Table 3.
Childhood obesity and risk of developing adult obesity
| Study (#, first author) (n = 24) | Type of study (P, R) | Country | Total sample n (% male) | Childhood obesity | Analysis/results | ± (f/m)* | P value | If data adjusted |
|---|---|---|---|---|---|---|---|---|
| (24) Corvalan | P | Guatemala | 1,559 (43%) | BMI at 5 years | Correlation | + (f/m) | <0.05 | |
| (31) Ezzahir | P | France | 127 (49%) | BMI gain at 6 years | OR 1.9 | + (f/m) | <0.01 | Adjusted |
| (CI – 1.2, 3.0) | ||||||||
| (34) Freedman | P | USA | 626 (44%) | BMI at 5 years | Correlation | + (f/m) | <0.001 | Adjusted |
| (35) Garn | P | USA | 383 (44%) | Fatness (1–5 years) | RR 1.77† of adult fatness | + (f/m) | <0.05 | |
| (37) Gasser | P | Europe | 232 (52%) | Early childhood BMI | RR increased | + (f/m) | NR | |
| (38) Gigante | P | Brazil | 2,250 (100%) | Wt for ht Z score at 2 + 4 years | anova | + (m) | <0.001 | Adjusted |
| (43) Guo | P | USA | 459 (50%) | BMI percentile >75th in childhood† | RR of high BMI | |||
| (>25 kg m−2) at 18 years | + (m) | NR | ||||||
| 1.1–2.4 | + (f) | NR | ||||||
| 1.3–3.1 | ||||||||
| RR of high BMI | ||||||||
| (>25 kg m−2) at 30 years | ||||||||
| 1.1–1.4 | + (m) | |||||||
| 1.2–1.8 | + (f) | |||||||
| (45) Guo | P | USA | 555 (50%) | Child BMI 95th vs. 50th percentile | OR 2.91 | + (f) | NR | N/A |
| (CI – 1.34, 6.34) | ||||||||
| OR 3.73 | + (m) | NR | ||||||
| (CI – 2.12,6.55) | ||||||||
| (46) Guo | P | USA | 347 (48%) | BMI at age 3 | anova | + (f) | <0.05 | |
| obese vs. non-obese @ 35 years | (m) | NSD | ||||||
| (49) Hawk | P | UK | 621 (51%) | Body fat in childhood (skinfolds) | Correlation with adult fatness | + (f/m) | NR | |
| (51) He | P | Sweden | 3,650 (51%) | BMI at 5 years | RR 4.24 overwt at 18 years | + (f/m) | NR | |
| (55) Kindblom | R | Sweden | 612 (100%) | BMI 1–4 years | Correlation | + (m) | <0.001 | Adjusted |
| (58) Kubo | P | Japan | 244 (0%) | BMI 3 months–5 years | Correlation | + (f) | <0.001 | |
| (63) Landhuis | P | New Zealand | 972 (52%) | BMI at 5 years | OR 1.4 | + (f/m) | <0.001 | Adjusted |
| (CI – 1.18, 1.67) | ||||||||
| Obesity at 32 years | ||||||||
| (72) Magarey | P | Australia | 155 (57%) | BMI at 2 years | RR 2.72 Overwt at 20 years | + (f/m) | NR | Adjusted |
| (76) McCarthy | P | UK | 679 (54%) | BMI at 5 years | Correlation | Adjusted | ||
| 3 months, 1.5 years | (f/m) | NSD | ||||||
| 5 years | + (f/m) | <0.001 | ||||||
| (89) Prokopec | P | Czech republic | 158 (51%) | BMI at 1 year (fat vs. lean) | RR 1.8 | + (f/m) | ||
| Obesity (>75th percentile) at 18 years | NR | |||||||
| (96) Rolland-Cachera | P | France | 102 (64%) | BMI >75th percentile | RR 2.0 Obesity (>75th percentile) at 20 years | + (f/m) | NR | |
| (100) Sachdev | P | India | 1,526 (58%) | Child BMI at 6 months | Correlation | + (f/m) | NR | Adjusted |
| (103) Schack-Nielsen | P | Denmark | 4,306 (NR) | BMI at 1 + 3 years | Correlation | Adjusted | ||
| 1 year | + (f/m) | <0.05 | ||||||
| 3 years | + (f/m) | <0.05 | ||||||
| (105) Schroeder | P | Guatemala | 866 (58%) | BMI 1–5 years | Correlation | |||
| 1–3 years | + (f/m) | |||||||
| 5 years | + (f/m) | |||||||
| (107) Siervogel | P | USA | 459 (50%) | BMI ≥2 years§ | Significant log OR with high adult BMI | + (f) | ||
| BMI ≥5 years§ | Significant log OR with high adult BMI | + (m) | ||||||
| (127) Whitaker | R | USA | 854 (36%) | BMI >85th percentile | OR 1.3 | (f/m) | NSD | Adjusted |
| (CI – 0.7, 2.5) | ||||||||
| 1–2 years | ||||||||
| OR 4.1 | + (f/m) | |||||||
| (CI – 2.5, 6.7) | ||||||||
| 3–5 years | ||||||||
| (128) Williams | P | New Zealand | 925 (50%) | BMI 3 + 5 years | Correlation | + (f/m) | <0.05 |
*
(f/m) = females/males. Results are shown for adult female offspring (f), adult male offspring (m) or both (f/m).
†
For two studies (35,49), the early marker was evaluated with adult fat mass (not body mass index [BMI]), with body fat measured by skinfold measures.
†
Early measures were completed ‘in childhood’. Specific ages were not provided.
§
A significant relationship (the log odds ratio) between childhood BMI and a high adult BMI was found in females beginning at 2 years of age and higher, and in males beginning at 5 years of age and higher.
CI, confidence interval; NSD, no significant difference. No association; NR, not reported; OR, odds ratio; P, prospective study; R, retrospective study.