Hypovitaminosis A in Reptiles: Squamous Metaplasia, Aural Abscesses, and the Narrow Dosing Margin

Bottom line

Hypovitaminosis A is a husbandry disease of insectivorous lizards and carnivorous/omnivorous chelonians, and it is a diagnosis of diet history plus lesion pattern — not one you make from swollen eyelids alone. Vitamin A is required for normal epithelial maturation; deficiency produces squamous metaplasia and hyperkeratosis across glandular and mucosal epithelia, which is why one patient presents with blepharoedema, the next with an aural abscess, and a third with respiratory disease [1][3][5]. Correct the diet first. If you reach for parenteral vitamin A, know that the top of the published "safe" band (5000-10,000 IU/kg) and the bottom of the published toxic band (50,000-100,000 IU/kg) are only five-fold apart, and that iatrogenic hypervitaminosis A — epidermal ulceration and sloughing managed as a burn over months — is the predictable cost of overshooting [2].

Pathophysiology & causes

Vitamin A drives epithelial differentiation, and its absence converts secretory and ciliated epithelium into stratified keratinising epithelium. Boyer frames the mechanism directly: "Vitamin A is an essential hormone that activates genes for maturation of immature epidermal cells," and without adequate levels the result is hyperkeratosis and squamous metaplasia [3]. The Merck Veterinary Manual states the same in one line: "Lack of vitamin A leads to squamous cell metaplasia and hyperkeratosis" [1].

The downstream consequences follow from where that metaplastic, keratinising epithelium sits:

• Ocular/adnexal glands. Loss of normal secretion and keratin accumulation under the lids produces the classic palpebral and conjunctival oedema. Elkan and Zwart's 1967 description of ocular disease in young terrapins established this as a systemic epithelial process with concurrent pathology in kidney, liver and pancreas, not an isolated orbital disease [4].
• Respiratory and auditory tract. Holladay and colleagues documented mucosal hyperplasia and squamous metaplasia in conjunctiva, pharynx, trachea and auditory tube of box turtles, with "accumulated keratin-like material in the middle ear cavity" [5]. Ciliated respiratory epithelium replaced by keratinising squamous epithelium loses mucociliary clearance, which is the mechanistic route to secondary pneumonia rather than a primary viral or bacterial event.
• Ductal epithelium in advanced disease. In a well-characterised western pond turtle (Actinemys marmorata) case, squamous metaplasia and keratin accumulation extended to the nasal vestibule and sinus mucous glands, the pancreatic ducts (thickened, hyperkeratotic) and the renal collecting ducts (distended and occluded by keratin debris), along with ureters and urinary bladder [6]. Merck notes that "edema secondary to nephropathy can ensue" [1].
• Integument. Poor shedding is part of the same lesion set; Boyer lists dysecdysis among presenting signs in insectivorous lizards [3]. See the dysecdysis hub for the differential workup when shedding failure is the chief complaint.

Dietary causes

The deficiency is almost always dietary and almost always accompanied by other husbandry failures. In de la Navarre's ornate box turtle case the abscesses "were the result of overall poor husbandry combined with a diet deficient in either beta-carotene, preformed vitamin A retinol, or retinyl ester" [8]. The recurring offenders in practice are all-lettuce or otherwise inappropriate herbivore rations, unsupplemented feeder insects fed without gut-loading, and exclusively muscle-meat or low-quality commercial diets.

Preformed vitamin A vs carotenoids — a genuinely contested point

The conventional teaching is that many reptiles convert carotenoids poorly and therefore require preformed vitamin A. Merck states that "The commonly used nutritional supplement beta carotene cannot be metabolized by many reptiles" [1]. Mayer and Huang write that "Herbivorous reptiles typically do not experience vitamin A deficiency because of their plant-based diet and their ability to synthesize vitamin A," while "some reptiles, such as carnivorous turtles and other carnivorous reptiles (e.g., box turtles), are not able to synthesize beta-carotene well" [2]. Boyer likewise reports that the commonly affected insectivores are leopard geckos, panther, veiled and Jackson's chameleons and anoles, while herbivores and omnivores such as tortoises, iguanas and bearded dragons are generally unaffected [3].

Against that, Cojean and colleagues ran ten clinically healthy female leopard geckos through ten weeks of supplementation — one group on vitamin A, one on beta-carotene, with insects gut-loaded accordingly. Epithelial squamous metaplasia "was not observed in any of the geckos," and hepatic vitamin A concentration was significantly higher in the carotenoid-supplemented group than in the vitamin A-supplemented group (p = 0.03); the authors concluded that in leopard geckos, beta-carotene supplementation "allows sufficient vitamin A hepatic storage" [7].

Synthesis: carotenoid-conversion capacity is species-specific and not settled. It is defensible to rely on gut-loaded, carotenoid-rich prey in leopard geckos on the strength of [7]; it is not defensible to extrapolate that to carnivorous chelonians, where the available guidance says conversion is poor [1][2]. As of 2026 the leopard gecko study remains a small (n = 10), single-species experiment.

Clinical presentation

Merck's presenting picture is "poor growth, anorexia, stomatitis, and blepharedema" resulting from squamous metaplasia, with "Secondary respiratory infections, aural abscesses, and edema secondary to nephropathy" following [1]. Stomatitis overlapping with this presentation should be worked up on its own terms — see the stomatitis / mouth rot hub.

Chelonians. Bilateral palpebral and conjunctival oedema is the classic presenting sign. de la Navarre's ornate box turtle presented "depressed, weak, and lethargic with significant bilateral aural swellings, mild bilateral blepharoedema, and a mild mucopurulent nasal discharge," with a history of slowly progressive neck swellings, puffy eyes, periodic nasal discharge, and decreased appetite and activity [8]. Aural swelling in a box turtle is the single most vitamin-A-suggestive presentation you will see, though the causal evidence is weaker than the association implies (below).

Insectivorous lizards. Boyer describes patients presenting with ocular disease unresponsive to antibiotics, blepharospasm, cheilitis, anorexia and dysecdysis. In leopard geckos he lists "Mucoid-to-solid cellular debris under eyelids, ulcerative keratitis, periocular gland abscessation," while chameleons show "difficulty capturing prey with the tongue" and dull coloration [3].

Aural abscesses in box turtles: what the evidence actually supports

The association is real and the causal claim is soft. Read the primary literature before you commit to the vitamin A narrative:

• Lesion evidence (supportive). Brown and colleagues evaluated the tympanic epithelium in 27 box turtles (10 with aural abscesses and 17 without). Lesions in affected turtles included "hyperplasia, squamous metaplasia, hyperemia, cellular sloughing, granulomatous inflammation, and bacterial infection," more severe than in unaffected turtles and more severe on the abscessed side in unilateral cases. Organs from 21 turtles showed minimal histopathologic change elsewhere. The authors concluded the changes were "consistent with a syndrome that may involve hypovitaminosis A" [9].
• Contaminant/vitamin A evidence (mixed). Holladay and colleagues found three organochlorine compounds and total organochlorines significantly higher in abscessed turtles, with a nonsignificant trend toward reduced vitamin A: mean serum and hepatic vitamin A were 71% and 49% respectively of levels in turtles without abscesses [5].
• Evidence against a simple model. Sleeman and colleagues studied 68 box turtles (40 with aural abscesses, 28 without) and found hepatic vitamin A positively correlated with pathologic score (r = 0.32, P = 0.01), which they explicitly flagged as "contrary to the expected result" [11]. Kroenlein and colleagues then failed to reproduce the lesion experimentally: red-eared sliders maintained on dietary vitamin A of 0 or 5 international units/g, with or without an organochlorine mixture (2 mg/kg chlordane, 0.25 mg/kg aroclor, 1 mg/kg lindane), showed no difference in degree of tympanic squamous metaplasia or in plasma or liver vitamin A over roughly six months [12].
• Epidemiology. In 46 free-living cases, county human population density, year and season of admission, weight and sex did not affect risk; geographic location was the only associated risk factor, with cases clustered into two multi-county regions [13].
• Microbiology. Joyner and colleagues cultured 8 turtles with aural abscesses and 15 controls. Seven of eight affected turtles had growth, yielding 10 isolates; Morganella morganii (n = 2) was the only species isolated from more than one abscessed turtle. No single agent was responsible, consistent with aerobic bacteria being "secondary opportunistic invaders of environmental origin" rather than primary pathogens [10].

Practical reading: treat the abscess surgically and treat the husbandry, but do not let "aural abscess" alone justify an aggressive parenteral vitamin A course.

Differential diagnosis

Before committing to hypovitaminosis A, rule out the conditions that produce the same periocular or aural picture. de la Navarre's own differential list for the presenting box turtle was "bilateral aural abscesses, hypovitaminosis A, upper/lower respiratory tract disease, other infections, systemic disease, trauma, and chemical or parasitic inflammation" [8].

• Primary ocular disease and corneal ulceration — Boyer notes leopard geckos develop ulcerative keratitis within this syndrome, so fluorescein-stain every ocular presentation rather than assuming vitamin A [3].
• Foreign body and retained shed under the lids.
• Bacterial conjunctivitis and primary upper respiratory tract disease.
• Trauma; in one leopard gecko case series, radiographs incidentally identified a healed tail fracture [15].
• Generalised husbandry failure. Vitamin A deficiency rarely arrives alone — the same deficient diet and suboptimal thermal environment drive nutritional secondary hyperparathyroidism and shedding failure. Audit the whole enclosure, and cross-check the metabolic bone disease hub and the gout hub when the diet history is poor.

Diagnosis

Diet history is the primary diagnostic. Boyer states plainly that "Diagnosis of vitamin A deficiency is based on dietary history and clinical signs," because circulating levels are unreliable indicators [3]. Establish exactly what the animal eats, whether feeder insects are gut-loaded, and what supplement (preformed retinol vs carotenoid) is actually used and how often.

Tissue vs circulating assays. For a definitive antemortem diagnosis, de la Navarre notes that "a vitamin A assay of the liver, or a large volume of blood is required" [8]. Mayer and Huang similarly observe that liver biopsy results "can be considered more accurate" than blood but are more invasive, that fine-needle aspiration may not be diagnostic, and that "Serum levels can vary significantly compared with hepatic vitamin A levels, and serum levels might therefore not be practical for a diagnosis because of these normal fluctuations" [2]. They list IDEXX, the Animal Health Diagnostic Center at Cornell, and the veterinary diagnostic laboratories at Michigan State and Iowa State as testing laboratories; serum should be separated from the clot, refrigerated and kept away from light before transport, with turnaround of 3 to 7 business days after sample receipt [2].

Reference intervals are species-limited — this is the real constraint. A validated interval exists for one species. Louth and colleagues analysed plasma from 40 healthy captive-bred leopard geckos (20 males, 20 females) by UPLC-MS/MS on a 50 µl sample volume and established a plasma retinol reference interval of 0.06 to 0.55 µg/mL [14]. Mayer and Huang report plasma retinol figures for other groups — "In captive aquatic turtles, plasma retinol levels are 0.03 to 0.364 mg/mL and in tortoises, 0.034 to 0.415 mg/mL. Plasma retinol levels in lizards and snakes have been reported between 0.049 to 0.372 mg/mL and 0.012 to 0.049 mg/mL" [2]. Those values are transcribed as published; note that the stated units (mg/mL) differ by three orders of magnitude from the µg/mL scale of the validated gecko interval [14], so confirm units with your laboratory before interpreting a chelonian result against them. Outside the leopard gecko, treat any single plasma retinol value as supporting evidence, not a diagnosis.

Treatment: diet first, parenteral vitamin A second

Dietary correction is the definitive treatment and the only one with no toxic ceiling. Merck's recommendation is dietary improvement plus supplements containing preformed vitamin A, and it explicitly warns that "Injectable vitamin A is best avoided, as hypervitaminosis A can cause skin erythema and sloughing" [1]. Mayer and Huang reach the same conclusion: rather than an injectable, "consider providing oral medication and changing the diet to include more vegetables with vitamin A," a regimen that "will help treat hypovitaminosis A and is unlikely to cause hypervitaminosis A" [2]. Where injection is unavoidable, they note it is better to give a fat-soluble than a water-soluble formulation, "because water-soluble vitamin A can be absorbed more quickly and thus potentially be more toxic" [2].

Parenteral dosing — the published regimens disagree by more than an order of magnitude. All vitamin A use in reptiles is extra-label. Transcribed exactly, from most to least conservative:

Regimen	Source
1000-2000 IU vitamin A/kg parenterally, once a week for 2 wk (mild cases) to 6 wk (severe cases) — with the caveat that "Higher doses of vitamin A do not speed recovery and can result in hypervitaminosis A"	de la Navarre, chelonians with aural abscess [8]
5000 IU/kg IM as a single treatment (injectable vitamin A supplement in reptiles, especially turtles)	Lewbart reptile formulary [16]
5000-66 666 IU vitamin A palmitate per kg IM, SC, or PO every 1 to 2 weeks for 2 treatments — explicitly described as empiric	Boyer, insectivorous lizards [3]
An initial IM injection of 66,667 IU/kg followed by weekly injections of 10,000 IU/kg, which "appeared to be safe and effective" in three cases	Louth, leopard geckos [15]

Read that table against the toxicity section below before you prescribe. The upper bound of Boyer's empiric range and the initial dose in the Louth case series both fall inside the band that Mayer and Huang label toxic [2][3][15]. That is a real, unresolved conflict in the literature, and it is the reason a conservative starting dose plus dietary correction is the defensible default in a first-opinion setting — particularly in a small chelonian, where the volume error margin on a concentrated preparation is unforgiving.

Monitoring response biochemically. In Louth's three leopard geckos, initial plasma retinol values were all below the published reference range; two days after the initial injection values had risen into the reference range at 0.11-0.24 µg/mL, fell to 0.08-0.15 µg/mL by seven days, and were 0.22-0.26 µg/mL at six months and 0.21-0.24 µg/mL at one year [15]. That seven-day decline is the argument for the weekly repeat interval rather than a single large loading dose.

Hypervitaminosis A: the iatrogenic complication that defines the dosing margin

This is the central patient-safety issue of the condition, and it is doctor-caused. Mayer and Huang describe hypervitaminosis A as an overload "which in reptiles can be iatrogenically induced by injection of vitamin A to reptiles suspected of having hypovitaminosis A" [2].

The numbers, transcribed verbatim: "The safe dose for vitamin A administration is 5000-10,000 IU/kg, and a toxic dose is roughly 100 times higher (50,000 to 100,000 IU/kg)" [2]. Note the internal inconsistency in that sentence — the prose says "100 times higher," but the bracketed figures are ten-fold apart. Plan around the figures, not the multiplier: therapeutic 5000-10,000 IU/kg; toxic 50,000-100,000 IU/kg. And note that band-to-band ten-fold is not the margin you actually dose against: the top of the safe band (10,000 IU/kg) and the bottom of the toxic band (50,000 IU/kg) are only five-fold apart, which is narrow for a fat-soluble vitamin dosed into a 300 g turtle. The same authors add that hypervitaminosis A "can still be induced, even when a 'safe dose' of supplement is given as an injection and an increased amount of vitamin A is consumed as part of the diet for a prolonged period" [2] — so total intake, not the injection alone, is what you are titrating.

Mechanism. Excess vitamin A overwhelms hepatic function and storage capacity and leads to tissue damage, seen clinically first as dry, flaky skin [2].

Clinical picture. "The signs are dry, scaly skin; skin ulceration and/or sloughing to various degrees; depression; lethargy; anorexia; weight loss; and/or dehydration" [2]. The areas of loose skin near the front legs and neck are most affected and become swollen from cell damage when water-soluble vitamin A has been given; in severe cases "the dermis and muscle can become exposed and they can die of secondary dehydration," with secondary bacterial or fungal infection of skin lesions capable of leading to sepsis [2]. In chameleons specifically, excess vitamin A can potentially lead to nutritional metabolic bone disease through vitamin D interference [2]. Differential diagnoses for these skin lesions are infection (bacterial, fungal, viral, parasitic) and/or trauma [2].

Management, if it has already happened. Prior administration of vitamin A cannot be reversed, so treatment is supportive: decrease and potentially discontinue vitamin A intake; treat skin lesions like burn wounds with hydrotherapy, antiseptic solution soaks, systemic antimicrobials and analgesia; provide fluid therapy at 10 to 30 mL/kg/day in acute cases; consider nutritional support with or without a feeding tube [2]. Reported analgesic options (all extra-label): morphine in chelonians at 1.5 mg/kg IM or SC and in lizards at 10 mg/kg IM or SC, or meloxicam at 0.2 to 0.3 mg/kg PO or IM q24h, "but the patient must be carefully monitored for respiratory depression" [2]. Photobiomodulation has shown promising results in iguanas at 10 Joules/cm2 [2]. Warn the owner about the timeline up front: "Treatment should continue until wounds have healed, which can take 4 to 6 months" [2]. Prognosis depends on the extent of skin lesions and is fairly good unless there are large areas of sloughing [2].

Aural abscess management

Surgical debridement is the treatment; antibiotics alone will not resolve caseous chelonian exudate.

Stabilise first. In chelonians with long-standing abscesses or evidence of systemic disease, de la Navarre delays surgery and gives systemic antibiotics, fluids and supportive care for 3-4 d, which reduces local inflammation and intraoperative haemorrhage and allows correction of fluid and electrolyte imbalance and warming into the preferred optimal temperature zone; most presenting turtles are debilitated and respond to 2-3 d of supportive care before surgery [8].

Anaesthesia. The author reports good success with propofol at 10.0 mg/kg IV (extra-label) [8].

Technique. Sterile preparation matters especially if culture and sensitivity specimens are anticipated. The described approach is "a full-thickness horizontal incision through the entire tympanum from the 3 o'clock position across the center of the tympanum to the 9 o'clock position," then "a second vertical incision... from the 12 o'clock position down through the center of the tympanum to the 6 o'clock position" — together forming a cross [8]. Inflammatory debris is removed with small ear loops or forceps, ideally in one piece, taking care not to damage the columella (stapes) during debridement and flushing, and evaluating all aspects of the tympanic cavity — especially the caudal extent — because caseous material can extend far in chronic cases [8].

Lavage and closure. The tympanic cavity and eustachian tube are gently but liberally lavaged; de la Navarre prefers diluted chlorhexidine at 1 part chlorhexidine to 50 parts saline, noting that povidone-iodine remains controversial because in vitro concentrations in excess of 1% killed fibroblasts and free iodine is inactivated by serum proteins present postoperatively in the tympanic cavity [8]. Because debris and fluid can be flushed through the eustachian tube into the oropharynx, examine the oral cavity frequently and position the turtle head-down while flushing to let material drain by gravity [8]. The site is then filled with an appropriate antibiotic ointment and allowed to heal by secondary intention, with the veterinarian or client continuing daily lavage and packing until healed [8].

Culture and antimicrobials. Submit the caseous plug for culture and sensitivity. de la Navarre's own wording is conditional — "If desired, specimens for culture and sensitivity should be collected at this time" [8] — but the microbiology argues for doing it routinely: expect a mixed, opportunistic environmental flora rather than a single pathogen, and let sensitivity results, not empiricism, drive the systemic course [10]. The decision to use systemic antibiotics depends on physical and laboratory findings; if presurgical antibiotics were started, they are continued for a prescribed course [8].

Analgesia. Provide it. Appropriate anaesthesia is what allows thorough debridement and "alleviates the pain associated with the treatment of the disease process" [8]; extend cover post-operatively using the reptile analgesic options above [2].

Adjunctive ocular care. Cellular debris under the eyelids can be carefully removed with a moistened cotton-tip applicator and digital pressure, with an ophthalmic antibiotic ointment at clinician discretion [8].

Monitoring, prognosis, and prevention

Expected course. After treatment of the aural abscess and correction of husbandry, "Symptoms generally resolve gradually within 2-6wk depending on the severity of presenting clinical signs" [8]. Most aural abscesses respond well to surgical treatment and heal completely; recurrences are usually the result of either inadequate surgical debridement or failure to address the underlying predisposing causes [8]. Auditory compromise from the disease, the surgery and subsequent fibrosis is presumed but does not appear to have long-term effects on reptiles commonly treated for this condition [8].

Monitoring. Serial plasma retinol is now interpretable in the leopard gecko against a validated interval [14], and the Louth case series demonstrates a usable monitoring cadence — recheck days 2 and 7 after the first injection, then at 6 and 12 months [15]. In other species, monitor clinically: resolution of blepharoedema, return of appetite, normal ecdysis, and re-examination of the corrected diet.

Prevention. de la Navarre's prescription is species-appropriate environmental maintenance plus providing growing box turtles "with either a natural source of beta-carotene, or supplementing their diet with an oral source of preformed vitamin A in the form of a commercial product specifically formulated for the species" [8]. For insectivores, gut-load feeder insects on a diet containing vitamin A, "at least 8% calcium, multivitamins, trace minerals, proteins, carbohydrates, and fat" [3]. Mayer and Huang summarise the safest strategy as oral supplementation — gut-loading insects, dietary supplementation, and vitamin A-rich vegetables — noting that "Vitamin A from natural foods rarely leads to an overdose" [2]. The broader framing is worth repeating to owners: nutritional disorders of captive reptiles remain very common despite increasing knowledge of husbandry, are often diagnosed late, and "all nutritional disorders seen in captive reptiles are preventable" [17].

Frequently Asked Questions

What parenteral vitamin A dose should I give a box turtle with an aural abscess?

The most conservative published chelonian regimen is 1000-2000 IU vitamin A/kg parenterally, once a week for 2 wk in mild cases to 6 wk in severe cases, with the explicit caveat that "Higher doses of vitamin A do not speed recovery and can result in hypervitaminosis A" [8]. A reptile formulary entry gives 5000 IU/kg IM as a single treatment, especially in turtles [16], and the published "safe dose" band is 5000-10,000 IU/kg [2]. All of this is extra-label. Given that the toxic band starts at 50,000 IU/kg [2] — only five-fold above the top of the safe band — start low, correct the diet, and do not escalate for a slow clinical response.

How much vitamin A is toxic in a reptile?

"The safe dose for vitamin A administration is 5000-10,000 IU/kg, and a toxic dose is roughly 100 times higher (50,000 to 100,000 IU/kg)" [2]. The bracketed figures are ten-fold apart, so treat 50,000 IU/kg as the threshold of concern — and note that the margin you actually dose against is narrower still, since the top of the safe band (10,000 IU/kg) and the bottom of the toxic band (50,000 IU/kg) are only five-fold apart. Toxicity can also occur at a nominally safe injected dose if dietary vitamin A intake is elevated for a prolonged period [2].

Can I diagnose hypovitaminosis A from a blood sample?

Only partly, and only confidently in one species. A validated plasma retinol reference interval of 0.06 to 0.55 µg/mL exists for leopard geckos, derived from 40 healthy captive-bred animals on a 50 µl sample [14]. Outside that, serum levels vary significantly relative to hepatic levels and may not be practical for diagnosis [2], and definitive antemortem confirmation classically requires a hepatic vitamin A assay or a large volume of blood [8]. Boyer's position stands for most patients: diagnosis rests on dietary history and clinical signs [3].

Does an aural abscess in a box turtle prove vitamin A deficiency?

No. Tympanic epithelium in affected turtles does show hyperplasia, squamous metaplasia, cellular sloughing and granulomatous inflammation "consistent with a syndrome that may involve hypovitaminosis A" [9], and affected turtles have shown mean serum and hepatic vitamin A at 71% and 49% of unaffected turtles as a nonsignificant trend [5]. But hepatic vitamin A has also been found positively correlated with pathologic score (r = 0.32, P = 0.01) in 68 box turtles, "contrary to the expected result" [11], and dietary vitamin A restriction plus organochlorine exposure failed to induce tympanic squamous metaplasia experimentally in red-eared sliders over roughly six months [12]. Treat the abscess and the husbandry; do not use the abscess alone to justify aggressive parenteral dosing.

Should I culture an aural abscess, and can I choose an antibiotic empirically?

Culture it. de la Navarre's wording is conditional — "If desired, specimens for culture and sensitivity should be collected at this time" — with the practical tip that the caseous material is best removed in one piece [8]. The microbiology argues for doing it routinely: in 8 affected box turtles, 7 had growth yielding 10 isolates, and Morganella morganii (n = 2) was the only species recovered from more than one abscessed animal; no single agent was responsible, which the authors interpreted as aerobic bacteria acting as "secondary opportunistic invaders of environmental origin" [10]. That flora is too heterogeneous for reliable empirical selection, so let sensitivity results drive the systemic course [10].

Is gut-loaded beta-carotene enough, or must I give preformed vitamin A?

Species-dependent, and genuinely contested. Merck states that beta carotene "cannot be metabolized by many reptiles" [1], and carnivorous turtles including box turtles are described as unable to synthesise beta-carotene well [2]. But in leopard geckos, ten weeks of beta-carotene supplementation (gut-loaded insects plus weekly carrot juice) produced significantly higher hepatic vitamin A than vitamin A supplementation (p = 0.03) with no squamous metaplasia in any of the ten geckos [7]. Reasonable practice: carotenoid-based supplementation is supported in leopard geckos [7]; use preformed vitamin A in carnivorous chelonians [1][2].

A colleague has already overdosed a turtle with injectable vitamin A — what now?

Prior administration cannot be reversed, so management is supportive [2]. Decrease and potentially discontinue vitamin A intake; treat the skin lesions as burn wounds with hydrotherapy, antiseptic soaks, systemic antimicrobials and analgesia; give fluids at 10 to 30 mL/kg/day acutely; add nutritional support with or without a feeding tube [2]. Extra-label analgesic options include morphine at 1.5 mg/kg IM or SC in chelonians or 10 mg/kg IM or SC in lizards, or meloxicam at 0.2 to 0.3 mg/kg PO or IM q24h with careful monitoring for respiratory depression [2]. Set expectations early: healing can take 4 to 6 months, and prognosis is fairly good unless there are large areas of sloughing [2].

Which route is safest when supplementation really is indicated?

Oral. Merck advises that "Injectable vitamin A is best avoided, as hypervitaminosis A can cause skin erythema and sloughing" [1], and Mayer and Huang note that overdose risk "can be decreased by administering vitamin A in an oral, rather than injectable, formulation" [2]. If you must inject, a fat-soluble formulation is preferred over a water-soluble one, since water-soluble vitamin A is absorbed more quickly and is potentially more toxic [2]. In small chelonians in particular, a single oral dose may be more suitable than injection [2].