Type 1 diabetes in your twenties

The handover that does not stop in the clinic letter

The clinic letter that moves you from the paediatric team to the adult service is the formal handover. It is not the actual one. The actual handover is the year either side of it, and the work it carries is rarely written down. It is you booking the prescription request the first time the surgery emails to ask. It is you ringing the manufacturer when the pump alerts at 11pm and the only person in the household who knows the device sequence is asleep in another city. It is the first time you explain the CGM to a flatmate, and the first time the explanation lands without the family voice doing the work.

The literature on this transition is honest about what gets lost in it. Wentzell 2022 (Pediatric Diabetes) describes the structural drop-off that follows the handover for many young adults: HbA1c rises in the first year of adult care for a meaningful fraction of the cohort, the clinic attendance pattern thins, and the supports that used to be invisible (parental reminders, parental knowledge of the device, parental conversations with the school) leave a gap that is rarely named in the adult clinic letter. The ISPAD 2024 chapter on transitions (Ch 17) names the same pattern from the paediatric side and asks the adult service to expect it.

If you are reading this on the early side of the handover, the first reframe is that the drop-off pattern is structural, not personal. The second is that the adult clinic conversation goes differently if you ask for the foundations review explicitly: a TDD check, an ICR check, an ISF check, an IOB-readable review of the last fortnight on the device, a hypo-awareness conversation. The next four sections walk through those foundations in the shape they take in adult care.

The six foundations, the shape they take in adult care

The foundations themselves do not change between the paediatric and adult years. The shape of the conversation about them does. In adult care you are the one quoting the numbers, the one asking for the review, the one deciding which device-pattern question to bring this quarter. Naming the six in one place and the shape each one takes in the twenties is the work of this section.

If you do not currently know one of these six for yourself, the foundations cluster on GNL holds the deeper read on each. The reader who wants the algorithm-strength conversation next moves to the section below.

Algorithm strength, AIT, and the IOB-versus-strength trade-off

The AID conversation in the twenties is rarely about the device choice and almost always about the settings inside the device that the user has the most control over. The two settings that come up again and again, across CamAPS FX, MiniMed 780G, Tandem Control-IQ, and Omnipod 5, are the active insulin time (AIT) and the algorithm’s target dial. Naming what each does and how they interact is the work of this section.

Active insulin time is the figure the device uses to estimate how long ago a bolus is still active. A shorter AIT means the device counts a bolus as cleared faster; the algorithm sees less IOB on the screen and is more willing to micro-bolus on a rise. A longer AIT means the device counts the same bolus as still active for longer; the algorithm sees more IOB and is more conservative on the next correction. Neither is universally correct. A shorter AIT in a reader who already runs on the edge of overlapping boluses pushes the overnight pattern toward hypo. A longer AIT in a reader whose post-meal line consistently runs high pushes the same pattern toward sustained hyperglycaemia.

The target dial sits beside AIT in the same conversation. A lower target asks the algorithm to deliver more aggressively to pull the line down; a higher target asks it to deliver more conservatively to keep the line steady. The AID literature (and the AID Optimiser positioning that GNL has locked) frames the two settings together: the work is not on either one in isolation, but on the trade-off the two of them describe together.

The AID Optimiser on GNL holds the educational ladder for this conversation. The AID Optimiser is a Grade D educational synthesis on a Grade A and B evidence base, reviewed by manufacturer global medical leads at CamAPS, MiniMed, Tandem, and Insulet; it is never endorsed by, validated by, or co-developed with them. It is not a substitute for the conversation with your diabetes team. The vocabulary GNL uses for the trade-off, and the vocabulary the manufacturer documentation uses, is “shorter AIT” and “lower target”. The words “aggressive” and “responsive” are out of the canon; they collapse two distinct levers into one feeling.

The clinic conversation that follows from this framing has two stages. The first is reading the last fortnight of data on the device and asking which of the two patterns the algorithm is showing: stacked-correction hypos overnight, or sustained post-meal rises that the algorithm is not catching. The second is agreeing one setting change at a time, reading the data a fortnight later, and only then moving the second lever. The team owns the change; you own the noticing.

The single biggest mistake the early twenties make with AID

From the adult-clinic conversations I have, one pattern accounts for more frustration with AID in this age band than any other. The user takes a meal bolus. The line rises faster than the algorithm catches. The user takes a correction. The line keeps rising. The user takes another correction. By the time the algorithm has caught up, the IOB is high enough that the overnight tip the other way is severe; the algorithm goes into its safety mode for an hour or two; the morning waking reading is high because the algorithm has under-delivered through the back end of the night. The user concludes the algorithm is broken and the cycle repeats the next time the post-meal line runs away.

The honest read is that the algorithm is doing what it is designed to do, and the work is to wait. Every AID system on the UK market counts the IOB from your manual corrections against the next algorithm-driven micro-bolus. Stacking corrections feeds the algorithm a phantom signal that there is more glucose to bring down than there actually is; it then over-delivers and the overnight low follows.

The behavioural lever for breaking the pattern is small and unglamorous. Set the device to show the IOB number on the home screen rather than buried two taps deep. Agree a personal rule with the team for the minimum wait between a correction and the next consideration of another. Read the post-correction line in the app fifteen minutes later rather than five. The pattern is the same shape on every AID system; the fix is the same too.

First questions to bring to the adult clinic

The adult clinic review is fifteen to twenty minutes. The conversation goes further if you walk in with the questions already shaped. From the conversations I have with younger adults moving through the handover, the questions that move the review forward most reliably are not the broad ones (“am I doing well?”) but the specific ones that anchor a foundation or a settings conversation.

Bring the seven-day TDD and ask whether it is moving in a direction the team has noticed. Bring the post-meal pattern for the meal that is consistently running high or low and ask which lever (ICR, AIT, target) the team would consider first. Bring the last severe-hypo incident or the last week with the most lows and ask for an IAH score. Bring the device-foundations question you have not asked yet (where is IOB on this screen, how does this system handle exercise, what is the algorithm doing through the dawn window). Bring the prescription question that has been in the back of your head for a month.

If any of those questions does not get an answer in the allotted time, ask for it to be added to the notes for the next review. Adult care is a continuous conversation; the work of the twenties is keeping the conversation continuous when the household around it is shifting. The team is the route through every setting change on this page; the page is the route into the conversation, not the substitute for it.