Correction
22 Sep 2015: Furer A, Afek A, Beer Z, Derazne E, Tzur D, et al. (2015) Correction: Height at Late Adolescence and Incident Diabetes among Young Men. PLOS ONE 10(9): e0139183. https://doi.org/10.1371/journal.pone.0139183 View correction
Figures
Abstract
Background
Short stature was suggested as a risk factor for diabetes onset among middle age individuals, but whether this is the case among young adults is unclear. Our goal was to assess the association between height and incident diabetes among young men.
Methods and Findings
Incident diabetes was assessed among 32,055 men with no history of diabetes, from the prospectively followed young adults of the MELANY cohort. Height was measured at two time points; at adolescence (mean age 17.4±0.3 years) and grouped according to the US-CDC percentiles and at young adulthood (mean age 31.0±5.6 years). Cox proportional hazards models were applied. There were 702 new cases of diabetes during a mean follow-up of 6.3±4.3 years. There was a significant increase in the crude diabetes incidence rate with decreasing adolescent height percentile, from 4.23 cases/104 person-years in the <10th percentile group to 2.44 cases/104 person-years in the 75th≤ percentile group. These results persisted when clinical and biochemical diabetes risk factors were included in multivariable models. Compared to the 75th≤ percentile group, height below the 10th percentile was associated with a hazard ratio (HR) of 1.64 (95%CI 1.09–2.46, p = 0.017) for incident diabetes after adjustment for age, body mass index (BMI), fasting plasma glucose, HDL-cholesterol and triglyceride levels, white blood cells count, socioeconomic status, country of origin, family history of diabetes, sleep quality and physical activity. At age 30 years, each 1-cm decrement in adult height was associated with a 2.5% increase in diabetes adjusted risk (HR 1.025, 95%CI 1.01–1.04, p = 0.001).
Citation: Furer A, Afek A, Beer Z, Derazne E, Tzur D, Pinhas-Hamiel O, et al. (2015) Height at Late Adolescence and Incident Diabetes among Young Men. PLoS ONE 10(8): e0136464. https://doi.org/10.1371/journal.pone.0136464
Editor: Jennifer L. Baker, Institute of Preventive Medicine, DENMARK
Received: March 25, 2015; Accepted: August 3, 2015; Published: August 25, 2015
Copyright: © 2015 Furer et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: The data of the current cohort is a restricted military database. Researchers who wish to use this database need to contact members of the Medical Corps of the Israeli Defense Forces or the Corresponding Author: Gilad Twig, MD, PhD, The Israel Defense Forces Medical Corps, and Department of Medicine 'B', Chaim Sheba Medical Center, Tel Hashomer, Ramat Gan 52621 Israel. Gilad.Twig@gmail.com.
Funding: This study was supported by the IDF Medical Corp and the Israeli ministry of defense. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The incidence rate of type 2 diabetes among children and young adults is increasing worldwide. Classic risk factors such as obesity or genetic variants account for only part of the increase in diabetes incidence [1], thereby emphasizing the need to explore additional risk factors to improve diabetes risk prediction.
Low stature was associated with abnormal fasting plasma glucose (FPG) [2] and diabetes in some studies [3–7], but not in all [8–11]. A recent meta-analysis concluded that height was associated with an increased diabetes risk in women, but not in men and indicated a high degree of heterogeneity among studies when controlling for diabetes risk factors [12]. The age of participants included in these studies is another source for ambiguity. Most studies included middle the characterized the height-diabetes association included middle aged participants spanning a range of over 3 decades [8,13,14], leaving paucity of data regarding young adults.
The Metabolic, Lifestyle and Nutrition Assessment in Young Adults (MELANY) cohort is a large ongoing, prospective study assessing risk factors for cardiovascular disease and diabetes [15–18] among young adults. This database also includes multiple measurements of height from late adolescence to young adulthood. Our goal was to assess the association between height at age 17 years and at young adulthood and incident diabetes among 32,055 young and apparently healthy men of this cohort.
Research Design and Methods
The MELANY cohort
The MELANY cohort is part of an ongoing investigation of the Israel Defense Forces (IDF) Medical Corps [19]. Israeli army personnel, older than 25-years of age, remaining in military service beyond the 2 to 3 years of mandatory service, are referred every 3 to 5 years for a routine health examination and screening tests at a screening center. At each visit the participants complete a detailed questionnaire assessing demographic, nutritional, lifestyle, and medical factors. Height and weight are measured, and a complete physical examination performed. Blood samples are drawn following a 14-hour fast and analyzed immediately. All medical information is recorded in the same central database, independent of scheduled visits, thereby facilitating ongoing, uniform follow-up as described previously [19]. All participants in the MELANY cohort, independent of their rank and position, have similar access to medical services which are provided free of charge [17]. In addition, prior to enlistment in the military at age 17 years, all MELANY participants underwent the mandatory, standard IDF pre-recruitment medical evaluation, comprising a physical examination including measurements of weight and height, assessment of cognitive performance and evaluation of socio-demographic data [17].
Study population
Fig 1 shows a schematic diagram of the study design and its outcomes. Included in this study were men with complete measurements of weight and height at age 17 years, who attended the screening assessment at least once between January 1, 1995 and March 8, 2011. Since the first routine metabolic screen is performed at the first scheduled visit, participants who developed diabetes (type 1 or 2) prior to their first visit (n = 63) and those with a follow-up of less than 1 year from enrollment to the diagnosis of diabetes (n = 3,009) were excluded from the analysis (Fig 1). Participants with missing weight and height data at first visit were excluded from analysis (n = 1,251). The Institutional Review Board of the IDF Medical Corps approved this study without the need for participants' informed consent, given the assurance of strict maintenance of participants’ anonymity during data analyses. The MELANY dataset included 4,497 women, 28 of whom developed diabetes. This small number of incident cases precluded meaningful statistical analyses, and thus this study included only male subjects.
Follow-up and outcomes
Follow-up began after the participants’ first visit to the screening center and ended at the time of diabetes diagnosis, death or retirement from military service or March 8, 2011, whichever came first. Screening for diabetes was performed at each visit to the center based on FPG levels. Incident cases of diabetes were based on a physician’s diagnosis of diabetes according to the American Diabetes Association criteria by documenting either two FPG levels ≥126 mg/dL (7.0 mmol/L) or a glucose level ≥200 mg/dL two hours after ingestion of 75 grams of glucose, conducted in cases in which the examining physician deemed the test necessary. All laboratory studies were performed on fresh samples, in an ISO-9002 quality-assured, core facility laboratory.
Study variables
The Centers for Disease Control and Prevention (CDC) data were selected by the Israel Ministry of Health as the routine reference for anthropometric data for children, and have been reported to be appropriate for assessing Israeli children [20]. Accordingly, height and weight at adolescence were categorized according to the percentile cut-off points of the CDC for age (in months) and sex as previously reported [21]. Socio-economic status was based on place of residence and was obtained from records of the Israeli Ministry of Interior, which stratifies all municipalities according to 10 socio-economic decile groups determined by the Israeli Central Bureau of Statistics [17]. This scale considers age distribution, available workforce, level of unemployment, level of education (proportion of undergraduate students and those entitled to a high school diploma), average per capita income and proportion of income support recipients. Socioeconomic status was categorized into 3 groups: low (1st-4th deciles), medium (5th -7th) and high (8th -10th) as reported previously [17]. Education was modeled as a categorical variable of low or high levels, with a threshold of 11 completed years of school education. This cut-off was chosen since it represents the maximum potential school instruction at the time of height measurement at adolescence. The intelligence score (IS), a Wechsler Adult Intelligence Score-equivalent IQ measure [22] and an independent risk for diabetes in this cohort [17], was treated as a continuous variable. Low intelligence scores reflect percent of scores lower than 100 points-equivalent in the Wechsler Adult Intelligence Scale. Country of origin, classified by the father’s or grandfather’s country of birth, was categorized into five geographical areas [17]: former USSR countries, Asia (non-USSR), Africa (excluding South Africa), Western (comprised of non-USSR Europe, North and South America, South Africa, Australia and New Zealand) and Israel. Country of birth was similarly classified. Data on mother's country of origin were not systematically obtained, and therefore were not included. Height at adulthood, body mass index (BMI), triglyceride and high-density lipoprotein cholesterol (HDL-c) level, FPG, and white blood cell (WBC) count at enrollment (first visit at the screening center) were treated as continuous variables. Smoking status (current smoker, past-smoker, never smoked), physical activity (≥150 min/week, <150 min/week, inactivity), breakfast consumption (frequent, sometimes, none), and family history of diabetes (yes, no) were treated as categorical variables. The mini sleep questionnaire (MSQ), a comprehensive index of sleep quality [23] and an independent risk factor for diabetes in this cohort [18] was treated as a continuous variable.
Statistical analysis
Continuous variables were summarized using means and standard deviation (SD), or intra-quartile ranges (IQR) when variables did not exhibit normal distribution. Counts with percentages were used for categorical variables. The cohort was divided into 5 groups based on the CDC age-adjusted height percentiles at pre-recruitment evaluation [24]: <10th, 10–24th, 25–49th, 50–74th and ≥75th percentiles. P for trend were calculated using a linear regression where the independent variable was the 5 height categories and the dependent variable was the mean or proportion for each covariate (as shown in Table 1). ANOVA was performed to compare the mean of continuous variables among the study groups. Dunnet T3 and Bonferoni multiple comparison tests were used when homogeneity of variance test were rejected or not, respectively. Cox proportional hazard models were used to estimate the Hazard Ratios (HR) and 95% confidence intervals (CI) for developing diabetes among the 5 adolescent height groups and additionally, using height at enrollment as a continuous variable. The differences in follow-up among the 5 height groups were compared with ANOVA and Dunnet T3 post-hoc multiple comparison test. Several models were used to assess the diabetes-height relationships after adjusting for possible confounders such as age, sex, socioeconomic status and mediators such as FPG and physical activity as follows: Model 1, age at enrollment and birth year; Model 2, age at enrollment, birth year and BMI; Model 3, the variables of model 2 and metabolic risk factors (FPG, HDL-c and triglyceride levels and WBC count); Model 4, the variables of model 2 and lifestyle risk factors (physical activity, smoking status, sleep quality, breakfast consumption); Model 5, the variables of model 2 and socio-genetic risk factors (country of origin, family history of diabetes, IS, education, socioeconomic status). The final model (model 6) included covariates that were significantly (p<0.05) associated with diabetes risk in the age-adjusted model. As the effect of obesity status on the height-diabetes association was controlled by adjustment to BMI [9] or weight [25], we also analyzed the height-diabetes association using weight instead of BMI in the multivariate models (models 2b-6b at Table A in S1 File). The association between height at adulthood and incident diabetes was also assessed by dividing the cohort into 6 group with 5-cm increments (≤165 cm, 165< and ≤170 cm, 170< and ≤175 cm, 175< and ≤180 cm, 180< and ≤185 cm and ≥185 cm). The relationship between incidence rate of diabetes (per 1,000 person-years) and the median height of each group was fitted with linear and quadratic models.
Tests for linear trend across the height percentile groups at age 17 years and the diabetes risk estimates were assessed separately for each of the above-mentioned models using the midpoint of each height percentile group. Log minus log plots for each variable were inspected to verify the assumption of proportionality of the hazards. For all variables used in the model co-linearity was assessed using the Pearson correlation and the variance inflation factor (VIF). The maximal R recorded was 0.357 (triglyceride level and BMI at enrollment) with a maximal VIF of 1.34 (tolerance of 0.74) for triglyceride level. Subjects with missing data were excluded from the multivariable analyses (12.4% in model 6). Analyses were performed with SPSS statistical software, version 19.0 (SPSS, Inc., Chicago, IL, USA).
Results
Table 1 shows the baseline characteristics of the cohort at age 17 years and upon enrollment into the MELANY cohort, according to the 5 height percentile groups at age 17 years. At the pre-recruitment (17 year old) assessment, subjects of short-stature (<10th percentile) had BMI values lower by 0.4 kg/m2 than the high-stature group (≥75th percentile; p = 0.001). Participants of Western origin predominate the tallest group (37%) as compared to pre-dominance of North African (33%) and Asian origin (33%) in the <10th height percentile group. With respect to the tallest group, the <10th height percentile group had a higher fraction of participants with lower intelligence (+17.1%), lower SES (+12%), immigrants (+6%) and low-level of education (+10%). At the first visit to the screening center the ≥75th height percentile had higher fasting plasma glucose (+1.0 mg/dL; p = 0.001) and higher IFG rates (+3.9%) despite a higher BMI (+0.57 kg/m2; p<0.001) compared to the low stature group. The ≥75th height percentile group had also lower rates of physical inactivity (61% vs. 67%), family history of diabetes (12.7% and 15.2%) and positive smoking history (-8%) as compared to the low-stature group. Of note, the height measured at young adulthood increased in all height percentile groups, compared to measurements in adolescence, with the largest increase (+4.59 cm) recorded in the <10th height percentile group, compared with an increase of 2.40 cm in the ≥75th height percentile group. Height at adolescence and adulthood were highly correlated (R = 0.92).
There were 702 new cases of diabetes diagnosed during 202,549 person-years of follow up. The incident rate of diabetes was 2.44/103 person-years for participants in the ≥75th height percentile group of height and increased with decreasing height percentile groups to 4.23/103 person-years among those with height <10th height percentile (Table 2). There was no difference in the length of follow-up among the study groups (p>0.25 in all comparison), with the exception of participants in the 10th-24th percentile group who had longer follow-up; 6.51 years vs. 6.11–6.33 in the other study groups (p<0.021 compared to all other groups).
The independent effect of height percentile at age 17 years on the risk for incident diabetes persisted in the series of Cox models adjusted for socioeconomic, family history of diabetes, lifestyle and metabolic risk factors for incident diabetes as shown in Table 2. In the multivariable analysis adjusted for age, birth year, BMI, FPG, HDL-c, triglycerides level, WBC count, socioeconomic status, country of origin, family history of diabetes, IS, MSQ score and physical activity (model 6), participants in the lowest height group had 64% higher risk (HR 1.64, 95%CI 1.09–2.46, p = 0.017) for incident diabetes compared to the tallest group. Fig 2A depicts the Cox survival curves for cumulative diabetes incidence across the study groups after controlling for the variables in model 6. When a forward stepwise model was applied to the risk factors included in model 6, height at adolescence was included in the seventh step and was preceded by FPG, BMI, age, family history of diabetes, IS and MSQ sleep quality score.
(a) Cox regression survival curves by US CDC height percentile groups were adjusted for age, birth year, BMI, FPG, HDL-c, triglycerides level, WBC count, socioeconomic status, country of origin, family history of diabetes, intelligence score, MSQ score and physical activity (model 6). Follow-up data and risk estimates are shown in Table 2. (b) The risk for diabetes incident is shown for each 1-cm decrement in height at adulthood (enrollment) for the different models used in the study (Table 2).
The height measured at young adulthood (mean age 31.0±5.6 years) was also found to be independently associated with incident diabetes after controlling for the factors in models 1–6 (Fig 2B). In the multivariable analysis (model 6), the risk for incident diabetes increased by 2.5% for every 1-cm decrement in height (HR 1.025, 95%CI 1.01–1.04, p = 0.001) with similar results obtained in sensitivity analyses for the categories of socioeconomic status, country of origin or IS (Figure A in S1 File). We found similar results when height at adolescence was introduced to the model (model 6) for those measured between 17 and 18 years (n = 27,300; HR 1.024 95%CI = 1.008–1.041, p = 0.002). We also assessed the height-diabetes association using height at adulthood categories with 5 cm increments (Table 3). The incidence rate of diabetes decreased from 6.1 to 3.7 cases per 1,000 person-years in those shorter than 166 cm and those in the 171–175cm group, respectively, and plateaued for individuals taller than 175 cm. For clarity, this is shown in Fig 3. A linear and quadratic fit had a goodness (R2) of 0.73 and 0.96, respectively. These risk estimates persisted in multivariate analysis (model 6 inTable 2) with a significant increase for those shorter than 170 cm and a borderline significance for those in the 171–175cm group.
Data is shown based on the division shown in Table 3. A linear and quadratic fit had a goodness (R2) of 0.73 and 0.96, respectively.
Note that hazard ratio across categories were adjusted to model 6 in Table 2.
Discussion
The current study of 32,055 men with 202,549 person-years of follow-up is to the best of our knowledge, the largest prospectively-designed cohort study assessing the relationship between height and incident diabetes. We demonstrated an inverse relationship between the risk for diabetes and height at both late adolescence and adulthood after adjustment for clinical and biochemical diabetes risk factors. The adjusted risk for incident diabetes increased by approximately 2.5% for every 1 cm decrement in height measured at adulthood (95%CI = 1.010–1.040, p = 0.001), with an adjusted HR of 1.64 (95%CI = 1.09–2.46, p = 0.017) among subjects of short stature (<10th percentile) at adolescence compared to those in the upper height quartile. We found that the inverse relationship between height at adulthood and diabetes risk was not homogenous throughout the entire height range with a comparable risk for those taller than 175 cm and borderline risk for those at height 170–175cm. This finding is in agreement with Njostad et al [9] who also reported an inverse relationship between height and diabetes incidence with a cutoff of 168cm based on 87 cases. While most of the other studies, although not all [9], support a similar inverse association between height and incidence diabetes [3–5,7], the later was mostly examined among middle aged participants that spanned an age range higher than 3 decades [8,13,14].
There is a large heterogeneity in adjustment for confounders and mediators of diabetes. Socioeconomic status is associated with diabetes [26] and also with height [27,28], but was not controlled for in nearly half of the studies evaluating the relationship between height and diabetes [12]. In our study, the height-diabetes association persisted when three socioeconomic-related variables were controlled for (area of residence, education level, and cognitive performance). Limited access to healthcare services is another important factor that is associated with both low socioeconomic status and under-diagnosis of diabetes [29,30]. In this respect, the MELANY cohort is advantageous in minimizing these effects as all participants had similar and free of charge access to medical services with a scheduled screening program, independent of their rank and position.
Family history of diabetes was rarely adjusted in some of the studies assessing the relationship between height or hip circumference and diabetes [31,32]. In our study the height-diabetes association persisted after controlling for a family history of diabetes and participants' origin. It is noteworthy that the MELANY cohort comprises people from a wide range of different backgrounds and origins, since Israel is considered a “young” country with a relatively high rate of immigration. The young age of our participants at enrollment may limit the accuracy of reporting family history of diabetes, as it is possible that first degree relatives of participants ruling out a history of diabetes might develop diabetes in the future. For these participants, short stature may carry an additional risk to the risk associated with a positive family history of diabetes.
The participants' age at assessment is another source of ambiguity. The literature evaluating the height-diabetes association is predominantly based on studies including middle aged or older participants, or those with age ranges spanning 3 decades or more [6–8,13,14,25]. This age heterogeneity can bias the height-diabetes association either by the age-dependent decline in height that occurs in midlife onwards, or by an age-dependent susceptibility to obesity and metabolic abnormalities among young compared to older individuals [33], especially when metabolic variables are not controlled for [12]. In our study height was measured at two time intervals; late adolescence and young adulthood. The strict time interval of height measurement and adjustment for sex and age (by month) using US CDC height percentiles, together with a second measurement in young adulthood, minimizes the age-dependent bias.
Several childhood and metabolic mechanisms may underlie the association between height and diabetes. These include premature birth, irrespective of size for gestational age [34], fetal growth retardation especially when followed by a postnatal obesogenic environment [35], rapid weight gain in the first 3 months of life [36], low rates of linear growth during infancy [37], altered insulin-like growth factor-I levels [38], and vitamin D level and/or receptor polymorphism [39,40].
Our findings should be viewed in the context of the growing fraction of unexplained cases of diabetes. It is estimated that as many as half of the increase in the prevalence of diabetes cases occurring during the last three decades were not related to traditional risk factors [1,41]. The fact that the height-diabetes association persisted after adjustment for diabetes risk factors generally assessed in clinical settings, raises the possibility that height should be added to diabetes risk stratification among young men.
This study has several limitations. Firstly, our data does not include anthropometric measurements other than height, such as leg length, hip or waist circumference, and therefore could not assess the relative specificity of height compared to other anthropometric indices [4,42,43]. Secondly, no antibody data were available and as such the type of diabetes (eg, Type 1) could not be ascertained. However, we recently reported for this cohort that over 98% of diabetes-diagnosed cases were not prescribed insulin during the first year after diabetes diagnosis, thereby supporting type 2 predominance [17]. With that respect it is noteworthy that diabetes prevalence and age of onset was similar to other cohorts studying risk factors for type 2 diabetes of young adults [44–47].Finally, our study was limited to men. The strengths of the study include the large number of participants, standardized assessments and repeat measurements, rather than self-reported measurements [48] of weight and height, with adjustment for a large number of clinical, social, lifestyle and biochemical diabetes risk factors.
In conclusion, we found that height at adolescence or young adulthood was inversely associated with diabetes risk among young and apparently healthy young men after adjustment for clinical and biochemical diabetes risk factors. Our results suggest that height should be included in the diabetes risk stratification among young men.
Supporting Information
S1 File. Multivariable assessment of Hazard Ratios (HR) for developing diabetes by US CDC height percentile groups at age 17 years for different clusters of risk factors using weight rather than BMI (Table A).
Sensitivity analysis for the effect of socioeconomic status (SES), country of birth and intelligence score on the height-diabetes association (Figure A).
https://doi.org/10.1371/journal.pone.0136464.s001
(DOCX)
Acknowledgments
This study was supported by a research grant of the IDF Medical Corp and the Israeli ministry of defense. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Author Contributions
Conceived and designed the experiments: GT. Performed the experiments: ED GT. Analyzed the data: AF ED GT. Contributed reagents/materials/analysis tools: DT. Wrote the paper: AF GT. Commented on the manuscript: AA ZB ED DT BR OPH.
References
- 1. Menke A, Rust KF, Fradkin J, Cheng YJ, Cowie CC (2014) Associations between trends in race/ethnicity, aging, and body mass index with diabetes prevalence in the United States: a series of cross-sectional studies. Ann Intern Med 161: 328–335. pmid:25178569
- 2. Brown DC, Byrne CD, Clark PM, Cox BD, Day NE, Hales CN et al. (1991) Height and glucose tolerance in adult subjects. Diabetologia 34: 531–533. pmid:1916060
- 3. Janghorbani M, Amini M (2012) Associations of hip circumference and height with incidence of type 2 diabetes: the Isfahan diabetes prevention study. Acta Diabetol 49 Suppl 1: S107–114. pmid:22080142
- 4. Snijder MB, Dekker JM, Visser M, Bouter LM, Stehouwer CD, Kostense PJ et al. (2003) Associations of hip and thigh circumferences independent of waist circumference with the incidence of type 2 diabetes: the Hoorn Study. Am J Clin Nutr 77: 1192–1197. pmid:12716671
- 5. Snijder MB, Zimmet PZ, Visser M, Dekker JM, Seidell JC, Shaw JE (2004) Independent and opposite associations of waist and hip circumferences with diabetes, hypertension and dyslipidemia: the AusDiab Study. Int J Obes Relat Metab Disord 28: 402–409. pmid:14724659
- 6. Han TS, Feskens EJ, Lean ME, Seidell JC (1998) Associations of body composition with type 2 diabetes mellitus. Diabet Med 15: 129–135. pmid:9507913
- 7. Kumari M, Head J, Marmot M (2004) Prospective study of social and other risk factors for incidence of type 2 diabetes in the Whitehall II study. Arch Intern Med 164: 1873–1880. pmid:15451762
- 8. Lorenzo C, Williams K, Stern MP, Haffner SM (2009) Height, ethnicity, and the incidence of diabetes: the San Antonio Heart Study. Metabolism 58: 1530–1535. pmid:19586642
- 9. Njølstad I, Arnesen E, Lund-Larsen PG (1998) Sex differences in risk factors for clinical diabetes mellitus in a general population: a 12-year follow-up of the Finnmark Study. Am J Epidemiol 147: 49–58. pmid:9440398
- 10. Wang SL, Pan WH, Hwu CM, Ho LT, Lo CH, Lin SL et al. (1997) Incidence of NIDDM and the effects of gender, obesity and hyperinsulinaemia in Taiwan. Diabetologia 40: 1431–1438. pmid:9447951
- 11. Bozorgmanesh M, Hadaegh F, Zabetian A, Azizi F (2011) Impact of hip circumference and height on incident diabetes: results from 6-year follow-up in the Tehran Lipid and Glucose Study. Diabet Med 28: 1330–1336. pmid:21627688
- 12. Janghorbani M, Momeni F, Dehghani M (2012) Hip circumference, height and risk of type 2 diabetes: systematic review and meta-analysis. Obes Rev 13: 1172–1181. pmid:22943765
- 13. Liu J, Tan H, Jeynes B (2009) Is femur length the key height component in risk prediction of type 2 diabetes among adults? Diabetes Care 32: 739–740. pmid:19171722
- 14. Schulze MB, Heidemann C, Schienkiewitz A, Bergmann MM, Hoffmann K, Boeing H (2006) Comparison of anthropometric characteristics in predicting the incidence of type 2 diabetes in the EPIC-Potsdam study. Diabetes Care 29: 1921–1923. pmid:16873804
- 15. Twig G, Afek A, Shamiss A, Derazne E, Tzur D, Gordon B, et al. (2013) White blood cells count and incidence of type 2 diabetes in young men. Diabetes Care 36: 276–282. pmid:22961572
- 16. Twig G, Gerstein H, Ben-Ami Shor D, Derazne E, Tzur D, Afek A, et al. (2015) Coronary Artery Disease Risk among Obese Metabolically Healthy Young Men. Eur J Endocrinol 173(3):305–312. pmid:26041076
- 17. Twig G, Gluzman I, Tirosh A, Gerstein HC, Yaniv G, Afek A, et al. (2014) Cognitive function and the risk for diabetes among young men. Diabetes Care 37: 2982–2988. pmid:25092683
- 18. Twig G, Shina A, Afek A, Derazne E, Tzur D, Cukierman-Yaffe T, et al. (2015) Sleep quality and risk of diabetes and coronary artery disease among young men. Acta Diabetol., in press.
- 19. Tirosh A, Shai I, Tekes-Manova D, Israeli E, Pereg D, Shochat T, et al. (2005) Normal fasting plasma glucose levels and type 2 diabetes in young men. N Engl J Med 353: 1454–1462. pmid:16207847
- 20. Goldstein A, Haelyon U, Krolik E, Sack J (2001) Comparison of body weight and height of Israeli schoolchildren with the Tanner and Centers for Disease Control and Prevention growth charts. Pediatrics 108: E108. pmid:11731635
- 21. Twig G, Afek A, Shamiss A, Derazne E, Landau Rabbi M, Tzur D, et al. (2014) Adolescence BMI and trends in adulthood mortality: a study of 2.16 million adolescents. J Clin Endocrinol Metab 99: 2095–2103. pmid:24601695
- 22. Reichenberg A, Weiser M, Rapp MA, Rabinowitz J, Caspi A, Schmeidler J, et al. (2005) Elaboration on premorbid intellectual performance in schizophrenia: premorbid intellectual decline and risk for schizophrenia. Arch Gen Psychiatry 62: 1297–1304. pmid:16330717
- 23. Melamed S, Oksenberg A (2002) Excessive daytime sleepiness and risk of occupational injuries in non-shift daytime workers. Sleep 25: 315–322. pmid:12003162
- 24. Kuczmarski RJ, Ogden CL, Guo SS, Grummer-Strawn LM, Flegal KM, Mei Z, et al. (2002) 2000 CDC Growth Charts for the United States: methods and development. Vital Health Stat 11: 1–190.
- 25. Lawlor DA, Ebrahim S, Davey Smith G (2002) The association between components of adult height and Type II diabetes and insulin resistance: British Women's Heart and Health Study. Diabetologia 45: 1097–1106. pmid:12189439
- 26. Espelt A, Borrell C, Roskam AJ, Rodríguez-Sanz M, Stirbu I, Dalmau-Bueno A, et al. (2008) Socioeconomic inequalities in diabetes mellitus across Europe at the beginning of the 21st century. Diabetologia 51: 1971–1979. pmid:18779946
- 27. Jousilahti P, Tuomilehto J, Vartiainen E, Eriksson J, Puska P (2000) Relation of adult height to cause-specific and total mortality: a prospective follow-up study of 31,199 middle-aged men and women in Finland. Am J Epidemiol 151: 1112–1120. pmid:10873136
- 28. Batty GD, Leon DA (2002) Socio‐economic position and coronary heart disease risk factors in children and young people. 263–272 p. pmid:12506501
- 29. Zhang X, Geiss LS, Cheng YJ, Beckles GL, Gregg EW, Kahn HS (2008) The missed patient with diabetes: how access to health care affects the detection of diabetes. Diabetes Care 31: 1748–1753. pmid:18753665
- 30. Beagley J, Guariguata L, Weil C, Motala AA (2014) Global estimates of undiagnosed diabetes in adults. Diabetes Res Clin Pract 103: 150–160. pmid:24300018
- 31. Conway B, Xiang YB, Villegas R, Zhang X, Li H, Wu X, et al. (2011) Hip circumference and the risk of type 2 diabetes in middle-aged and elderly men and women: the Shanghai women and Shanghai men's health studies. Ann Epidemiol 21: 358–366. pmid:21458729
- 32. Parker ED, Pereira MA, Stevens J, Folsom AR (2009) Association of hip circumference with incident diabetes and coronary heart disease: the Atherosclerosis Risk in Communities study. Am J Epidemiol 169: 837–847. pmid:19224980
- 33. Twig G, Afek A, Derazne E, Tzur D, Cukierman-Yaffe T, Gerstein H, et al. (2014) Diabetes risk among overweight and obese metabolically healthy young adults. Diabetes Care 37: 2989–2995. pmid:25139886
- 34. Hofman PL, Regan F, Jackson WE, Jefferies C, Knight DB, Robinson EM, et al. (2004) Premature birth and later insulin resistance. N Engl J Med 351: 2179–2186. pmid:15548778
- 35. Chernausek SD (2012) Update: consequences of abnormal fetal growth. J Clin Endocrinol Metab 97: 689–695. pmid:22238390
- 36. Leunissen RW, Kerkhof GF, Stijnen T, Hokken-Koelega A (2009) Timing and tempo of first-year rapid growth in relation to cardiovascular and metabolic risk profile in early adulthood. JAMA 301: 2234–2242. pmid:19491185
- 37. Eriksson JG, Osmond C, Kajantie E, Forsen TJ, Barker DJ (2006) Patterns of growth among children who later develop type 2 diabetes or its risk factors. Diabetologia 49: 2853–2858. pmid:17096117
- 38. Sandhu MS, Heald AH, Gibson JM, Cruickshank JK, Dunger DB, Wareham NJ (2002) Circulating concentrations of insulin-like growth factor-I and development of glucose intolerance: a prospective observational study. Lancet 359: 1740–1745. pmid:12049864
- 39. Ortlepp JR, Metrikat J, Albrecht M, von Korff A, Hanrath P,Hoffmann R (2003) The vitamin D receptor gene variant and physical activity predicts fasting glucose levels in healthy young men. Diabet Med 20: 451–454. pmid:12786678
- 40. Tsur A, Feldman BS, Feldhammer I, Hoshen MB, Leibowitz G, Balicer RD (2013) Decreased serum concentrations of 25-hydroxycholecalciferol are associated with increased risk of progression to impaired fasting glucose and diabetes. Diabetes Care 36: 1361–1367. pmid:23393216
- 41. Bullard KM, Saydah SH, Imperatore G, Cowie CC, Gregg EW, Geiss LS, et al. (2013) Secular changes in U.S. Prediabetes prevalence defined by hemoglobin A1c and fasting plasma glucose: National Health and Nutrition Examination Surveys, 1999–2010. Diabetes Care 36: 2286–2293. pmid:23603918
- 42. Asao K, Kao WH, Baptiste-Roberts K, Bandeen-Roche K, Erlinger TP, Brancati FL (2006) Short stature and the risk of adiposity, insulin resistance, and type 2 diabetes in middle age: the Third National Health and Nutrition Examination Survey (NHANES III), 1988–1994. Diabetes Care 29: 1632–1637. pmid:16801590
- 43. Smith GD, Greenwood R, Gunnell D, Sweetnam P, Yarnell J, Elwood P (2001) Leg length, insulin resistance, and coronary heart disease risk: the Caerphilly Study. J Epidemiol Community Health 55: 867–872. pmid:11707479
- 44. Koskinen J, Magnussen CG, Sabin MA, Kahonen M, Hutri-Kahonen N, Laitinen T, et al. (2014) Youth Overweight and Metabolic Disturbances in Predicting Carotid Intima-Media Thickness, Type 2 Diabetes, and Metabolic Syndrome in Adulthood: The Cardiovascular Risk in Young Finns Study. Diabetes Care 37(7):1870–1877. pmid:24742659
- 45. Nguyen QM, Srinivasan SR, Xu JH, Chen W, Hassig S, Rice J, et al. (2011) Elevated liver function enzymes are related to the development of prediabetes and type 2 diabetes in younger adults: the Bogalusa Heart Study. Diabetes Care 34: 2603–2607. pmid:21953798
- 46. Vassy JL, Dasmahapatra P, Meigs JB, Schork NJ, Magnussen CG, Chen W, et al. (2012) Genotype prediction of adult type 2 diabetes from adolescence in a multiracial population. Pediatrics 130: e1235–1242. pmid:23071215
- 47. Vassy JL, Hivert MF, Porneala B, Dauriz M, Florez JC, Dupuis J, et al. (2014) Polygenic type 2 diabetes prediction at the limit of common variant detection. Diabetes 63: 2172–2182. pmid:24520119
- 48. Keith SW, Fontaine KR, Pajewski NM, Mehta T, Allison DB (2011) Use of self-reported height and weight biases the body mass index-mortality association. Int J Obes (Lond) 35: 401–408.