied among parameters. Reactance and phase angle at 5 kHz demonstrated weaker correlation between electrode types and substantially wider limits of agreement, indicating greater measurement disagreement at low frequency. Linear mixed-effects modelling indicated that between-dog differences accounted for the majority of observed variability. Body position had measurable effects on several MF-BIA parameters when the stifle was included in the electrode path, indicating that limb orientation and contact can influence trunk impedance readings.
Under tightly controlled anesthetized conditions with standardized preparation and stabilization, clip electrodes produced MF-BIA measurements with small mean biases relative to needle electrodes. The low intra-individual variability across repeated measures supports the repeatability of trunk MF-BIA when technical conditions are controlled. Nevertheless, the wider limits of agreement and weaker correlations for reactance and phase angle at 5 kHz indicate that electrode type influences low-frequency capacitive measurements more than higher-frequency resistance measures.
These findings imply that clip and needle electrodes should not be assumed fully interchangeable for all MF-BIA-derived variables, particularly when interpreting reactance and phase angle at low frequency. Because between-dog variability explained most measurement variance and body position affected readings when the stifle was in the path, consistent electrode placement and body positioning are important when comparing measurements across subjects or tracking longitudinal changes within individuals.
Limitations stated in the manuscript include the use of a small sample of anesthetized, adult Beagle dogs and fixed measurement sequence without randomization or blinding. The study therefore provides controlled-condition data but may not capture awake or clinical patient variability.
Clip electrodes can provide trunk MF-BIA measurements with small average bias relative to needle electrodes in anesthetized dogs, but they are not fully interchangeable with needle electrodes for all parameters, notably reactance and phase angle at low frequency (5 kHz). Investigators and clinicians should standardize electrode type, electrode placement, and body position for experimental comparison and for longitudinal monitoring within individual dogs. The manuscript’s data and supporting information are available with the article for further review.