Maternal Age at Birth and Childhood Type 1 Diabetes: A Pooled Analysis of 30 Observational Studies
Article Figures & Tables
Figures
- FIG. 1.
Meta-analysis of the unadjusted association between maternal age (per 5-year increase) and type 1 diabetes (including 14,724 case subjects) using the random effects model; studies are ordered by publication date.
- FIG. 2.
Scatter plot of odds ratio for diabetes per 5-year increase in maternal age by difference in response rates between case and control subjects (size of plotting symbol was proportional to precision of study; line was taken from meta-regression).
Tables
- TABLE 1
Characteristics of included studies investigating the association between maternal age and type 1 diabetes, ordered by publication date
First author, year* (reference) Design Country Type 1 diabetic subjects Control subjects Available confounders‡ Ascertainment method (year case subjects diagnosed) Age at dx (years) n† Resp rate (%) Source (matching criteria) n† Resp rate (%) BO BW GA MD CS BF# (months) Dahlquist, 1992 (9) C-C Sweden Swedish childhood diabetes register (78–88) 0–14 2,757 98 Medical birth registry (birth year, unit) 8,271 100 Bock, 1994 (10) C-C Denmark Hosp. admission from National Patient Registry (78–89) <16 837 98 Birth registry (age, sex) 837 NA Patterson, 1994 (11) C-C Scotland Hosp. admission/childhood diabetes register (76–88) 0–14 271 100 Maternal discharge records (age, sex, area) 1,340 100 ✓ ✓ ✓ ✓ ✓ ✓(any) Wadsworth, 1997 (25) C-C U.K. British Paediatric Association Surveillance Unit (92) 0–5 213 89 Health Authority Immunization Register 318 70 ✓ ✓ ✓ ✓(4) Gimeno, 1997 (26) C-C Brazil Diabetes association/ Hospital admission (95) 0–19 344 91 Unclear (neighborhood, sex, age)‖ 333 100 ✓ ✓ ✓ ✓(3) McKinney, 1999 (28) C-C England Yorkshire Childhood Diabetes Register (93–94) 0–15 220 94 General practitioner's records (age, sex) 423 82 ✓ ✓ ✓ ✓ ✓ ✓(any) Rami, 1999 (29) C-C Austria Vienna type 1 diabetes register (89–94) 0–14 103 86 Schools (age, sex) 373 80 ✓ ✓ ✓ ✓§ ✓ ✓(any) Bache, 1999 (19)** C-C Denmark Hospital admission (78–95) 0–14 857 100 Medical birth registry (month, sex, district) 1,404 100 ✓ Dahlquist, 1999 (14) C-C Bulgaria W. Bulgaria type 1 diabetes register (91–94) 0–14 125 73 Schools and policlinics (age) 440 79 ✓ ✓ ✓ ✓§ ✓ ✓(any) C-C Latvia Latvian type 1 diabetes register (89–94) 0–14 140 99 Population register (age) 301 79 ✓ ✓ ✓ ✓§ ✓ ✓(any) C-C Lithuania Lithuanian type 1 diabetes register (89–94) 0–14 117 94 Policlinics (age) 266 73 ✓ ✓ ✓ ✓§ ✓ ✓(any) C-C Luxembourg Luxembourg type 1 diabetes register (89–95) 0–14 59 100 Preschools and schools (age) 172 95 ✓ ✓ ✓ ✓§ ✓ ✓(any) C-C Romania Bucharest type 1 diabetes register (89–94) 0–14 81 74 Preschools and schools (age) 277 81 ✓ ✓ ✓ ✓§ ✓ ✓(any) Stene, 2001 (13) Cohort Norway Norwegian Childhood Diabetes Registry (89–98) 0–14 1,810 100¶ Norwegian medical birth registry 1,382,602 NA ✓ ✓ ✓ ✓ ✓ Visalli, 2003 (30) C-C Italy Lazio type 1 diabetes register (89–95) 0–14 139 100 Schools (age) 703 91 ✓ ✓ ✓ ✓ ✓(3) Stene, 2004 (31) C-C Norway Norwegian Childhood Diabetes Registry (98–00) 0–14 346 73 Norwegian population registry 1,626 56 ✓ ✓ ✓ ✓§ ✓ ✓(3) Sadauskaite-Kuehne, 2004 (15) C-C Sweden SE Sweden type 1 diabetes register (95–00) 0–15 442 100 Population register 1,084 73 ✓ ✓ ✓ ✓§ ✓(3) C-C Lithuania Lithuanian type 1 diabetes register (96–00) 0–15 281 100 Outpatient clinic‖ 807 95 ✓ ✓ ✓ ✓§ ✓(3) Sumnik, 2004 (32) C-C Czech Republic Czech Republic type 1 diabetes registry (95–00) 0–15 640 79 National Birth registry (age) 32,000 100 ✓ Marshall, 2004 (33) C-C England Morecambe Bay/E. Lancashire diabetes clinics (98) 0–15 196 83 Health Authorities (sex, birth date) 381 53 ✓ ✓ ✓ ✓§ ✓(any) Cardwell, 2005 (34) Cohort N. Ireland N. Ireland type 1 diabetes register (71–01) 0–14 990 92¶ Northern Ireland Child Health register 439,647 NA ✓ ✓ ✓ ✓ ✓ ✓(any) Sipetić, 2005 (35) C-C Serbia Belgrade Hospital admission (94–97) 0–16 105 91 Hospital outpatients with skin disease‖ (age, sex, area) 210 100 ✓ ✓ ✓ ✓§ ✓ ✓(4) Svensson, 2005 (36) C-C Denmark Danish register of childhood diabetes (96–99) 0–14 602 100 Danish population register (age, sex) 1,459 100 ✓ ✓ ✓ ✓ ✓ ✓(4) Bottini, 2005 (20)** C-C Sardinia Hospital diagnosis ? 189 ? Consecutive births in northern Sardinia 5,460 ? Polańka, 2006 (37) C-C Poland Upper Silesia Diabetes Register (89–96) 0–14 394 87 Central Bureau for Statistics 994,460 100 ✓ ✓ Wei, 2006 (38) C-C Taiwan School-based urine screening program & questionnaire (92–97) 0–18 260 87 Randomly selected negatives from screening program 533 88 ✓ ✓ ✓ ✓ ✓(3) Tenconi, 2007 (39) C-C Italy Pavia type 1 diabetes register (88–00) 0–19 99 85 Hospital (age, sex, week) 194 ? ✓ Haynes, 2007 (40) Cohort Australia W. Australian Children's Diabetes Register (80–02) 0–14 926 99¶ W. Australia Midwives' Notification System ∼557,707 NA ✓ ✓ ✓ ✓ Ievins, 2007 (41) Cohort England Hosp. admission [ICD diabetes code] (63–99) 0–14 410 — Oxfordshire/W. Berkshire maternity records 266,665 NA ✓ ✓ ✓ ✓ ✓ ✓(any) Borras Perez, 2007 (42) C-C Spain Catalonia type 1 diabetes register (97–08) 0–14 626 72 Catalonia Public Health Birth Register 3,320 98 ✓ ✓ ✓(any) Rosenbauer, 2008 (12) C-C Germany Nationwide hosp. based surveillance (92–95) 0–4 747 71 Local registration offices (age, sex, area) 1,820 43 ✓ ✓ ✓§ ✓(4) Waldhoer, 2008 (43) Cohort Austria Austrian diabetes register (89–05) 0–5 444 85¶ Birth certificate registry 1,435,385 NA ✓ ✓ ✓ -
*Year of publication.
-
†Number included in analysis of maternal age.
-
‡Tick denotes data recorded in study and available for analysis.
-
§Maternal type 1 diabetes used in analyses.
-
‖Not randomly selected and population based.
-
¶Percentage of case subjects identified in cohort.
-
#Duration of breast-feeding used in adjusted analysis shown in brackets.
-
**Only included in sensitivity analyses. dx, diagnosis; Resp, response; BF, breast-feeding (in months); BO, birth order; BW, birth weight; C-C, case-control; CS, cesarean section; GA, gestational age; Hosp., hospital; MD, maternal diabetes; NA, not applicable.
-
- TABLE 2
Meta-analyses of 30 studies investigating the association between maternal age and type 1 diabetes before and after adjustments for recorded confounders and in subgroups defined by study type and quality
Maternal age (years) Case subjects (n) Combined OR (95% CI) P Heterogeneity χ2 (P) I2 Overall (n = 30 studies) <20 764 0.88 (0.74−1.04) 0.12 81.4 (<0.001) 64 20–25 3,919 0.95 (0.89−1.00) 0.05 36.1 (0.17) 20 25–30 5,433 1.00 (ref.) 30–35 3,274 1.05 (0.97−1.13) 0.28 59.1 (0.001) 51 ≥35 1,334 1.10 (1.01−1.20) 0.03 36.4 (0.16) 20 Per 5-year increase 14,724 1.05 (1.02−1.09) 0.006 97.7 (<0.001) 70 Adjusted for gestational age, birth weight, and birth order*(n = 20 studies) <20 403 0.95 (0.77−1.17) 0.65 42.7 (0.001) 56 20–25 1,846 0.90 (0.84−0.97) 0.003 20.9 (0.34) 9 25–30 2,826 1.00 (ref.) 30–35 1,709 1.05 (0.93−1.19) 0.40 46.4 (<0.001) 59 ≥35 737 1.12 (0.97−1.29) 0.14 33.0 (0.024) 42 Per 5-year increase 7,521 1.06 (1.00−1.12) 0.05 66.5 (<0.001) 71 Adjusted for all available confounders as shown in Table 1 (n = 30 studies) <20 736 0.89 (0.74−1.07) 0.22 88.9 (<0.001) 67 20–25 3,715 0.93 (0.87−0.99) 0.02 36.2 (0.17) 20 25–30 5,147 1.00 (ref.) 30–35 3,105 1.08 (0.99−1.18) 0.10 62.4 (<0.001) 54 ≥35 1,251 1.12 (1.02−1.24) 0.02 39.9 (0.09) 27 Per 5-year increase 13,954 1.06 (1.01−1.11) 0.01 116.9 (<0.001) 75 Cohort studies (n = 5 studies) <20 269 0.80 (0.65−0.99) 0.04 9.3 (0.06) 57 20–25 1,105 0.89 (0.82−0.96) 0.003 3.8 (0.43) 0 25–30 1,681 1.00 (ref.) 30–35 1,057 0.99 (0.88−1.12) 0.93 8.7 (0.07) 54 ≥35 468 1.08 (0.96−1.22) 0.21 5.2 (0.26) 23 Per 5-year increase 4,580 1.06 (1.01−1.11) 0.03 12.7 (0.01) 69 Case-control studies (n = 25 studies) <20 495 0.91 (0.73−1.14) 0.41 71.5 (<0.001) 66 20–25 2,814 0.97 (0.91−1.05) 0.47 28.9 (0.22) 17 25–30 3,752 1.00 (ref.) 30–35 2,217 1.07 (0.97−1.19) 0.20 49.6 (0.002) 52 ≥35 866 1.12 (0.99−1.25) 0.07 30.9 (0.16) 22 Per 5-year increase 10,144 1.05 (1.00−1.11) 0.04 84.6 (<0.001) 72 Studies with a low risk of bias† (n = 14 studies) <20 518 0.81 (0.70−0.94) 0.005 20.8 (0.08) 38 20–25 2,547 0.90 (0.86−0.96) <0.001 9.3 (0.75) 0 25–30 3,648 1.00 (ref.) 30–35 2,195 1.08 (0.99−1.18) 0.10 23.8 (0.03) 45 ≥35 904 1.18 (1.06−1.32) 0.003 18.3 (0.14) 29 Per 5-year increase 9,812 1.10 (1.06−1.14) <0.001 27.6 (0.01) 53 -
*Includes only studies for which adjustments for birth weight (in categories <2.5, 2.5−3, 3−3.5, 2−4.5, >4.5 kg), gestational age (in categories ≤ 37, 38−41, ≥42 weeks), and birth order (in categories first, second, or third born or later) could be made.
-
†Excluding case-control studies that have control subjects who were not randomly selected (or population based) or studies in which the response rate in either the case or control subjects was less than 80% (or unknown) as shown in Table 1.
-
In this Issue
