Canine Diabetes Mellitus: Insulin Therapy Adjustment Framework

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-08-11

Key Takeaways

• Insulin therapy adjustment in canine diabetes mellitus is an iterative process requiring interpretation of serial blood glucose curves, clinical signs, and insulin pharmacokinetics, with no single curve being diagnostic in isolation.
• Key parameters for dose adjustment include the nadir glucose (critical for hypoglycemia risk), glucose range across the curve (reflecting duration and adequacy), and clinical signs (PU/PD, appetite, weight) as the ultimate ground truth for metabolic control.
• The Somogyi effect, characterized by rebound hyperglycemia following hypoglycemia, must be suspected when a curve shows a low nadir followed by a marked rise, necessitating dose reduction rather than increase.
• Insulin glargine and detemir offer longer durations of action than traditional NPH, but individual responses vary significantly, making product selection and potential switching a critical component of therapy refinement.
• Hypoglycemia, defined as blood glucose below 60 mg/dL with signs or below 50 mg/dL regardless of signs, is the primary acute complication and mandates immediate dose reduction by 25% to 50%.
• Troubleshooting non-responding patients requires systematic investigation of insulin handling/administration errors, dietary inconsistencies, concurrent diseases (e.g., hyperadrenocorticism, hypothyroidism, infection), and true insulin resistance, with doses exceeding 1.5 U/kg per injection warranting further diagnostics.

This article provides a structured framework for adjusting insulin therapy in diabetic dogs once the initial diagnosis has been established and stabilization has begun. It is written for practicing veterinarians who manage canine diabetics in clinical settings and need a repeatable, evidence-informed method for interpreting glucose data and titrating doses. The framework addresses blood glucose curve interpretation, dose adjustment algorithms, insulin product selection between glargine and detemir, and recognition and management of hypoglycemia. The goal is to convert serial glucose measurements and clinical observations into defensible therapeutic decisions.

Insulin adjustment in canine diabetes is iterative. Each dose change generates new data that must be interpreted in the context of the previous curve, the dog's clinical signs, and the pharmacokinetic profile of the chosen insulin. No single curve is diagnostic in isolation. The framework presented here assumes the reader has confirmed hyperglycemia with concurrent glucosuria or consistent clinical signs, ruled out or managed concurrent disease, and initiated insulin therapy. What follows is the decision architecture for refining that therapy.

At a Glance

Physiologic Basis for Insulin Adjustment

The therapeutic objective in canine diabetes is to mimic physiologic insulin secretion closely enough to control clinical signs and minimize glucose toxicity without inducing hypoglycemia. Endogenous insulin secretion in healthy dogs is pulsatile and meal-responsive. Exogenous insulin cannot replicate this precisely, so the clinician aims for a practical approximation: a glucose curve that stays within an acceptable range for most of a 12 to 24 hour period, with a nadir that does not threaten neurologic function.

Glucose toxicity complicates early adjustment. Persistent hyperglycemia impairs residual beta cell function and reduces peripheral insulin sensitivity, creating a self-perpetuating cycle. As glycemic control improves, insulin requirements may fall. This means that a dose that was appropriate at stabilization may become excessive weeks later. The framework must therefore include scheduled reassessment, also reactive adjustment.

Counterregulatory responses also shape curve interpretation. Hypoglycemia triggers glucagon, epinephrine, cortisol, and growth hormone release. The resulting hyperglycemia can appear hours after the actual hypoglycemic event, producing a curve that looks like insulin underdosing when the real problem is relative overdose. This phenomenon, the Somogyi effect, is a recognized failure mode in canine insulin adjustment and must be considered before increasing any dose.

Insulin Pharmacology in Dogs

Pharmacokinetic Differences Between Insulin Types

Neutral protamine Hagedorn (NPH) insulin has historically been the first-line choice in canine diabetes due to cost and availability. Its peak effect occurs roughly 4 to 8 hours after injection in dogs, with a duration of action of 8 to 14 hours. This requires twice-daily dosing in most dogs and produces a glucose curve with a discernible nadir.

Long-acting analogs differ meaningfully. Insulin glargine forms microprecipitates at physiologic pH, slowing absorption and producing a relatively flat time-action profile in humans. In dogs, glargine has a duration of action that can approach 12 to 24 hours, though individual variation is substantial. Insulin detemir is acylated to bind albumin, which delays absorption and prolongs action. Published canine studies show detemir has a duration of action similar to or slightly shorter than glargine, with a more pronounced peak in some dogs.

The choice between glargine and detemir is not settled by a single definitive trial. Current guidance from the American College of Veterinary Internal Medicine consensus statements acknowledges that both can be effective in dogs but that individual responses vary. Some dogs show better control with one product than the other, and switching products is a reasonable step when a dog fails to respond adequately to the initial choice. The MSD Veterinary Manual similarly notes that insulin selection should be individualized and that response, not product label, determines success.

Duration of Action and Curve Timing

The expected duration of action dictates when glucose measurements should be taken. For twice-daily NPH, a 12 hour curve with measurements every 2 hours captures the peak and nadir. For glargine or detemir, a 24 hour curve may be necessary to document duration of action, particularly if the dog shows pre-injection hyperglycemia that suggests the insulin is not lasting the full interval.

A common error is measuring glucose only at the expected nadir. This misses duration problems. If the dog is hyperglycemic before the next injection, the insulin is not covering the full interval, regardless of what the nadir shows. The curve must bracket the entire dosing interval to support a duration decision.

Blood Glucose Curve Interpretation

Curve Components and What They Mean

A blood glucose curve provides four pieces of information: the starting glucose, the nadir, the time to nadir, and the glucose range across the interval. Each component supports a different adjustment decision.

The starting glucose reflects the residual effect of the previous dose and the fasting state. A high starting glucose with a low nadir suggests the insulin is potent but short-acting. A high starting glucose with a high nadir suggests underdosing. The time to nadir indicates the absorption and peak profile of the insulin in that individual dog. A delayed nadir may mean the insulin is absorbed slowly, which can be normal for glargine but problematic for NPH if it pushes the peak beyond the expected window.

The glucose range, defined as the difference between the highest and lowest readings, captures the overall stability of glycemic control. A narrow range with a high nadir indicates the insulin is having little effect. A wide range with a low nadir indicates excessive insulin effect at peak with inadequate duration. Both patterns require different corrective actions.

Target Ranges and Thresholds

The ACVIM consensus statements provide the most widely referenced target ranges for canine diabetics. The goal is to maintain blood glucose between roughly 100 and 250 mg/dL for most of the day, with a nadir that does not fall below approximately 80 to 100 mg/dL. These are clinical targets, not absolute thresholds. Some dogs show resolution of clinical signs with glucose values above this range, and forcing lower values increases hypoglycemia risk without added benefit.

The nadir is the critical safety parameter. If the nadir falls below 80 mg/dL, the dose must be reduced regardless of how high the rest of the curve sits. Hypoglycemia is the only acute life-threatening complication of insulin therapy, and no degree of glycemic improvement justifies risking it.

Dose Titration: The First 30 Days

The initial adjustment period begins after the starting dose has been in place for 7 to 10 days. Do not alter the dose before this point unless hypoglycemia occurs. The first recheck should include a full blood glucose curve, body weight, and owner-reported clinical signs. The goal of the first adjustment is not perfect regulation. It is to establish the direction and magnitude of change needed while avoiding hypoglycemia.

The decision to increase, decrease, or hold the dose rests on three inputs: the lowest glucose value on the curve, the duration of insulin effect, and the clinical signs reported by the owner. These three inputs are weighted differently depending on where the patient sits in the treatment timeline. Early in therapy, the lowest glucose value dominates the decision. Later, duration of effect and clinical stability become more important.

Decision Rules for Dose Adjustment

The following framework applies to intermediate-acting insulins such as NPH or lente preparations. For each rule, the glucose values refer to blood glucose measured by a validated glucometer or laboratory analyzer.

These thresholds align with the monitoring parameters described in the ACVIM consensus statements on diabetes management. The percentages are starting points. Individual patients may require larger or smaller increments depending on their response and their proximity to the hypoglycemic range.

The 10% Rule and Its Limits

A common teaching is to adjust insulin by no more than 10% to 25% per change. This rule protects against hypoglycemia from cumulative dose increases. It has limits. A dog receiving 0.5 U/kg twice daily that needs a 20% increase will move to 0.6 U/kg. A dog receiving 1.5 U/kg twice daily that needs a 20% increase will move to 1.8 U/kg. The latter change is larger in absolute units and carries more risk.

When the current dose is high, consider smaller percentage increases. When the current dose is low, the full percentage increase is usually safe. Always recalculate the dose per kilogram after each change. Weight loss during the adjustment period will increase the per-kilogram dose even if the total units stay the same.

Troubleshooting the Non-Responding Patient

A dog that remains hyperglycemic despite dose increases requires a structured investigation. The most common causes, in order of frequency, are insulin administration failure, dietary inconsistency, concurrent disease, and true insulin resistance.

Insulin Handling and Administration Errors

Before any dose increase beyond 1.0 U/kg per injection, verify the mechanics of administration. Observe the owner drawing up and injecting the insulin. Common errors include using the wrong syringe, failing to mix the suspension, injecting into a skin fold instead of subcutaneously, and rotating sites inconsistently. The MSD Veterinary Manual emphasizes that client education on insulin handling is a recurring need, not a one-time event.

Check the insulin bottle for crystallization, clumping, or expiration. Verify that the insulin is stored refrigerated and not frozen. Confirm that the concentration of the insulin matches the syringe. U-40 insulin with U-100 syringes produces a fivefold dosing error.

Concurrent Disease and Insulin Resistance

If administration is correct and the dose exceeds 1.5 U/kg per injection without adequate response, investigate concurrent disease. Dental disease, urinary tract infection, pancreatitis, hyperadrenocorticism, hypothyroidism, and neoplasia all cause insulin resistance. A minimum database should include a complete blood count, serum biochemistry, urinalysis with culture, and imaging as indicated.

The distinction between inadequate dosing and true insulin resistance matters for the adjustment plan. Inadequate dosing responds to continued increases. Insulin resistance requires treatment of the underlying condition. The two can coexist. A dog with dental disease may also be underdosed. Treat the concurrent disease first, then reassess the insulin requirement.

Hypoglycemia: Recognition and Response

Hypoglycemia is the limiting complication of insulin therapy. The threshold for intervention is a blood glucose below 60 mg/dL (3.3 mmol/L) with clinical signs, or below 50 mg/dL (2.8 mmol/L) regardless of signs. Clinical signs include lethargy, weakness, ataxia, tremors, seizures, and collapse. Some dogs show no signs until glucose falls very low.

When hypoglycemia is confirmed, the insulin dose should be reduced by 25% to 50%. The dog should be rechecked with a blood glucose curve within 3 to 5 days. Do not simply hold the next dose and resume at the previous dose. The dose that caused hypoglycemia will cause it again.

The Somogyi Effect

The Somogyi effect describes hypoglycemia followed by rebound hyperglycemia. It occurs when an excessive insulin dose drives glucose down, triggering counter-regulatory hormone release that pushes glucose back up. The blood glucose curve shows a low point followed by a steep rise to high values. The owner may report polyuria and polydipsia that worsen after dose increases.

The Somogyi effect is less common in dogs than in cats, but it occurs. When a curve shows a nadir below 80 mg/dL (4.4 mmol/L) followed by glucose above 250 mg/dL (13.9 mmol/L) within 2 to 4 hours, suspect rebound. The correct response is to decrease the dose, not increase it. Increasing the dose in response to the rebound hyperglycemia worsens the cycle.

Insulin Type Selection and Switching

Most dogs are started on an intermediate-acting insulin such as NPH or a lente preparation. When response is poor or duration is short, switching to a longer-acting insulin may help. The choice between insulin types depends on the curve findings and the owner's ability to manage twice-daily dosing.

The duration ranges in this table are clinical generalizations. Individual variation is substantial. When switching insulins, reduce the starting dose by 25% to 50% and perform a blood glucose curve within 5 to 7 days. The ACVIM consensus statements note that the pharmacokinetic differences between insulin types justify different starting doses, but the clinical response must guide subsequent adjustments.

When to Switch

Switch insulin types when the current insulin cannot achieve a duration of effect that covers the inter-dose interval without causing hypoglycemia. A dog on NPH twice daily that shows glucose rising above 300 mg/dL (16.7 mmol/L) for 4 or more hours before each injection is a candidate for a longer-acting insulin. A dog that develops hypoglycemia at the nadir but remains hyperglycemic at the end of the interval has a duration problem, not a dose problem.

Do not switch insulins in response to a single poor curve. Confirm the pattern with a second curve. Ensure that administration technique and diet are consistent before attributing the problem to the insulin type.

Documentation and Recheck Scheduling

Every dose adjustment should be documented with the date, the dose in units and per kilogram, the curve values, and the clinical signs. This documentation supports the next adjustment decision. Without it, each recheck starts from incomplete information.

The recheck schedule follows a predictable pattern. After each dose change, recheck in 7 to 10 days. Once a stable dose is achieved, recheck in 3 to 4 weeks. Stable dogs on a fixed dose can be rechecked every 3 to 6 months with a curve and physical examination. Owners should be encouraged to keep a log of water consumption, urine output, appetite, and body weight between visits.

Fructosamine measurement provides a 2 to 3 week average of glucose control. It is useful when a full curve is not possible or when the curve is difficult to interpret. Fructosamine does not replace the curve for dose adjustment decisions. It confirms overall control but does not identify the nadir or the duration of effect.

Recognized Complications and Early Detection

The principal treatment-related complications in canine diabetes are hypoglycemia, insulin resistance, and diabetic ketoacidosis (DKA). Hypoglycemia remains the most common and most dangerous complication of insulin therapy. Early detection depends on owner recognition of clinical signs, including lethargy, ataxia, weakness, tremors, and altered mentation, combined with periodic blood glucose measurement. Any glucose reading below 3.3 mmol/L (60 mg/dL) in a dog receiving insulin warrants immediate dose reduction regardless of clinical signs. Subclinical hypoglycemia, defined as glucose values between 3.3 and 4.4 mmol/L (60 to 80 mg/dL) without clinical signs, should prompt a 10 to 20% dose reduction at the next scheduled administration.

Insulin resistance is defined as a requirement exceeding 1.5 U/kg per injection in dogs. Early detection relies on serial glucose curves that show persistently elevated nadirs above 11.1 mmol/L (200 mg/dL) despite escalating doses. The clinician should suspect resistance when dose increases produce no measurable change in curve shape or nadir over two consecutive rechecks. Underlying causes include hyperadrenocorticism, hypothyroidism, diestrus in intact bitches, infection, pancreatitis, and exogenous glucocorticoid administration. The ACVIM consensus statements provide structured guidance on diagnostic investigation of suspected insulin resistance, including recommended endocrine testing and imaging protocols ACVIM consensus statements on endocrine disease.

DKA develops when absolute or relative insulin deficiency permits uncontrolled lipolysis and ketogenesis. Early warning signs include progressive polyuria, polydipsia, vomiting, and anorexia in a previously regulated diabetic. Point-of-care beta-hydroxybutyrate measurement is more sensitive than urine ketone dipsticks and should be performed in any diabetic dog presenting with vomiting or lethargy. The MSD Veterinary Manual describes the clinical progression of DKA and emphasizes that early intervention with fluid therapy and short-acting insulin improves outcomes MSD Veterinary Manual guidance on diabetic ketoacidosis.

Common Clinical Errors and Corrective Actions

A common error among less experienced clinicians is adjusting insulin based on a single glucose reading instead of the complete curve. A single pre-insulin value provides no information about nadir timing or duration of action. Corrective action is to obtain a full 12-hour curve before any dose change, except when hypoglycemia is documented or strongly suspected.

Another frequent mistake is increasing the dose when the pre-insulin glucose is high without considering the nadir. If the nadir is already within target range, increasing the dose will cause hypoglycemia. The dose should be adjusted based on the nadir, not the peak. Clinicians also commonly misinterpret the Somogyi effect as insulin resistance and increase the dose, worsening the rebound hyperglycemia. The discriminating feature is that the Somogyi effect produces a characteriztic curve with a low nadir followed by a steep rise, whereas true insulin resistance produces a flat, elevated curve throughout.

Limitations of Current Evidence

The evidence base for canine insulin adjustment relies heavily on expert opinion and extrapolation from human medicine instead of large randomised controlled trials. Direct comparative studies of insulin types in dogs remain limited, and the pharmacokinetic data that exist come from small populations with considerable inter-individual variation. The ACVIM consensus statements acknowledge that many recommendations, including specific target glucose ranges and titration intervals, derive from clinical experience instead of prospective trials ACVIM consensus statements on diabetes management.

Expert opinion still differs on several points. The optimal target nadir range varies between authorities, with some recommending 5.6 to 8.3 mmol/L (100 to 150 mg/dL) and others accepting up to 11.1 mmol/L (200 mg/dL) to reduce hypoglycemia risk. The frequency of glucose curve monitoring during stabilization is similarly contested, with recommendations ranging from every 3 to every 14 days. The role of continuous glucose monitoring in routine canine diabetic management remains an area of active investigation, and published data on its accuracy in dogs are limited.

Referral and Escalation Criteria

Referral to a specialist is warranted when insulin resistance persists despite doses above 1.5 U/kg per injection, when the clinician cannot identify an underlying concurrent disease, or when repeated hypoglycemic episodes occur despite appropriate dose reduction. Specialist consultation is also appropriate for dogs requiring more than two insulin type changes without clinical improvement, or when the owner cannot achieve reliable glucose monitoring at home.

Laboratory involvement is indicated for suspected concurrent endocrinopathy, particularly hyperadrenocorticism or hypothyroidism, where dynamic testing such as low-dose dexamethasone suppression or thyroid stimulation may be required. The AVMA practice resources outline laboratory quality assurance standards that support reliable diagnostic testing in practice settings AVMA laboratory and diagnostic resources.

Regulatory reporting is rarely required in canine diabetes management. However, if a veterinary team identifies a suspected adverse drug reaction to a licensed insulin product, reporting to the relevant national pharmacovigilance scheme is appropriate. The World Organization for Animal Health terrestrial standards address reporting obligations for notifiable diseases, but diabetes mellitus is not a notifiable condition in most jurisdictions WOAH terrestrial animal health standards. Clinicians should consult their local regulatory authority for specific adverse event reporting requirements.

Frequently Asked Questions

How Should I Adjust Insulin When the Owner Cannot Perform a Full Blood Glucose Curve at Home?

When a full curve is not feasible, rely on serial single glucose measurements paired with clinical signs. Measure blood glucose at the expected time of peak insulin effect, typically 6 to 8 hours after injection for intermediate-acting insulins, and again immediately before the next dose. The pre-injection value reflects trough glucose and the peak value reflects maximal insulin effect. If the owner can obtain only one reading, the pre-injection sample is more useful for safety, as it identifies the lowest point of the dosing cycle. Adjust doses by 10% based on trends across three consecutive days of single readings, and always prioritize owner-reported signs of hypoglycemia over any single glucose number. The MSD Veterinary Manual provides guidance on practical monitoring strategies when intensive sampling is unavailable.

What Is the Safest Approach When a Dog Presents With Severe Hyperglycemia but No Ketosis?

Hospitalization for curve generation is indicated when glucose exceeds 500 mg/dL or when clinical signs are severe. Administer the usual insulin dose at the usual time, then generate a curve every 2 hours for 12 hours. Do not administer additional insulin based on a single high reading, as this risks stacking and subsequent hypoglycemia. If the dog is eating and drinking, maintain the current dose and reassess after 24 hours. If glucose remains above 400 mg/dL throughout the curve, rule out injection errors, insulin inactivation, and concurrent disease before increasing the dose. The ACVIM consensus statements address the importance of identifying underlying causes of insulin resistance before escalating therapy.

How Do I Manage Insulin Adjustment in a Dog With Concurrent Glucocorticoid Therapy?

Glucocorticoids induce insulin resistance through hepatic gluconeogenesis and reduced peripheral glucose uptake. Expect insulin requirements to rise by 25% to 50% during treatment, and anticipate that requirements will fall again after discontinuation. Recheck glucose curves 5 to 7 days after starting or stopping glucocorticoids. Increase the dose by 10% to 20% if the curve shows persistent hyperglycemia, but never increase more frequently than every 5 to 7 days. When glucocorticoids are tapered, reduce insulin proactively to avoid hypoglycemia. The MSD Veterinary Manual discusses drug interactions that affect insulin sensitivity in diabetic dogs.

What Should I Do When the Owner Reports Polyuria and Polydipsia but the Glucose Curve Appears Acceptable?

A curve that shows acceptable glucose values does not exclude poor glycaemic control if clinical signs persist. First confirm the curve was performed correctly, as stress hyperglycemia during hospitalization can mask a poorly regulated state. Ask the owner to measure urine glucose and ketones at home, and correlate those findings with water intake. If the curve is genuinely acceptable but clinical signs continue, investigate concurrent disease, particularly urinary tract infection, hyperadrenocorticism, or exocrine pancreatic insufficiency. Recheck the curve within 7 days instead of assuming the current dose is correct. The ACVIM consensus statements emphasize that clinical response, not glucose numbers alone, defines successful regulation.

How Should I Advise Owners Who Cannot Afford Frequent Glucose Curve Assessments?

Cost constraints are common and should be addressed directly. A single pre-injection glucose measurement performed every 3 to 7 days, combined with careful monitoring of water intake, urine output, and body weight, provides adequate data for safe titration in most dogs. Ask owners to keep a simple log of daily water consumption measured in cups or milliliters, and to record any episodes of weakness, lethargy, or collapse. Use these clinical endpoints as the primary titration guide, reserving full curves for cases where the response is ambiguous or complications are suspected. The AVMA practice resources offer guidance on communicating cost-effective monitoring strategies to clients while maintaining safety standards.

How Do I Explain the Difference Between Detemir and Glargine to a Client When Cost Is a Factor?

Explain that both insulins are long-acting analogues, but they differ in duration and consistency. Detemir often requires twice-daily dosing in dogs and may provide more stable control in some individuals, while glargine has a longer duration that suits once-daily protocols in certain patients. Cost differences are real, but the more expensive product is not automatically better. The deciding factors are the glucose curve shape, the dog's response, and the owner's ability to administer injections consistently. If cost forces a switch, transition at the same dose, then recheck a curve within 5 to 7 days. The MSD Veterinary Manual compares insulin formulations and supports product selection based on individual patient response instead of price alone.

Related Clinical & Scientific Guides

• Feline Hepatic Lipidosis: Nutritional and Medical Management
• Canine Respiratory Infection: Diagnostic Approach and Treatment
• Canine Respiratory Virus: Diagnostic and Management Considerations

References and Further Reading

• NH2-terminal pro-brain natriuretic peptide and risk of diabetes.. 2013.
• Bone health in diabetes: considerations for clinical management.. 2009.
• Effect of oestrogen plus progestin on the incidence of diabetes in postmenopausal women: results from the Women's Health Initiative Hormone Trial.. 2004.
• Plasma copeptin and the risk of diabetes mellitus.. 2010.
• Adiposity is a major determinant of plasma levels of the novel vasodilator hydrogen sulphide.. 2010.
• Effects of carnosine supplementation on glucose metabolism: Pilot clinical trial.. 2016.
• ACVIM Consensus Statements. Journal of Veterinary Internal Medicine.
• MSD Veterinary Manual, Professional Edition. MSD Veterinary Manual.
• American Veterinary Medical Association Practice Resources. American Veterinary Medical Association.

Related Articles

• Canine and Feline Chronic Diarrhea: Diagnostic and Therapeutic Framework
• Feline Diabetes Mellitus: Remission Induction and Monitoring
• Canine Autoimmune Disorders: Clinical Recognition and Diagnostic Framework
• Canine Immune-Mediated Arthropathy: Diagnostic and Management Framework
• Canine Immune-Mediated Hemolytic Anemia: Diagnostic and Management Framework

This article is educational professional reference material for veterinary audiences. It is not a substitute for veterinary diagnosis, individual clinical judgment, current product labeling, or applicable regulatory requirements.