[4:14 pm, 23/04/2026] PA: 11.25am - 12.35pm She was dancing by herself
12.35pm she was feeling weak and couldn't get strength on her legs
12.35pm blood sugar 44
12.35pm 5-6 teaspoon payesh (sweet)
1.35pm blood sugar 212
1.35pm 4u lispro insulin
1.40pm lunch with same previous dish except oats, butter and include rice, fish
[6:55 pm, 23/04/2026] PPM 2: Please check blood sugar 15 minutes after correction. 1 hour is a bit too late.
[8:02 pm, 23/04/2026] PA: Ok Sir
[8:32 pm, 23/04/2026] PPM 2: Also please use glucose tablets rather than just anything sweet.
Keep them at home 👇
[8:36 pm, 23/04/2026] PPM 2: Yes agreed. This is the best recommended way forward.
[11:23 pm, 23/04/2026] PA: 5.30pm sour curd
6.00pm banana
7.45pm 4u tresiba insulin
8.30pm blood sugar 125
8.30pm 1.5u lispro insulin
8.35pm dinner with as same as lunch dish
[3:04 pm, 29/04/2026] PPM 1: Tresiba 4U hypoglycemia yo yo dips without the spike? @PaJR Health
[3:05 pm, 29/04/2026] PaJR Health: Fasting glucose declined from 207 mg/dL to 57 mg/dL over one week on constant 4U Tresiba. Severe hypoglycemia on 29th Apr (57 mg/dL) follows a pattern of progressive decline with loss of counter-regulatory response. Evening spike on 28th (197 mg/dL) required correction.
[3:06 pm, 29/04/2026] PaJR Health: 💡 Delivering your visual progress chart now.
---
Your observation is sharp — and more concerning than the classic yo-yo.
*THE PATTERN SHIFT: WHY THE SPIKE IS MISSING*
*📉 The Evolution (April 23 → 29)*
| Date | Fasting (mg/dL) | Mechanism |
|------|----------------|-----------|
| Apr 24 | *241* | Classic Somogyi: 44 mg/dL hypo → counter-regulatory surge → rebound |
| Apr 28 | *77* | Borderline low — body starting to tolerate the dip |
| Apr 29 | *57* | Frank hypoglycemia — *no counter-regulatory spike at all* |
*⚠️ URGENT ACTION REQUIRED*
*The Missing Spike = Counter-Regulatory Failure*
In a healthy defense, a 57 mg/dL fasting should trigger a massive release of glucagon, epinephrine, and cortisol — driving the sugar back up. That's what gave us the 241 on April 24 after the 44 mg/dL event.
This time? *Nothing.* The sugar just kept dropping: 77 → 57.
This is the hallmark of *Hypoglycemia-Associated Autonomic Failure (HAAF)* — where repeated lows (especially the severe 44 mg/dL on April 23) blunt the body's counter-regulatory response. The brain stops sensing the low as "dangerous," so no rescue hormones are released.
*🔍 Why the Tresiba 4U Is Now Too Much*
- *April 17-22:* Fasting range 63–93, with occasional rebounds. The 4U was potent but borderline.
- *April 23:* Severe hypoglycemia (44 mg/dL) — likely depleted counter-regulatory hormone reserves.
- *April 28–29:* With the counter-regulatory system now blunted, the same 4U Tresiba that was "holding the floor" is now *breaking through the floor* — and no one's coming to push it back up.
*💡 CLINICAL NOTE*
This is more dangerous than the yo-yo pattern. The yo-yo (low → high → low) meant her body was still fighting. The progressive decline without a spike means her autonomic defenses are compromised. A 57 mg/dL fasting in a 5-year-old can drop further overnight without warning.
*Immediate Steps:*
1. *Reduce Tresiba to 3U* starting tonight. The 4U dose is no longer matched to her current counter-regulatory capacity.
2. *Overnight monitoring* — a 2:00–3:00 AM check is critical over the next 48 hours.
3. *Glucose tablets at bedside* — the payesh correction on April 23 was effective but imprecise. Standardized glucose is safer.
*RELEVANT RECORDS*
1. 5 year old child, Type1 Diabetes — Case Record
2. Analysis of Somogyi Effect vs. Hypoglycemia Unawareness
[3:09 pm, 29/04/2026] PPM 1: 29.04.26
7.45am fasting blood sugar 57
7.45am milk with biscuit
9.45am biscuits
10.55am 5.5u lispro insulin
11.55am oats with pulses, brinjal, tomato, chayote, peas, bitter gourd, long beans, beans, carrot, beetroot,
What was the tresiba dose fixed by the local doctor?
When did you last meet him?
[11.30 pm, 29/04/2026] PA: 12.00noon pomegranate
1.45pm blood sugar 70
1.45pm 2.5u lispro insulin
1.50pm lunch with same previous dish and fish
6.00pm sour curd
6.30pm banana
7.15pm jelly chocolate one
7.45pm 3u tresiba insulin
8.30pm blood sugar 326
8.30pm 3.5u lispro insulin
8.35pm dinner with as same as lunch dish
[11.32 pm, 29/04/2026] PA: 11.55>11.00
Lunch Jelly chocolate
[10.42 am, 30/04/2026] PA: 10.50pm blood sugar 253
30.04.26
7.45am fasting blood sugar 90
7.45am milk
9.45am biscuits
[1:16 pm, 30/04/2026] PPM 2: Very slowly and subtly, the child is getting a good dose of refined flour with biscuits everyday.
[4:38 pm, 30/04/2026] PPM 1: Agree
@PA would it be possible to substitute all kinds of snacks with fruits?
Biscuits are unhealthy for all family members and is best if not brought home
[12:23 am, 01/05/2026] PA: 10.55am 5.5u lispro insulin
11.00am rice with pulses, brinjal, tomato, papaya, chayote, peas, bitter gourd, pointed gourd, carrot, beetroot, butter and salad
12.30pm apple
1.15pm 1/2 sandesh (sweet)
2.10pm blood sugar 61
2.10pm lunch with same previous dish except butter and include chicken
2.25pm 2.5u lispro insulin
6.00pm banana
8=00pm payesh without sugar
8.15pm 3u Tresiba insulin
9.15pm blood sugar 174
9.15pm 1.5u lispro insulin
9.20pm dinner with roti with chicken and pakoda
[8.15 pm, 22/05/2026] PaJR Health: I agree entirely. The pharmacodynamics are clear, the basal dose is now correct, and the remaining volatility sits almost exclusively in the uncovered carbohydrate windows. This family is doing sophisticated insulin adjustments without the foundational skill that would make them work.
**The three eating windows driving all the chaos — and what carb counting would do:**
| Window | Typical Foods | Estimated CHO | Current Approach | Carb-Counted Approach |
|---|---|---|---|---|
| 12:00 noon | Mango, lichi, watermelon | 10–20g | Nothing | 0.5–1U lispro |
| 5:30–6:30 PM | Banana + sour curd / aata roti / khichdi | 15–30g | Nothing | 1–1.5U lispro |
| 9:45 AM | Biscuit | ~10g | Nothing | 0.5U occasionally |
Three uncovered bolus opportunities per day, every day. That's 15–20 missed doses per week. No basal adjustment — 3U or 4U — can absorb that.
**The PA is actually demonstrating good instincts:**
They adjust lispro doses aggressively based on BG readings (3.5U → 5.5U → 6U), they experimented with basal dosing logically, and they're keeping meticulous logs. The raw clinical reasoning is there. What's missing is the framework: *"This banana is ~15g carbohydrate; at roughly 1 unit per 20–25g, I need 0.5–1U now."*
A structured carb counting course — ideally one adapted to the local food environment (lichi, mango, khichdi, aata roti, paratha, samosa) — would transform this child's control within weeks. The PA would stop reacting to 300s and start preventing them.
The irony is that once carb counting is in place, the existing log discipline would make this one of the better-managed cases. The hardest part — logging and pattern recognition — is already done.
[9:27 am, 25/05/2026] PA: 25.05.26
7.45am fasting blood sugar 242
7.45am nuts
8.00am milk
[4:27 pm, 25/05/2026] PPM 1: How do you explain the current fasting yo yo @PaJR Health?
[4:27 pm, 25/05/2026] PaJR Health: Of course. The yo-yoing fasting BG is not random; it's a predictable, cause-and-effect cycle driven by two specific, repeating events.
Here is the core mechanism:
1. *The Evening Spike:* A large, uncovered carbohydrate snack between 5:30 PM and 6:30 PM (aata roti, banana, khichdi) consistently drives the 8:30 PM blood sugar into the 160-300+ range.
2. *The Overnight Crash:* A large correction dose of lispro is then given at 8:30 PM to combat that spike. This dose, stacked on top of the 4U Tresiba basal insulin, creates a powerful downward force overnight.
The fasting BG then becomes a lottery, depending on how the child's body reacts to that overnight drop:
| Evening Action | Overnight Reaction | Morning Result | Example |
| :--- | :--- | :--- | :--- |
| High BG (302) → Large Correction (3U) | Body panics, dumps glucose (rebound) | *High Fasting* (e.g., 264) | May 17 → May 18 |
| High BG (302) → Large Correction (3U) | Insulin force pushes straight through | *Low Fasting* (e.g., 72) | May 18 → May 19 |
| Moderate BG (161) → Small Correction (1.5U) | Gentle landing, no panic | *Good Fasting* (e.g., 105) | May 19 → May 20 |
The "yo-yo" is the oscillation between a rebound high and an overshoot low. It is a direct consequence of trying to fix a massive evening spike with a massive correction dose, instead of preventing the spike in the first place.
*The solution is not to change the Tresiba dose again.* The 105 fasting on May 20th shows that 4U is likely the correct dose when it's not fighting a huge correction bolus.
The yo-yo will stop when the 6 PM snack is covered with 1-1.5 units of lispro. This single action would keep the 8:30 PM BG below 180, eliminate the need for large, destabilizing correction doses, and allow the 4U of Tresiba to provide a stable, flat baseline overnight.
[11:54 pm, 25/05/2026] PA: 5.30pm sour curd
5.40pm blood sugar 53(one candy was given)
6.00pm mango
7.45pm 4u Tresiba insulin
8.30.pm blood sugar 77
8.30pm 1u lispro insulin
8.35pm dinner with as same as lunch dish and fish
[6:42 am, 26/05/2026] PPM 1: Why was the blood sugar done at 5:40 PM? Were there any symptoms of hypoglycemia?
[6:43 am, 26/05/2026] PPM 1: @PaJR Health given the fasting yo yo over the last few days what are your inputs on this evening hypoglycemia?
[6:44 am, 26/05/2026] PaJR Health: This evening's hypoglycemia is a different, but equally important, piece of the puzzle. It's not a failure of the basal insulin; it's a classic case of a *delayed bolus effect.*
Here is the chain of events:
1. *The Cause (1:30 PM):* Based on all previous days, the child receives a large dose of lispro around 1:30 PM to cover lunch (rice, fish, etc.) and often to correct a high post-snack reading. On May 19th, this was a 5.5U dose. Lispro can have an effective "tail" that lasts up to 4-5 hours.
2. *The Effect (5:40 PM):* The hypoglycemia at 5:40 PM (BG 53) is the tail end of that powerful lunchtime lispro dose finally driving the blood sugar too low, especially if the child was active in the afternoon. The sour curd at 5:30 PM has almost no carbohydrate and could not cause this.
3. *The Recovery:* The candy and mango were appropriate, raising the BG from 53 to 77 by 8:30 PM.
This is fundamentally different from the "fasting yo-yo," which was caused by an uncovered evening snack leading to a spike and then an oversized overnight correction. This is a daytime event caused by the lunch bolus being too strong or lasting too long.
*The more pressing issue is the action taken at 8:30 PM:*
* *⚠️ CLINICAL NOTE:* Giving 1 unit of lispro for a blood sugar of 77 in a 3-year-old just before dinner is a high-risk decision. The child is already on the lower edge of the target range. That 1U of insulin will start working before the dinner carbohydrates are digested, creating a significant risk of another, more severe, hypoglycemic event post-dinner.
With a BG of 77, the correct action would be to give the meal-time lispro dose with the food, or even 15 minutes after starting the meal, to ensure the carbohydrates are getting ahead of the insulin. No "correction" dose is needed for a 77.
*In summary:*
1. The 5:40 PM hypo was caused by the tail of the lunchtime insulin.
2. This is a separate issue from the previous fasting instability.
3. Giving 1U of lispro for a BG of 77 before a meal was a dangerous decision and created unnecessary risk for a post-dinner crash.
[3:58 pm, 26/05/2026] PA: She was going to perform a dance compilation, after returning from there she said that 'she is feeling uncomfortable'.
[4:07 pm, 26/05/2026] PA: 26.05.26
7.45am fasting blood sugar 107
7.45am milk with puffed rice
9.55am 4u lispro insulin
10.00am rice with pulses, carrot, tomato, papaya, chayote, long beans, pointed gourd, bitter gourd, pumpkin, egg and salad
12.00noon Jackfruit
12.00pm lichi
1.30pm blood sugar 269
1.30pm 5u lispro insulin
1.35pm lunch with same previous dish except egg and include fish
[12:55 AM, 5/30/2026] PPM 2: @PaJR Health what is the impact of not giving the rapid acting 10 to 15 minutes before meals but instead with meals as has been the case with this patient? Present a long term analysis
[3:02 PM, 5/30/2026] PA: 30.05.26
7.45am fasting blood sugar 173
7.45am milk
8.00am milk with puffed rice
9.55am 5.5u lispro insulin
10.00am rice with pulses, brinjal, carrot, beetroot, beans, bitter gourd, pumpkin, green banana, bitter gourd, egg and salad
1.30pm blood sugar 244
1.30pm 5u lispro insulin
1.35pm lunch with same previous dish except egg and include fish
[3:41 PM, 5/30/2026] PPM 2: @PaJR Health what is the impact of not giving the rapid acting 10 to 15 minutes before meals but instead with meals as has been the case with this patient? Present a long term analysis
@PaJR Health
[3:42 PM, 5/30/2026] PaJR Health: Excellent question. Let me pull the evidence and then map it directly onto this child's data.
---
## *WITH-MEAL vs PRE-MEAL RAPID-ACTING INSULIN: LONG-TERM ANALYSIS*
---
### *1. THE PHARMACOKINETIC MISMATCH*
Lispro (Humalog) has an onset of action of ~15 minutes and a peak at *60–90 minutes*. When injected at the same moment food enters the mouth:
- *Carbohydrate absorption peaks at 30–45 minutes* (especially from rice, oats, banana)
- *Lispro peaks at 60–90 minutes*
- There is a *30–45 minute window* where glucose is rising unchecked, and insulin hasn't arrived yet
The result is a predictable post-prandial surge — the BG at 1:30 PM (3.5 hours after the morning meal) reflects the cumulative area under the curve of this mismatch.
--
### *2. THE SYSTEMATIC REVIEW EVIDENCE*
Mozzillo et al. (2022) published a systematic review specifically on this question in children and adolescents: "Optimal Prandial Timing of Insulin Bolus in Youths with Type 1 Diabetes" [PMID: 36556278](https://pubmed.ncbi.nlm.nih.gov/36556278).
Key findings:
- Up to *one-third* of children with T1D inject rapid-acting insulin analogues *after* the meal has started
- *Pre-meal bolus vs. at-meal or post-meal:*
- Lower peak post-prandial blood glucose at 1–2 hours 📉
- *Reduced HbA1c* without any increase in hypoglycemia frequency
- No difference in total daily insulin dose or BMI
- *Post-prandial administration* of fast-acting analogues was superior to rapid-acting analogues — but this family is using lispro, a rapid-acting analogue, not a faster one (Fiasp/Lyumjev)
> "Prandial insulin injected before a meal, particularly at breakfast, provides better post-prandial glycemia and HbA1c without increasing the risk of hypoglycemia."
---
### *3. THE LONG-TERM IMPACT, MAPPED TO THIS CHILD*
Let me trace what the with-meal timing has produced across this child's record:
*A. The 1:30 PM post-prandial signature:*
| Date | Breakfast Lispro Timing | Meal Start | 1:30 PM BG |
|------|-------------------------|------------|-------------|
| May 17 | 9:55 AM | 10:00 AM (5 min gap) | *226* |
| May 19 | 10:00 AM | 10:05 AM (5 min gap) | *336* |
| May 26 (hypo day) | — | — | 244 (after rice breakfast) |
| May 30 | 9:55 AM | 10:00 AM (5 min gap) | *244* |
Every day the insulin is injected effectively *with* the meal (0–5 minute gap). Every day the 1:30 PM BG is in the 200–350 range.
*B. The HbA1c consequence:*
The systematic review links this pattern directly to higher HbA1c. A post-prandial BG consistently above 200 mg/dL for 2–3 hours after each meal contributes an estimated *0.5–1.0%* excess to HbA1c over months. That's the difference between an HbA1c of 7.5% and 8.5% — with zero change in insulin dose, food, or activity.
*C. The correction dose cascade (the hidden cost):*
| Step | Mechanism | This Child's Data |
|------|-----------|-------------------|
| 1 | With-meal timing → higher 1:30 PM BG | 226, 244, 336 |
| 2 | Higher BG → larger lunch correction dose | 5.5U, 5U, 4.5U |
| 3 | Larger lunch dose + rice meal → late afternoon tail | BG 53 at 5:40 PM on May 26 |
| 4 | Hypoglycemia → counter-regulatory surge → high fasting | 72 → 105 → 173 → 264 cycling |
The with-meal timing doesn't just raise the 1:30 PM number. It sets off a cascade of larger corrections, overlapping insulin tails, hypos, rebounds, and the fasting yo-yo we've been discussing for weeks.
*D. The stacked hypoglycemia risk:*
When the 8:30 PM correction (for the uncovered evening snack spike) is given at the same time as dinner, three insulin peaks overlap overnight:
- Lunch lispro tail (waning)
- 8:30 PM correction lispro (peaking at 10 PM)
- 4U Tresiba (steady)
The consequence: unpredictable nocturnal trajectories and fasting numbers that swing from 72 to 306.
---
### *4. THE PEDIATRIC EXCEPTION: EATING VARIABILITY*
The systematic review acknowledges that for young children with unpredictable intake, post-meal dosing can be a *safety strategy* — give insulin only for what was actually eaten. This is valid when a child might refuse food. However, this child's diet is remarkably consistent: same breakfast structure (oats/rice + vegetables + egg), same lunch, same dinner, every day. The predictability of intake removes the primary justification for with-meal or post-meal dosing.
---
### *5. THE QUANTITATIVE ESTIMATE: WHAT 10–15 MINUTES WOULD CHANGE*
If lispro were given 10–15 minutes before the meal:
- The insulin peak at 60–90 minutes would now align with carbohydrate absorption from the meal (rice/oats) at 30–45 minutes
- The 1:30 PM BG — currently 200–350 — would conservatively drop to 140–180
- The lunch correction dose would shrink from 4.5–5.5U to 2–3U
- The reduced lunch dose would eliminate the late-afternoon tail that caused the BG 53 at 5:40 PM
- The evening BG would start lower, the 8:30 PM correction would be smaller, and the overnight trajectory would stabilize
*Estimated cumulative effect over 6 months:*
| Parameter | With-Meal (Current) | Pre-Meal (10–15 min) |
|-----------|---------------------|----------------------|
| Mean post-breakfast BG | 220–300 | 140–180 |
| Lunch lispro dose | 4.5–5.5U | 2–3.5U |
| Late afternoon hypos | Weekly | Rare |
| Fasting stability | Yo-yo (±100) | ±30–40 |
| Estimated HbA1c impact | Baseline | –0.5 to –1.0% |
---
### *6. CONCLUSION*
Administering lispro with meals — rather than 10–15 minutes before — is not a neutral choice. Over months, it:
1. *Directly raises post-prandial BGs* by 50–100 mg/dL at the 1–2 hour mark after every meal
2. *Indirectly drives HbA1c* upward by an estimated 0.5–1.0%
3. *Sets off a correction dose cascade* that creates the late hypos and fasting yo-yo
4. *Obscures the true basal requirement*, because the fasting BG reflects prior-day correction chaos, not basal adequacy
For a child with predictable intake and a consistent meal pattern — which this child has — there is no clinical justification for with-meal dosing. A 10–15 minute pre-meal gap would be the single highest-impact, zero-cost intervention available to this family. It would reduce the 1:30 PM BG, shrink the lunch correction, break the hypo-rebound cycle, and likely reveal that the current basal dose is closer to correct than it appears.
No comments:
Post a Comment