Pediatric Obesity: Pharmacotherapy Selector
Step-by-step guide to selecting and initiating pharmacotherapy in children ≥12 years with obesity. All agents must be combined with behavioural and psychological interventions.
Based on CMAJ 2025 Clinical Practice Guideline (Ball et al., doi: 10.1503/cmaj.241456).
Step 1 — Confirm Eligibility
Before prescribing, confirm all three:
| Check | Criteria |
|---|
| Age | ≥ 12 years (no evidence for < 12) |
| Diagnosis | Obesity confirmed (BMIz using WHO/Canadian charts) |
| Behavioural program | Currently enrolled or starting alongside medication |
⚠️ Pharmacotherapy must always be co-prescribed with behavioural and psychological interventions. Medication alone is not recommended.
Step 2 — Assess Contraindications and Preferences
Before selecting an agent, screen for:
- GI conditions (IBD, gastroparesis) — caution with GLP-1RAs
- Eating disorder (binge eating, atypical anorexia) — psychiatric clearance before any pharmacotherapy
- Renal or hepatic impairment — affects metformin dosing
- Family access and affordability — GLP-1RAs are costly; metformin widely available
- Patient/family values — injection vs. oral, frequency, side effect tolerability
Step 3 — Select Agent
Option A — GLP-1 Receptor Agonists (Preferred where accessible)
Semaglutide, liraglutide, exenatide
Conditional recommendation; very low to low certainty
Best for: Significant BMIz reduction needed; cardiometabolic comorbidities (hypertension, dyslipidaemia, insulin resistance)
Evidence:
- BMIz: small to large reduction (semaglutide has largest effect — very large effect on BMIz; 1 RCT, 201 children aged 12–17)
- HRQoL: small benefit (semaglutide); little/no effect for others
- Triglycerides: large benefit (semaglutide); small for others
- HDL-C, LDL-C, total cholesterol, systolic BP: small benefit (semaglutide)
Common AEs (important — discuss upfront):
- GI: nausea, diarrhea, vomiting, abdominal pain (most common; 79% semaglutide vs. 82% placebo — many are trivial and self-limiting)
Serious AEs (uncertain risk):
- Cholelithiasis, appendicitis (11% semaglutide vs. 9% placebo in key RCT)
- Risk appears similar to control groups but uncertain — monitor
Semaglutide note: Stronger evidence than liraglutide/exenatide based on current data, but guideline recommends GLP-1RAs as a class due to limited paediatric RCTs.
Option B — Biguanides (Metformin)
Good first option, especially if GLP-1RAs unavailable or unaffordable
Conditional recommendation; low to moderate certainty
Best for: Insulin resistance present; oral route preferred; cost/access barrier to GLP-1RAs
Evidence:
- BMIz: moderate reduction
- Cardiometabolic: small benefits on lipids and insulin resistance
- HRQoL, anxiety, depression: little to no effect
AEs:
- No serious AEs reported in paediatric studies
- Mild-moderate GI (nausea, diarrhea) — more than placebo but manageable; start low and titrate
- Not associated with increased serious AEs (critical safety advantage over GLP-1RAs)
Option C — Lipase Inhibitors (Orlistat)
Use only if A and B are unavailable or contraindicated
Conditional recommendation; low certainty — least preferred
Evidence: Lacks evidence on HRQoL, depression, anxiety.
AEs:
- More serious AEs than control (critically important — higher risk)
- More mild-moderate GI AEs (oily stools, faecal urgency, leakage)
- Fat-soluble vitamin malabsorption — supplement required
⚠️ Avoid as first choice. Only use if GLP-1RAs and metformin are not feasible.
Step 4 — Initiate and Counsel
Before prescribing, cover all three with the child and family:
- Expected benefits: BMIz reduction is likely modest in most (not guaranteed); improvements in cardiometabolic markers; possibly HRQoL
- Expected harms: Discuss the relevant AE profile for the chosen agent
- Duration: Obesity is chronic — medication may be long-term; reassess regularly
- What to do if GI side effects occur: Reassure most are transient; hydration; dose titration strategy
Step 5 — Monitoring Plan
Review at 3 months, then every 6 months:
| Outcome | Why |
|---|
| BMIz | Primary efficacy marker |
| HRQoL (PedsQL or similar) | Critically important to families |
| Depression & anxiety screening | Critically important — monitor throughout |
| BP (systolic and diastolic) | Cardiometabolic target |
| Fasting lipids (total cholesterol, LDL-C, HDL-C, TG) | Cardiometabolic benefit expected |
| Fasting insulin / HOMA-IR | Insulin resistance marker |
| ALT | Hepatic safety (especially GLP-1RAs) |
| GI adverse events | Adherence-limiting; especially GLP-1RAs |
| Serious AEs (hospitalisation) | Especially cholelithiasis with GLP-1RAs |
Clinical Guardrails
- No pharmacotherapy under 12 years — no evidence base; do not extrapolate
- Always combine with behavioural program — medication alone is not guideline-concordant
- Screen for eating disorders before starting — pharmacotherapy may worsen body image disturbance
- Shared decision-making is mandatory — especially given low-to-moderate certainty evidence for all agents
- Discontinue if no benefit after adequate trial (typically 3–6 months) and reassess
- GLP-1RAs are not stepwise — can be offered alongside or before trying metformin based on clinical picture and family preference
Source
CMAJ 2025 Clinical Practice Guideline: Managing obesity in children.
Ball GDC et al. CMAJ 2025 April 14; 197:E372–89. doi: 10.1503/cmaj.241456
Updated version: June 3, 2025.