Table 2.

Association between BMI and type 1 diabetes and risk of bias across included studies (n = 10)

Study	Effect estimates overweight (95%CI)	Effect estimates obesity (95%CI)	Increment in risk for 1 unit of BMI	Adjusted to	NOS quality score
Carlsson et al. [25]	0.38 (0.10-1.36)a	1.16 (0.25–5.41)	0.92 (0.78–1.08)	Sex, age, smoking	7
Viner et al. [19]	-	3.1 (1.0–9.3)	1.8 (1.2–2.8)	Sex, maternal educational, birth weight, puberty stage, breastfeeding, social class	7
Harpsøe et al. [24]	1.42 (0.95–2.14)	2.67 (1.71–4.17)	1.07 (1.04–1.11)	Smoking, alcohol, parity, socio-occupational status	8
Tosur et al. [20]	1.37 (1.05–1.79)b	–	1.00 (0.997–1.004)	Sex, number of Abs (single vs. multiple), DPTRS, HLA-genotype	8
Nucci et al. [21]	1.26 (0.71–2.21)b	1.59 (0.50–5.08)	N/Ac	Sex, birth weight z-score, birth length z-score, HLA risk, maternal type 1 diabetes, mode of delivery, breastfeeding duration (> \ < 6 months)	8
Herzog et al. [23]	1.88 (1.43–2.47)b	–	N/Ac	Sex, age (timescale), calendar time, fasting status (overnight fasting vs. no fasting), socioeconomic status, country of birth	7
Zucker et al. [15]	1.54 (1.23–1.94)	2.05 (1.58–2.66)	1.25 (1.17–1.32)	Sex, age, birth year, education, cognitive score	8
Ferrannini et al. [16]	N/Ad	N/Ad	Females: 2.25 (1.31–3.79)a Males: 1.27 (0.72–2.13)a	Age, ß-cell glucose sensitivity insulin sensitivity, projective 5-year risk of diabetes (high vs. moderate)	7
Steck et al. [18]	N/Ad	N/Ad	0.86 (0.72–1.02)	Unadjusted	7
So et al. [17]	N/Ad	N/Ad	1.2 (1.11–1.31)	Unadjusted	7

NOS Newcastle–Ottawa Scale, Abs antibodies; DPTRS Diabetes Prevention Trial-Type 1 Risk Score.

^aRR (risk ratio).

^bHazard ratios were reported for overweight or obesity.

^cNot applicable, as BMI was not modeled as a continuous variable.

^dNot applicable, as BMI was not applied into categories.