Early age at menarche may represent another indicator of metabolic risk for development of type 2 diabetes.
Abstract
Context:
Early age at menarche has been linked to an elevated risk of type 2 diabetes. However, the underlying mechanism is unclear.
Objective:
Our objective was to examine associations between age at menarche and type 2 diabetes risk factors.
Design, Participants, and Setting:
The BioCycle Study followed 253 healthy premenopausal women from the general population (Buffalo, NY) for one to two menstrual cycles.
Main Outcome Measures:
Age at menarche was self-reported. Body mass index and waist circumference were measured by trained personnel. Total body and trunk fat were measured by dual-energy x-ray absorptiometry. Fasting glucose, insulin, highly sensitive C-reactive protein, and SHBG levels were measured up to eight times per cycle. Insulin resistance (IR) and β-cell function were evaluated using the homeostasis model assessment (HOMA)-IR and HOMA-β.
Results:
The mean age at menarche was 12.5 ± 1.2 yr. After adjustment for age, race, education, and physical activity, early menarche (≤12 yr) was significantly associated with an increase of 1.35 kg/m2 in body mass index (P = 0.01), 2.52% in percent total body fat (P = 0.004), 3.02% in percent trunk fat (P = 0.004), 0.15 μIU/ml in (log)insulin (P = 0.02), 0.15 U in (log)HOMA-IR (P = 0.03), and 0.16 U in (log)HOMA-β (P = 0.01) compared with average menarche (12–14 yr). No associations were found for SHBG or highly sensitive C-reactive protein.
Conclusions:
Early onset of menarche is associated with unfavorable metabolic phenotypes compared with average onset of menarche in healthy premenopausal women, including reduced insulin sensitivity and β-cell function and greater total and trunk fat.
The average age at menarche is declining in many countries (1, 2), coincidental with the trend of increasing prevalence of type 2 diabetes (T2D) over the past few decades. Recent cohorts have found that earlier age at menarche was associated with an elevated risk of T2D among middle- and older-aged women (3, 4). However, mechanisms that explain this association have yet to be identified. Although several studies have investigated associations of earlier age at menarche with T2D risk markers such as elevated fasting glucose (5–7) or insulin (5, 8) and insulin resistance (5, 9, 10), findings have been inconsistent. Furthermore, although menstrual cycle has been shown to affect glucose homeostasis (11), none of the previous studies considered the influence of menstrual cycle phase in premenopausal women. In addition, data on associations between age at menarche and other emerging biomarkers of T2D risk such as highly sensitive C-reactive protein (hsCRP) and SHBG are lacking. In this study, we aimed to investigate the associations between age at menarche and metabolic markers important for the development of T2D, taking into account the potential impact of menstrual cycle on glucose homeostasis in premenopausal women by using the longitudinal measures of metabolic markers over the menstrual cycle.
Subjects and Methods
Study population and design
The details of the study design and data collection of the BioCycle Study have been previously reported (12). Briefly, 259 healthy premenopausal women aged 18–44 yr with regular menstruation were recruited from Buffalo, NY. Participants were asked to attend 16 visits over two cycles (eight per menstrual cycle). Visits were scheduled using fertility monitors to correspond to specific phases of the menstrual cycle. Overall, 94% of women completed at least seven clinic visits per cycle. The University at Buffalo Health Sciences Institutional Review Board approved the study, and all participants provided written informed consent.
Data collection
Of 259 women who participated in the BioCycle Study, 253 provided information on their age at menarche and were included in the present analysis. Age at menarche was self-reported at baseline via a question asking, “How old were you when you had your first menstrual period?” Information on age, race, education, income, marital status, family history of diabetes, smoking, and alcohol consumption were also collected by self-report. Physical activity was measured by the International Physical Activity Questionnaire, which assessed vigorous and moderate intensity activities (13).
Anthropometry was measured by trained personnel at the beginning of the study. Standing height using a fixed stadiometer and weight using a calibrated scale were the average of two measurements. Body mass index (BMI) was calculated by dividing weight (kilograms) by the square of height (meters). Waist circumference (WC) was measured in duplicate with a tape applied horizontally midway between the iliac crest and lowest lateral portion of the rib cage. Body composition was measured at the completion of the study visits using a dual-energy x-ray absorptiometry scan from which total percent body fat (%BF) and percent trunk fat (%TF) were derived.
Fasting blood samples were drawn the morning of each clinic cycle visit. Serum insulin, SHBG, estradiol, progesterone, LH, FSH, and hsCRP were measured using the competitive chemiluminescent enzymatic immunoassay. Fasting plasma glucose was assayed using the hexokinase-based methodology on a Beckman LX20 autoanalyzer. Insulin resistance [homeostasis model assessment for insulin resistance (HOMA-IR)] and secretion [HOMA for B-cell function (HOMA-β)] were calculated based on the homeostasis model (14).
Statistical analysis
Age at menarche was evaluated as a continuous variable in years and then grouped into three categories based on the distribution (tertiles). Women with menarcheal age in the first, second, and third tertiles were categorized as early (range, 9–12 yr), average (range, >12 to <14), and late (range, 14–16 yr) menarche, respectively. All measures except for BMI, WC, %BF, and %TF were log-transformed for normality. Descriptive data were expressed as mean ± sd if not specified otherwise. One-way ANOVA and χ2 tests were applied to compare continuous variables and categorical variables, respectively. Linear mixed-effects models with random intercepts were applied for the outcomes and exposures with repeated measurements (e.g. fasting insulin and glucose, estradiol, progesterone, SHBG, hsCRP, HOMA-IR, and HOMA-β). Covariates such as age, race, education, and physical activity were adjusted in the multivariate models to examine the influence of potential confounding. All statistical analyses were performed using STATA version 9.0 (Stata Corp., College Station, TX). Statistical significance was set at P ≤ 0.05 (two-tailed).
Results
Among the 253 women with information on age at menarche, 60% were Caucasian, 20% were African Americans, 15% were of Asian, and 5% were other race. The mean age at menarche was 12.5 ± 1.2 yr. Women in the early menarche group had greater BMI, higher levels of %BF, %TF, fasting insulin, HOMA-IR, and HOMA-β compared with women with average menarche (Table 1).
Table 1.
Baseline characteristics according to the tertiles of age at menarche of 253 healthy, premenstrual women from the BioCycle study
| Age at menarche (yr) |
Total cohort | ||||
|---|---|---|---|---|---|
| Early menarche, tertile 1 (9–12 yr) | Average menarche, tertile 2 (>12 to <14 yr) | Late menarche tertile 3 (14–16 yr) | P value | ||
| n (%)a | 127 (50.2) | 78 (30.8) | 48 (19.0) | 253 (100) | |
| Age at menarche (yr) | 11.4 (0.7) | 13.0 (0) | 14.3 (0.5) | <0.001 | 12.5 (1.2) |
| Age at study baseline (yr) | 26.8 (8.3) | 28.5 (8.7) | 31.2 (3.2) | 0.31 | 27.3 (8.2) |
| Ethnicity (%) | 0.25 | ||||
| Caucasian | 54.3 | 70.5 | 56.3 | 59.7 | |
| African-American | 23.6 | 12.8 | 20.8 | 19.8 | |
| Asian | 18.1 | 10.3 | 16.7 | 15.4 | |
| Others | 3.9 | 6.4 | 6.3 | 5.1 | |
| Education, % with college degree or above | 35.4 | 46.2 | 47.9 | 0.18 | 41.1 |
| Married (%) | 18.9 | 18.1 | 35.4 | 0.06 | 24.9 |
| Leisure PA (MET) | 1179.7 (2157.7) | 1336.8 (1588.7) | 1461.5 (1703.6) | 0.66 | 1281.6 (1911.2) |
| Total PA (MET) | 5401.9 (6751.9) | 5030.7 (5119.8) | 5942.3 (7592.6) | 0.74 | 5390.0 (6453.9) |
| Current smoking (%) | 3.9 | 2.6 | 6.3 | 0.19 | 4.0 |
| Alcohol, ≥3 drinks/wk | 44.1 | 38.5 | 35.4 | 0.38 | 40.7 |
| BMI (kg/m2) | 24.8 (3.8) | 23.5 (3.8) | 21.0 (3.7) | 0.02 | 24.1 (3.8) |
| BMI categories | 0.17 | ||||
| Overweight (25 ≤ BMI < 30 kg/m2) | 29.1 | 25.3 | 18.8 | 26.1 | |
| Obese (BMI ≥30 kg/m2) | 11.8 | 6.4 | 10.4 | 9.9 | |
| Waist circumference (cm) | 75.8 (8.4) | 73.3 (8.5) | 73.3 (8.5) | 0.17 | 74.9 (8.6) |
| %BF | 30.8 (5.8) | 28.3 (5.8) | 28.1 (6.2) | 0.004 | 29.5 (6.0) |
| %TF | 26.6 (7.1) | 23.6 (7.0) | 23.4 (8.0) | 0.006 | 25.1 (7.3) |
| Fasting insulin (μIU/mL)b | 6.4(1.8) | 5.0 (1.9) | 5.3 (1.8) | 0.02 | 5.7 (1.9) |
| Fasting glucose (mg/dl) | 87.0 (6.0) | 88.8 (7.5) | 85.2 (6.2) | 0.01 | 87.2 (6.6) |
| HOMA-IRb | 1.4 (1.8) | 1.1 (2.0) | 1.1 (1.9) | 0.03 | 1.2 (1.9) |
| HOMA-βb | 96.9 (1.8) | 72.9 (2.0) | 90.2 (1.8) | 0.007 | 87.6 (1.9) |
| hsCRP (mg/liter)b | 1.0 (2.9) | 0.8 (3.4) | 0.9 (4.0) | 0.45 | 0.9 (3.2) |
| SHBG (nmol/liter)b | 41.1 (1.6) | 45.1 (1.9) | 43.2 (1.6) | 0.70 | 42.7 (1.7) |
| Follicular estradiol (pg/ml)b | 52.9 (1.8) | 49.1 (1.9) | 48.4 (1.9) | 0.58 | 50.3 (1.9) |
| Luteal estradiol (pg/ml)b | 112.4 (1.8) | 121.8 (1.7) | 98.1 (1.7) | 0.11 | 112.3 (1.7) |
| Follicular progesterone (ng/ml)b | 0.39 (1.8) | 0.38 (1.8) | 0.36 (1.8) | 0.61 | 0.38 (1.82) |
| Luteal progesterone (ng/ml)b | 4.6 (3.4) | 4.0 (3.4) | 3.7 (4.1) | 0.56 | 4.1 (3.6) |
| Ovulatory FSH (mIU/ml)b | 6.7 (1.7) | 6.9 (1.8) | 7.1 (1.6) | 0.86 | 6.8 (1.7) |
| Ovulatory LH (ng/ml)b | 10.7 (2.2) | 10.9 (2.5) | 10.7 (2.5) | 0.99 | 10.7 (2.3) |
Unless indicated otherwise, data are presented as mean (sd). All biomarkers and hormone values are averages of two cycles. Conversion factors are, for glucose, mg/dl × 0.0555 = mmol/liter; for insulin, μIU/liter × 6.945 = pmol/liter; for estradiol, pg/ml × 3.671 = pmol/liter; for progesterone, ng/ml × 3.18 = nmol/liter; for FSH, mIU/ml × 1.0 = IU/liter; and for LH, mIU/ml × 1.0 = IU/liter. MET, Metabolic equivalents; PA, physical activity.
The tertiles have unequal number of participants because a large number of women had menarcheal age at 13 yr.
Geometric mean (sd) was calculated because of skewed distribution.
After controlling for age, race, education, and physical activity, women in the early menarche group had significantly greater mean levels of fasting insulin (regression coefficient β = 0.15 μIU/ml; P = 0.02), HOMA-IR (β = 0.15; P = 0.02), and HOMA-β (β = 0.16; P = 0.01) compared with women in the average menarche group over the menstrual cycle (Table 2). Women in the late menarche group did not have significantly different insulin, HOMA-IR, or HOMA-β values. No significant associations were found between age at menarche and fasting glucose, SHBG, or hsCRP in either crude or multivariate adjusted models. We also conducted stratified analyses by race. Age at menarche was significantly and inversely associated with fasting insulin, HOMA-IR, and HOMA-β in Caucasians. Corresponding regression coefficients were −0.072 (P = 0.033), −0.073 (P = 0.035), and −0.069 (P = 0.037), respectively. Associations of age at menarche with fasting insulin, HOMA-IR, and HOMA-β among African-Americans, however, were not statistically significant, although the magnitude and the direction of the associations were similar to those in Caucasians. Corresponding regression coefficients were −0.063 (P = 0.18), −0.064 (P = 0.20), and −0.070 (P = 0.094), respectively.
Table 2.
Association of age at menarche with metabolic markers for T2D
| Outcomes | Age at menarche |
||||
|---|---|---|---|---|---|
| Early menarche, tertile 1 (9–12 yr) |
Average menarche, tertile 2 (>12 to <14 yr) β (se) | Late menarche, tertile 3 (14–16 yr) |
|||
| β (se) | P | β (se) | P | ||
| Fasting glucosea (mg/dl) | |||||
| Crude model | −0.65 (1.17) | 0.58 | Reference | −1.11 (1.48) | 0.45 |
| Multivariate adjusted model | −0.11 (1.16) | 0.92 | Reference | −0.63 (1.47) | 0.67 |
| Fasting insulina (log) (mmol/ml) | |||||
| Crude model | 0.19 (0.07) | 0.006 | Reference | 0.08 (0.09) | 0.31 |
| Multivariate adjusted model | 0.15 (0.07) | 0.02 | Reference | 0.08 (0.08) | 0.32 |
| HOMA-IRa (log) | |||||
| Crude model | 0.21 (0.07) | 0.005 | Reference | 0.10 (0.09) | 0.26 |
| Multivariate adjusted model | 0.15 (0.07) | 0.03 | Reference | 0.08 (0.09) | 0.36 |
| HOMA-βa (log) | |||||
| Crude model | 0.24 (0.07) | 0.001 | Reference | 0.13 (0.09) | 0.14 |
| Multivariate adjusted model | 0.16 (0.06) | 0.01 | Reference | 0.09 (0.08) | 0.27 |
| SHBGa (log) (nmol/liter) | |||||
| Crude model | −0.11 (0.07) | 0.09 | Reference | −0.01 (0.08) | 0.83 |
| Multivariate adjusted model | −0.10 (0.06) | 0.11 | Reference | −0.02 (0.08) | 0.83 |
| hsCRPa (log) (mg/liter) | |||||
| Crude model | 0.16 (0.14) | 0.26 | Reference | 0.16 (0.17) | 0.35 |
| Multivariate adjusted model | 0.22 (0.13) | 0.09 | Reference | 0.22 (0.16) | 0.17 |
| BMIb (kg/m2) | |||||
| Crude model | 1.26 (0.53) | 0.02 | Reference | −0.07 (0.69) | 0.92 |
| Multivariate adjusted model | 1.35 (0.54) | 0.01 | Reference | 0.13 (0.68) | 0.85 |
| Waist circumferenceb (cm) | |||||
| Crude model | 1.66 (1.23) | 0.18 | Reference | −0.83 (1.57) | 0.60 |
| Multivariate adjusted model | 1.94 (1.18) | 0.10 | Reference | −0.25 (1.50) | 0.87 |
| %BFb | |||||
| Crude model | 2.44 (0.86) | 0.005 | Reference | −0.23 (1.12) | 0.84 |
| Multivariate adjusted model | 2.52 (0.85) | 0.004 | Reference | −0.06 (1.10) | 0.95 |
| %TFb | |||||
| Crude model | 2.93 (1.06) | 0.006 | Reference | −0.34 (1.38) | 0.86 |
| Multivariate adjusted model | 3.02 (1.04) | 0.004 | Reference | −0.02 (1.34) | 0.99 |
Multivariate adjusted models were adjusted for age, race, education, and physical activity.
Estimated by mixed-effects models.
Estimated by the linear regression models.
Age at menarche was also inversely associated with measures of body adiposity in linear regression models (Table 2). Compared with women in the average menarche group, women in the early menarche group had greater BMI (β = 1.30 kg/m2; P = 0.02), %BF (β = 2.51; P = 0.003), and %TF (β = 3.02; P = 0.004) even after adjustment for age, race, education, and physical activity.
Discussion
In this study among healthy premenopausal women, we found that earlier menarche (≤12 yr) was associated with elevated levels of fasting insulin and measures of insulin resistance (HOMA-IR) and β-cell function (HOMA-β) as well as adulthood total adiposity (i.e. BMI and %BF) and central adiposity (i.e. %TF), compared with average menarche (12–14 yr). We did not find significant association of menarcheal age with fasting glucose, SHBG, or hsCRP levels.
Findings from the present study on association of early menarche with insulin resistance and β-cell dysfunction are in line with the results from previous studies (5, 8–10). However, we did not find an association between early menarche and increased fasting glucose level as previous studies did (5–7). This may be partly because our population was relatively younger and healthier. Changes in measures of glucose homeostasis have been observed over the menstrual cycle in the BioCycle Study (11). We expanded findings from previous studies by considering the influence of menstrual cycle phase in the present study. We applied longitudinal analytic approach to handle the repeated measures over the menstrual cycle to maximize the statistical power. Our findings were also consistent with previous studies (5, 8–10) supporting a strong inverse association between age at menarche and body adiposity in adulthood. In the present study, body adiposity was measured using the dual-energy x-ray absorptiometry scan compared with using the BMI in previous studies. To our knowledge, our study was also the first one to examine the association of age at menarche with SHBG and hsCRP in premenopausal women.
The biological mechanism by which women with earlier age at menarche are associated with increased insulin resistance and β-cell dysfunction is not well understood. It has been proposed that early maturation in women may reflect a longer duration of positive energy balance or a greater accumulation of body fat (15). Adipose tissue is now recognized as a metabolically active organ that secrets many adipokines (16). Although the etiology of excessive body adiposity in relation to insulin resistance has not been fully established, a growing body of evidence supports a possible causal association. Increased adiposity in obese individuals can be characterized as a stage of low-grade inflammation that is thought to contribute to local insulin resistance. This loss of insulin sensitivity within adipose tissue can subsequently lead to systemic insulin resistance and/or hyperinsulinemia through several pathways, including uncontrolled release of fatty acids, secretion of inflammatory cytokines, and alterations in the balance of adipokines (17). In further exploratory analysis, the significant associations between age at menarche and fasting insulin, HOMA-IR, and HOMA-β were attenuated after adjustment for adulthood adiposity. However, these results should be interpreted with caution because the current data do not allow us to evaluate whether the association is independent of body adiposity due to several constraints. In particular, in the present study, body adiposity was measured in the same time window as insulin, HOMA-IR, and HOMA-β, such that temporality could not be established.
The present study was strengthened by repeated measurements of key outcomes across menstrual cycle phases, high follow-up rate, and direct measures of body adiposity. Nevertheless, several limitations should be acknowledged. One limitation of our study is that age at menarche was reported by recall. Therefore, misclassification of age at menarche was possible. However, previous studies have shown that women's recalled age at menarche in the middle age was highly correlated (r = 0.79) with the information reported in childhood (18). Another limitation is we were unable to account for unmeasured factors such as birth weight and childhood adiposity, which may influence the onset of menarche and T2D. It is plausible that the associations of age at menarche with insulin resistance and adulthood adiposity are confounded by childhood adiposity because obesity tends to track between childhood and adulthood. However, results from other longitudinal studies that followed girls from childhood to young adulthood tend to suggest that childhood adiposity and age at menarche are associated with adulthood adiposity independently (19, 20). Also, due to the observational nature of the present study, we cannot completely rule out the impact of residual confounding.
In conclusion, results from the present study suggest that early menarche is associated with unfavorable metabolic phenotypes implicated in the development of T2D compared with average age of menarche in premenopausal women. Further study and follow-up are required to determine whether early age at menarche may also represent another indicator for later development of T2D in adulthood.
Acknowledgments
This work was supported by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health.
Disclosure Summary: Authors have nothing to declare.
Footnotes
- %BF
- Percent body fat
- BMI
- body mass index
- HOMA-β
- homeostasis model assessment for β-cell function
- HOMA-IR
- homeostasis model assessment for insulin resistance
- hsCRP
- highly sensitive C-reactive protein
- T2D
- type 2 diabetes
- %TF
- percent trunk fat
- WC
- waist circumference.
References
- 1. Anderson SE, Must A. 2005. Interpreting the continued decline in the average age at menarche: results from two nationally representative surveys of U.S. girls studied 10 years apart. J Pediatr 147:753–760 [DOI] [PubMed] [Google Scholar]
- 2. Harris MA, Prior JC, Koehoorn M. 2008. Age at menarche in the Canadian population: secular trends and relationship to adulthood BMI. J Adolesc Health 43:548–554 [DOI] [PubMed] [Google Scholar]
- 3. He C, Zhang C, Hunter DJ, Hankinson SE, Buck Louis GM, Hediger ML, Hu FB. 2010. Age at menarche and risk of type 2 diabetes: results from 2 large prospective cohort studies. Am J Epidemiol 171:334–344 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Lakshman R, Forouhi N, Luben R, Bingham S, Khaw K, Wareham N, Ong KK. 2008. Association between age at menarche and risk of diabetes in adults: results from the EPIC-Norfolk cohort study. Diabetologia 51:781–786 [DOI] [PubMed] [Google Scholar]
- 5. Frontini MG, Srinivasan SR, Berenson GS. 2003. Longitudinal changes in risk variables underlying metabolic syndrome X from childhood to young adulthood in female subjects with a history of early menarche: the Bogalusa Heart Study. Int J Obes Relat Metab Disord 27:1398–1404 [DOI] [PubMed] [Google Scholar]
- 6. Heys M, Schooling CM, Jiang C, Cowling BJ, Lao X, Zhang W, Cheng KK, Adab P, Thomas GN, Lam TH, Leung GM. 2007. Age of menarche and the metabolic syndrome in China. Epidemiology 18:740–746 [DOI] [PubMed] [Google Scholar]
- 7. Saquib N, Kritz-Silverstein D, Barrett-Connor E. 2005. Age at menarche, abnormal glucose tolerance and type 2 diabetes mellitus: The Rancho Bernardo Study. Climacteric 8:76–82 [DOI] [PubMed] [Google Scholar]
- 8. Remsberg KE, Demerath EW, Schubert CM, Chumlea WC, Sun SS, Siervogel RM. 2005. Early menarche and the development of cardiovascular disease risk factors in adolescent girls: the Fels Longitudinal Study. J Clin Endocrinol Metab 90:2718–2724 [DOI] [PubMed] [Google Scholar]
- 9. Feng Y, Hong X, Wilker E, Li Z, Zhang W, Jin D, Liu X, Zang T, Xu X, Xu X. 2008. Effects of age at menarche, reproductive years, and menopause on metabolic risk factors for cardiovascular diseases. Atherosclerosis 196:590–597 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Kivimäki M, Lawlor DA, Smith GD, Elovainio M, Jokela M, Keltikangas-Järvinen L, Vahtera J, Taittonen L, Juonala M, Viikari JS, Raitakari OT. 2008. Association of age at menarche with cardiovascular risk factors, vascular structure, and function in adulthood: the Cardiovascular Risk in Young Finns study. Am J Clin Nutr 87:1876–1882 [DOI] [PubMed] [Google Scholar]
- 11. Yeung EH, Zhang C, Mumford SL, Ye A, Trevisan M, Chen L, Browne RW, Wactawski-Wende J, Schisterman EF. 2010. Longitudinal study of insulin resistance and sex hormones over the menstrual cycle: The BioCycle Study. J Clin Endocrinol Metab 95:5435–5442 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Wactawski-Wende J, Schisterman EF, Hovey KM, Howards PP, Browne RW, Hediger M, Liu A, Trevisan M. 2009. BioCycle study: design of the longitudinal study of the oxidative stress and hormone variation during the menstrual cycle. Paediatr Perinat Epidemiol 23:171–184 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Craig CL, Marshall AL, Sjöström M, Bauman AE, Booth ML, Ainsworth BE, Pratt M, Ekelund U, Yngve A, Sallis JF, Oja P. 2003. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc 35:1381–1395 [DOI] [PubMed] [Google Scholar]
- 14. Wallace TM, Levy JC, Matthews DR. 2004. Use and abuse of HOMA modeling. Diabetes Care 27:1487–1495 [DOI] [PubMed] [Google Scholar]
- 15. van Lenthe FJ, Kemper CG, van Mechelen W. 1996. Rapid maturation in adolescence results in greater obesity in adulthood: the Amsterdam Growth and Health Study. Am J Clin Nutr 64: 18–24 [DOI] [PubMed] [Google Scholar]
- 16. Galic S, Oakhill JS, Steinberg GR. 2010. Adipose tissue as an endocrine organ. Mol Cell Endocrinol 316:129–139 [DOI] [PubMed] [Google Scholar]
- 17. Gutierrez DA, Puglisi MJ, Hasty AH. 2009. Impact of increased adipose tissue mass on inflammation, insulin resistance, and dyslipidemia. Curr Diab Rep 9:26–32 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Must A, Willett WC, Dietz WH. 1993. Remote recall of childhood height, weight, and body build by elderly subjects. Am J Epidemiol 138:56–64 [DOI] [PubMed] [Google Scholar]
- 19. Demerath EW, Li J, Sun SS, Chumlea WC, Remsberg KE, Czerwinski SA, Towne B, Siervogel RM. 2004. Fifty-year trends in serial body mass index during adolescence in girls: the Fels Longitudinal Study. Am J Clin Nutr 80:441–446 [DOI] [PubMed] [Google Scholar]
- 20. Mamun AA, Hayatbakhsh MR, O'Callaghan M, Williams G, Najman J. 2009. Early overweight and pubertal maturation: pathways of association with young adults' overweight: a longitudinal study. Int J Obes (Lond) 33:14–20 [DOI] [PubMed] [Google Scholar]