School-age children with T1D (ages 7 to 14)

From carer-managed to child-managed, slowly

For families I see in the paediatric clinic at Birmingham Women’s and Children’s, the years from 7 to 14 are where the work of T1D starts to migrate. The parent still holds most of the cognitive load, the dose maths and the sick-day rules and the school relationship, but the child begins to do the doing. The first solo finger prick. The first lunch bolus the parent did not see. The first sleepover. None of this happens on a single day. It happens in small, uneven steps over several years, and the unevenness is the point.

What the evidence supports here is gradual responsibility transfer with the parent staying involved in the decision-making. International paediatric consensus (Cengiz and colleagues, ISPAD 2024, Hormone Research in Paediatrics) locks intensive basal-bolus therapy, multiple daily injections or pump, as the gold standard across every paediatric age band, with the regimen tuned by the diabetes team rather than left to develop on its own. The technology layer sits on top of that. The growing autonomy sits on top of both.

The thing the textbooks underplay is the parent’s grief in this window. The dose was yours; now it is shared. The kitchen was the world; now the school canteen is. Naming that out loud, to your team and to the child, tends to be more useful than carrying it alone.

The pump conversation, what changed in 2017

For two decades the question “pump or pens” sat as a preference call, with families and teams reading each other across the desk. That changed when Karges and colleagues published the largest contemporary comparison we have. In a German, Austrian and Luxembourgish registry of children, adolescents and young adults with T1D, those on pumps had lower rates of severe hypoglycaemia (incidence rate ratio 0.68), lower rates of severe diabetic ketoacidosis (IRR 0.61) and a small HbA1c advantage (Karges and colleagues 2017, JAMA). The cohort was 30,579 strong; the propensity-matched analysis paired the pump and pen groups on every important variable the registry recorded.

The detail that matters for this guide is the age stratification. The severe-hypoglycaemia advantage of the pump was present in school-age children and adolescents, but not in the 1.5 to 5 year preschool subgroup, where event rates were similar between pump and injection therapy. The pump did not, on its own, fix the youngest band. That is part of why preschool care has moved on to AID, where the algorithm reaches first. For ages 7 to 14, the pump on its own does change the hypo and DKA picture, and that is what shifted the clinical conversation.

For a family deciding now, the practical truth is that the pump is one route into the wider technology stack, and the conversation is with your paediatric diabetes team. The Karges figures are the reason the conversation is not a preference call any more.

Paediatric AID, what the pivotal trials show

The next layer up is automated insulin delivery, where the CGM reads the glucose, the algorithm decides, and the pump delivers, with the child or the parent still bolusing for meals. Two pivotal trials carry the school-age evidence base.

The MiniMed 780G paediatric pivotal trial enrolled 160 children aged 7 to 17 across thirteen US centres. HbA1c fell from 7.9 to 7.4 percent, time in range rose from 59.4 to 70.3 percent, and there were no episodes of severe hypoglycaemia or diabetic ketoacidosis across the three-month study phase (Pihoker and colleagues 2023, Diabetes Technology and Therapeutics). Pihoker also showed that the participants using the system’s lower target with the shorter active-insulin-time setting hit the highest time in range while keeping the lowest time below range, which is the counter-intuitive finding the algorithm-savvy clinics now teach in practice.

The Omnipod 5 pivotal trial enrolled 112 children aged 6 to 13.9 and 129 adults. The children’s group moved time in range from 52.5 to 68.0 percent, an extra 3.7 hours per day in target, with HbA1c down from 7.67 to 6.99 percent (Brown and colleagues 2021, Diabetes Care). The extension study at three years showed the gains were durable, and 95 percent of children who completed the trial chose to continue. Both are single-arm pivotal trials, which is the trial design regulators require for AID approval; real-world post-approval datasets are now in the tens of thousands.

AID from diagnosis, the honest reading

A reasonable question, when a family has just left the hospital with a new diagnosis, is whether starting an AID system immediately changes what comes next. The UK CLOuD trial answered the hardest version of that question. Researchers randomised 97 newly diagnosed young people aged 10 to 16.9 years, within three weeks of diagnosis, to either CamAPS FX closed-loop or to standard sensor-augmented therapy, then followed them for four years (Ware and colleagues 2024, Diabetes Care).

The primary hypothesis was that the early algorithm would soften the immune cascade and preserve residual beta-cell function. The primary outcome was negative. C-peptide decline was no different in the two groups at 24 months and remained no different at 48 months. The autoimmune destruction did not slow.

The secondary glycaemic outcomes were a different story. At 48 months, the closed-loop group had time in range of 61 percent against 50 percent in the control arm, and HbA1c of 7.1 percent against 7.9 percent. The proportion reaching the ADA HbA1c goal of under 7.0 percent was 59 percent against 22 percent. Severe hypoglycaemia and DKA rates were similar. Ware (2024) is the trial that gives the honest answer: closed-loop from diagnosis does not preserve beta cells, but it does produce a sustained glycaemic benefit over four years without raising safety events.

For a family with a fresh diagnosis, the framing matters. The case for early AID is now a glycaemic case, not a disease-modifying one. The strength of that glycaemic case is real, and the decision belongs with your paediatric diabetes team in the weeks after diagnosis.

Targets in this age band, what the evidence supports

ISPAD’s 2024 paediatric glycaemic-targets chapter (de Bock and colleagues, Hormone Research in Paediatrics) is the population-level reference for the school-age years. Two HbA1c targets, stratified by technology access. With CGM and a safely working AID system, the target is 6.5 percent (48 mmol/mol). Without that technology, it is 7.0 percent (53 mmol/mol). Time in range, glucose between 3.9 and 10.0 mmol/L (70 and 180 mg/dL), is targeted at over 70 percent; time below 3.9 mmol/L (70 mg/dL) at under 4 percent; time in tight range (3.9 to 7.8 mmol/L, 70 to 140 mg/dL) is the emerging metric the field is moving towards.

The cliff to avoid is reading these as universal pass-or-fail thresholds. The HbA1c-to-complications relationship is continuous and non-linear; the population-average target is the orientation, the individual target lives with the diabetes team. A child running 7.4 percent on a safely working AID is not failing. A child running 7.4 percent on multiple daily injections with no CGM is in a different conversation, with different next steps.

What the modern targets do is name what is achievable when technology is in the picture. The 0.5 percentage-point difference between the two ISPAD tiers maps almost exactly onto the technology gradient. That is the equity advocacy framing the rest of this guide carries. The pivot most families describe is not “we hit the number”. It is “we have the technology that lets us pursue the number safely”.

School, the written individualised plan

The school day is the longest stretch of the week where the parent is not in the room. For most children with T1D, it covers a meal, a snack, two or three break windows, physical education, and occasionally the kind of event that breaks the routine: a trip, an exam, an assembly that runs late. The work of holding all that together belongs to the school, the family and the diabetes team, with the child increasingly in the conversation.

ISPAD’s school-based care chapter (Lawrence and colleagues 2022, Pediatric Diabetes) sets the operational anchor: a written individualised diabetes care plan, co-authored by the family, the paediatric diabetes team and the school, refreshed annually or after any major regimen change. The plan covers monitoring at school (including CGM and data-sharing where used), meal-time insulin support, hypoglycaemia treatment with glucagon access named, hyperglycaemia and ketone recognition, physical activity adjustments, and the emergency contact pathway. School trips and exams sit inside the same plan rather than as a separate document.

Most paediatric diabetes teams in the UK can provide a template; if your school has not seen one before, asking for the team’s template is the cleanest place to start. CGM data-sharing with a designated staff member is a recognised best-practice option in the 2022 update; the configuration is set with the team and the family, not with the school in isolation.

For the child, the school plan is also where the conversation about visible diabetes lives. The pump on the waistband, the phone with the CGM trace, the snack in the desk. The school staff being on the same page as the family reduces how much of that the child has to defend themselves.