Beyond body mass index: Body composition in children and adolescents with type 1 diabetes, clinical implications and perspectives

Table 1 Body-composition assessment methods in paediatric type 1 diabetes

Method	Accuracy	Reproducibility	Radiation	Cost/resources	Availability	Paediatric T1D validation	Regional fat	Serial monitoring
Anthropometry (BMI, WC, skinfolds)	Indirect assessment	High for height and weight, operator-dependent for WC and skinfolds	None	Minimal equipment	Widespread	Not applicable for direct body composition	Indirect central adiposity	Suitable for growth and weight trends only
BIA	Equation- and hydration-dependent estimates	Acceptable under standardized conditions with the same device	None	Dedicated device, lower resource requirements than imaging	Variable	Limited	Device-dependent indirect estimates, not interchangeable with MRI- or CT-measured VAT	Conditional on standardized protocol
DXA	Precise for total and regional tissues, not universal for all compartments	High within the same device and software	Low-dose	Specialized equipment and personnel	Selected centres	Limited	Trunk fat, VAT estimation with selected systems	Limited mainly by resources and access
MRI	Direct assessment of specific tissue compartments	High with standardized protocols	None	Specialized equipment and personnel	Specialized centres	Limited	Direct VAT, SAT and ectopic fat assessment	Limited mainly by resources and access
CT	High spatial resolution	High with standardized protocols	Ionizing radiation	Specialized equipment and personnel	Specialized centres	Not established	Direct VAT, SAT and muscle assessment	Unsuitable for routine serial monitoring
ADP	Model-based total FM and FFM assessment	Good under standardized conditions	None	Dedicated specialized equipment	Mainly research centres	Not established	None	Feasible but access-limited

Accuracy and reproducibility depend on the measured compartment, device, prediction algorithm and acquisition protocol. Cost and availability are described in relative terms because they vary across healthcare settings. T1D: Type 1 diabetes; BMI: Body mass index; WC: Waist circumference; BIA: Bioelectrical impedance analysis; MRI: Magnetic resonance imaging; CT: Computed tomography; VAT: Visceral adipose tissue; DXA: Dual-energy X-ray absorptiometry; SAT: Subcutaneous adipose tissue; ADP: Air displacement plethysmography; FFM: Fat-free mass; FM: Fat mass.

Table 2 Summary of key paediatric evidence on body composition in type 1 diabetes

Clinical domain	Ref.	Design and population	Method	Main findings
Adiposity and trajectories	Zheng et al[19], 2022	Systematic review and meta-analysis, 24 studies, 1017 youth with T1D and 1045 controls	DXA	Youth with T1D had 1.2 kg more fat mass and 2.3 percentage points higher body-fat percentage. Lean mass did not differ. Earlier onset and higher insulin dose were associated with larger differences
	Davis et al[25], 2012	Prospective cohort, 30 newly diagnosed children and 14 controls, assessments at diagnosis, 6 weeks and 1 year	Whole-body DXA	The lower BMI and body-fat percentage observed at diagnosis normalised by 6 weeks after insulin initiation and remained stable at 1 year. Sex differences persisted
	Lipsky et al[73], 2016	Secondary longitudinal analysis of a dietary RCT, 136 youth, 18-month follow-up	DXA and CGM	Adiposity was not associated with HbA1c but was associated with more frequent hyperglycaemic excursions and lower 1,5-anhydroglucitol
	Kahkoska et al[45], 2018	Longitudinal SEARCH cohort, 742 youth with T1D, approximately 9 years of follow-up	Anthropometry-derived estimated body-fat percentage	Lower socioeconomic indicators were associated with less favourable adiposity trajectories, particularly among girls
	Halvadzhiyan et al[64], 2026	Retrospective comparison of two single-centre cohorts assessed 20 years apart, n = 293	BIA and anthropometry	The contemporary cohort showed a trend toward higher BMI and greater fat-free mass. Girls consistently had higher fat mass. CGM use was associated with lower BMI, HbA1c and insulin dose
Glycaemic and metabolic associations	Maffeis et al[78], 2026	Cross-sectional multicentre study, 364 children and adolescents aged 6-18 years	BIA, anthropometry and CGM	FMI, body-fat percentage and waist-to-height ratio were positively associated with HbA1c and inversely with time in range. FMI and waist-to-height ratio remained independently associated with achievement of glycaemic targets
	Torre et al[72], 2026	Cross-sectional single-centre study, 133 fully pubertal adolescents	BIA and CGM	BIA-derived visceral-fat estimates correlated weakly with current and mean 12-month HbA1c. The association with mean 12-month HbA1c was not significant after adjustment. FFM% correlated weakly with glucose CV and sensor glucose SD. TIR was inversely associated with estimated BMR in a separate multivariable model with BMR as the outcome
Fat distribution and vascular markers	Rangel et al[67], 2025	Cross-sectional study, 78 children and adolescents aged 5-18 years	Segmental BIA	Nearly half had excess adiposity, including about 11% of participants with normal BMI. Girls had higher body-fat percentage and a lower muscle-to-fat ratio. Planned exercise was associated with a more favourable profile
	Ozde et al[68], 2026	Cross-sectional case-control study, 150 youth with T1D and 152 age- and sex-matched controls	Ultrasound EATT/PrATT, cIMT and eGDR	Epicardial and perirenal fat thicknesses were higher in T1D and were associated with lower estimated insulin sensitivity, inflammatory markers, hepatic steatosis and cIMT
	Tas et al[71], 2026	Cross-sectional study, 49 late-pubertal youth with T1D	MRI hepatic fat, DXA and cIMT	Hepatic fat was low and was not associated with cIMT. Several adiposity and body-composition measures, rather than liver fat, were associated with cIMT
Muscle and physical function	Huerta-Uribe et al[84], 2024	Prospective cohort, 83 youth aged 6-18 years, 1-year follow-up	DXA and handgrip strength	Persistently higher handgrip strength was associated with a more favourable cardiometabolic profile and lower adiposity. Muscle strength provided information distinct from lean mass
	Muñoz-Pardeza et al[38], 2026	24-week parallel RCT, 62 youth aged 8-18 years	DXA, progressive resistance training	Training increased lean mass by 0.88 kg and whole-body-less-head bone mineral content by 32.4 g and reduced probable sarcopenia risk. Fat mass and anthropometry were unchanged
Skeletal health	Loxton et al[86], 2021	Systematic review and meta-analysis, 46 studies, 2617 youth with T1D and 3851 controls	DXA, pQCT and quantitative ultrasound	Youth with T1D had lower bone mineral measures across several skeletal sites and imaging modalities than healthy controls
	Mondkar et al[87], 2026	Retrospective longitudinal cohort, 200 children and adolescents, 4-year follow-up	DXA and pQCT	Total-body-less-head areal BMD z-score and trabecular volumetric BMD declined over time. Longer diabetes duration and adverse metabolic factors were associated with poorer skeletal outcomes

T1D: Type 1 diabetes; DXA: Dual-energy X-ray absorptiometry; BMI: Body mass index; RCT: Randomised controlled trial; CGM: Continuous glucose monitoring; HbA1c: Glycated haemoglobin; BIA: Bioelectrical impedance analysis; FMI: Fat mass index; CV: Coefficient of variation; TIR: Time in range; BMR: Basal metabolic rate; FFM: Fat-free mass; EATT: Epicardial adipose tissue thickness; PrATT: Perirenal adipose tissue thickness; cIMT: Carotid intima-media thickness; eGDR: Estimated glucose disposal rate; MRI: Magnetic resonance imaging; pQCT: Peripheral quantitative computed tomography; BMD: Bone mineral density.