"Adolescent BMI at Northern Israel –from trends, to associated variables and comorbidities, and to medical signatures" by Yossy Machluf et al.
Supplemental Content 1
The Israeli Defense service law
Most adolescents in Israel are obliged to service by law. The main populations which do not present themselves at local recruitment centers for medical examination and profiling, and are exempt from service, are: females who are married, pregnant or mothers, religious populations (while orthodox Jewish males can postpone their service, females can be released based on testimony to their religious lifestyle), and minorities (only Druze and Circassian males are obliged to service by law, whereas males of other minorities can volunteer. Females of all minorities are exempt from service). Therefore, these (mainly female) populations are under-represented in this study population. Consequently, findings among females are expected to relate mainly to the Jewish secular to traditional sectors.
Supplemental Content 2
The medical process and the computerized database
The medical process at the recruitment center involves preliminary documentation from the primary care physician at his/her health maintenance organization (HMO), and examinations that include blood pressure, pulse rate, visual acuity, and color vision testing—all performed by trained medical personnel. Weight and height are measured by trained technicians. The same models of altimeters and scales are used and instruments are calibrated and checked for accuracy by experienced technical personnel at regular intervals. Urinalysis is performed with a dipstick and read by a laboratory technician. If the dipstick analysis is abnormal, the urine sediment is examined by high-power microscopy. Abnormal findings in urine microscopy warrant further investigation. Examination by the physician at the recruitment center includes a systematic and thorough anamnesis including family history, habits, psychological evaluation, complete physical examination and referral if necessary for further investigation according to findings. Finally,a profile and appropriate Functional Classification Codes (FCCs) that describe medical status (condition and severity, as well as an element of occupation) are assignedto each recruit and stored in a computerized database. The FCCs are organized in sections dealing with medical conditions in the standard profiling manual used for Israeli Defense Force (IDF) recruits and soldiers. FCCs are similar to the international classification of disease (ICD) coding.
The computerized database contains further personal and demographic information about each recruit,as recorded by the Ministry of Internal affairs and validated (and updated) by interview and questionnaire. The data consist of identity number, date of birth, date of arrival to Israel (if not born in Israel), country of origin determined by paternal country of origin, gender, religion, parents' year of birth and place of residence. Further data include number of years of education as verified by personal interview and by documentation from the educational institution, intelligence score as determined by a personal psychometric test, number of children in the family and the recruit's place among them, parents' occupations.
There are various parameters that represent different aspects of education. The main one, used in this study, takes into account the number of years of study, the course of study and the level of the lasteducational institution.
Supplementary Content 3
Age and BMI distributions of the study population
BMI categories were defined according to gender-related percentiles for 17-year-old BMI on growth charts of the Israeli Ministry of Health.Below are charts presenting distributions of age (upper panel) and BMI (lower panel).
Supplementary Content4
Criteria for the diagnosis of each medical condition
As this study is aimed at providing an overarching characterization of the medical signature of populations stratified by gender and BMI categories, diverse medical conditions were grouped together, while focusing on the more common medical conditions among conscripts. No self- reported measurements were used in this study. The medical conditions were as follows:
Allergic rhinitis: Diagnosis was made according to previous documentation verified by ear, nose and throat (ENT) specialist or allergy specialist. Diagnosis required chronic rhinitis associated with allergens and/or season of year.
Asthma: Asthma diagnosis was established by previous medical documentation from the primary physician and thorough anamnesis, physical examination, and spirometry results together with expert pulmonologist examination. Only active asthma was included in the present analysis (wheezing, chronic cough, dyspnea on effort and or upon exposure to allergens less than 3 years before examination period).
Blood pressure (BP): Systolic and diastolic BP measures (herein SBP and DBP, respectively) were obtained by a sphygmomanometer, in a relaxed, sitting position. Categories were defined following age-adjusted (17 years old) and height-adjusted (174 cm and 162 cm for males and females, respectively) definitions of the Ministry of Health. Seven categories were defined: first, these with missing BP (either DBP=0 or SBP=0) were classified; followed by hierarchically classification into 1 of 5 abnormal BP categories, based on clinical severity,abnormal BP(either DBP=1–39 or SBP=1–79 or SBP-DBP<20),hypertension type II (>99th percentile + 5 mmHg:either DBP>90 or SBP>146/139 for males/females or recruits belonging to 1 of the 3 BP categories indicated below and having FCC indicative of taking medications due to hypertension type II), hypertension type I (≥95th percentile and ≤99th percentile + 5 mmHg: DBP=86–90 or SBP=135–146/129-139 for males/females), pre-hypertension (≥90th percentile and <95th percentile + 5 mmHg: DBP=80–85 or SBP=120–134/120–128 for males/females), hypotension(DBP=40–49 or SBP=80–89); and finally these with normal BP(DBP=50–79 and SBP=90–119) were classified. Diagnosis of hypertension also warranted further workup, including 10 BP measurements on alternate days, renal function tests, eye fundus examination, and cardiac transthoracic echo. Lipid profile, blood glucose, BMI and family history of cardiovascular disease were also examined to access risk factors.
Cardiac anomalies: Cardiac anomalies were diagnosed according to cardiologist examination and Echo Doppler study. Cardiac anomalies included valvular abnormalities such as bicuspid aortic valve, aortic valve insufficiency or stenosis, mitral valve insufficiency or stenosis, pulmonary stenosis or insufficiency, tricuspid valve insufficiency or stenosis. Non-valvular anomalies included patent ductus arteriosus, atrial septal defect (ASD), ventricular septal defect (VSD), coarctation of the aorta, tetralogy of Fallot, complete transposition of great vessels and dextrocardia.
Diabetes mellitus: Diabetes mellitus diagnosis was established by fasting blood glucose levels above 126 mg/dL for those examined after 1998, and above 140 mg/dL for those examined during previous years in repeated analysis, and or 2 h post-prandial glucose levels above 200 mg/dL. The change in the fasting glucose level threshold was made according to changes in definition as decided by the WHO and NIH.
Endocrine disorders: Endocrine disorders were diagnosed by an endocrinologist according to accepted criteria and included mostly thyroid gland disease, hormonal deficiencies, adrenal gland disorders, diseases of the pituitary gland, and parathyroid disorders.
Flat foot: Flat foot assessment included anamnesis and examination in standing position, standing on toes and assessment of foot arch flexibility, along with sub-talar movement and existence of valgus of heels in the standing position. Only severe flat feet were included in the analysis.
Gastrointestinal disorders: Gastrointestinal disorders were diagnosed by a gastroenterologist. They included gastro-esophageal reflux, inflammatory bowel diseases (Crohn's disease and ulcerative colitis), irritable bowel, celiac, liver diseases such as chronic hepatitis B or C, fatty liver, autoimmune hepatitis, cirrhosis, biliary tract disorders such as biliary calculi and peptic ulcers.
Hematological malignancy: Hematological malignancies were diagnosed at hemato-oncological units and included exact histopathology, stage and grade, bone marrow biopsy results, treatment protocols (chemotherapy, radiation, biological treatment, bone marrow transplant, etc.). Hematological malignancies included mainly acute leukemias (mostly ALL) and Hodgkin and non-Hodgkin lymphomas.
Hyperlipidemia: Hyperlipidemia was diagnosed as low-density lipoprotein (LDL) cholesterol levels above 150 mg/dL and/or triglyceride levels above 200 mg/dL.
Knee disorders: Knee disorders were verified by an orthopedic surgeon, and included: chondromalacia of patellae, meniscal derangements, and partial or complete ligament tears as diagnosed by an orthopedic surgeon.
Learning disorders/ADHDs: Learning disabilities and attention-deficit disorder (ADD)/attention-deficit hyperactivity disorder (ADHD) were diagnosed by specific psycho-diagnostic tests and verified by neurologist and or psychiatrist examination. In this analysis, only the severe learning disorders were included, such those accompanied with dyspraxia or coordination problems, or severe ADD/ADHD requiring daily consumption of medication (Ritalin or Concerta).
Malignant solid tumors: Malignant solid tumors were diagnosed by oncologists and verified by documentation from hospitals/oncological units, including histopathology, stage, grade, treatment protocols, clinical course, involvement of other organs and/or side effects of chemotherapy/radiotherapy and present condition. Solid malignancies included Wilms' tumor, neuroblastoma, testicular tumors, sarcomas, malignant brain tumors, gynecological malignancies among females and epithelial tumors,among others.
Mental disorders: Mental disorders were diagnosed by a psychiatrist, and severity was established according to criteria of the10th revision of the international classification of diseases (ICD)-10 of the WHO.Mental disorders included personality disorders, neuroses, adjustment disorders, post-traumatic stress disorders, psychosis, mood disorders, mental retardation, pervasive developmental disorders, eating disorders, somatoform disorders, alcohol and drug abuse.
Micro-hematuria: A diagnosis of isolated microscopic hematuria has been described in detail elsewhere1.Briefly,the definition was based on three urine samplesanalyzed in succession with a positive dipstick for blood and microscopic analysis of urinary sediment containing three or more erythrocytes per high-power-field examination, with a workup that included imaging of kidneys and urinary bladder, renal function test and a nephrologist's verification of diagnosis. It is also important to state that only isolated cases of microscopic hematuria were included as assessed by the urine results in the computerized database. Only those who had both positive blood urine according to the database and a specific FCC that indicated diagnosis of isolated hematuria were included in the analysis.
Migraine: The diagnosis of migraine relied on at least 1 year's documentation of headaches with typical characteristics of a migraine, which was verified by a neurologist.
Proteinuria: Proteinuria was defined as three successive urine samples that were positive for protein according to dipstick analysis, and the amount of more than 200 mg protein in 24 h urine collection, along with normal imaging of kidneys and bladder and normal renal function tests, and having a specific FCC that indicated proteinuria diagnosis. Diagnosis was verified by a nephrologist.
Refractive errors: Refractive errors were determined by eyesight examination by trained technical personnel and establishment of a refractive error by further examination by an optometrist or ophthalmologist. These included myopia, hypermetropia, anisometropia, amblyopia and astigmatism. Severity was determined by the finding of refraction and visual acuity after correction as well as assignment of FCC.
Scoliosis/Kyphosis: Spine deformities included scoliosis and kyphosis as established by clinical examination and X- ray imaging.
Sleep disorders: Sleep disorders included mainly sleep apnea syndrome and somnambulism diagnosed at a certified sleep disorders laboratory and verified by a sleep disorders specialist.
Supplemental Content 5
The main socio-demographic characteristics of the excluded and included study populations
The computerized medical database was established in 1987. The initial study population consisted of all records included in the database until 2010 (24 years, 158,255 records). Inclusion criteria were only set to enable the analysis of BMI trends and related associations, among adolescents. Records which met the inclusion criteria (having both height and weight measures as well as a completed medical profile at the age of 16–19 [inclusive] years of age) were included in the final study population (113,694 records). Of the 44,561 records (28.15% of initial study population) which were missing major data (height/weight measures, medical profile) or not defined as adolescents at the time of medical inspection (16–19 years old), and therefore were omitted: 29,227 subjects had not completed the medical profiling process,12,049 subjects completed their medical profiling process but not at 16–19 years of age, and 3,285records had a proper medical profile but lacked readings of either height or weight (or both). These criteria, illustrating a unified, strict andnon-biased selection approach, can be viewed as technical ones (selecting subjects with valid BMI which was determined at a given age), rather than medical or demographic, and thus are not expected to affect the socio-demographic characteristics of the final study population.
The excluded and included populations were very similar in most of their socio-demographic characteristics. Although minor differences were statistically significant, due to the very large size of the study, they were not of clinical or medical importance. Furthermore, all differences can be explained by the definition of the respective desired study population of adolescents (such as age at medical profiling) and of sub-populations exempt from service according to the IDF's service law (such as religious subjects).
For example, females comprise 41.2% and 41.4% of the initial and final study populations, respectively. Out of the 18,052 records of females which were omitted, 16,723 records (92.6%) lacked a medical profile, reflecting the exemption of females from compulsory service based on family status (married, pregnant or mothers), religious reasons (females can be released based on testimony to their religious lifestyle), or belonging to minorities (females of all minorities are exempt from service). Therefore, and as expected in light of the IDF's service law, females are under-represented in the study population.
Among the minorities/religion group, except Jews, only Druze and Circassian males are obliged to service by law, whereas males of other minorities can volunteer. As expected, Jews comprise 83.9% and 87.6% of the initial and final study populations, respectively. While Druze and Circassians comprise 3.4% of both study populations, Arab Christians and Muslims comprise 6.1% of the initial study population but only 4.2% of the final one. Subjects belonging to this minority group usually complete their medical profiling process, but not at 16–19 years of age, are under-represented in the included study population.
Recruits who were born between the years 1970 and 1993 comprise 93.2% and 99.2% of the initial and final study populations, respectively. Most of the recruits who were born before 1970 (8,315 subjects) either lacked a medical profile (2,469 subjects, 29.7%) or had completed their medical profiling process but not at 16–19 years of age (5,604 subjects, 64.7%), representing mainly cadets of military high schools or older volunteers for service and new immigrants. In addition, 71.3% of the recruits who were born since 1994 (2,385 subjects) did not completed the medical profiling and therefore lacked a medical profile.
Subjects who leave in urban environment comprise 52.5% and 49.9% of the initial and final study populations, respectively. Most ultra-orthodox male (and female) Jews who are exempt from service live in specific cities, and therefore 62% of the subjects who lacked a medical profile (18,114 of 29,227) belong to the urban environment.
Immigrants comprise 21.5% and 15.9% of the initial and final study populations, respectively. Many older immigrants had not completed the profiling process due to their age or family status, or did but at older than 19 years. Of the 16,037 new immigrants who were not included in the final study population, 7,760 subjects lacked a medical profile (48.4%) and 7,743 had completed their medical profiling process but not at 16–19 years of age (48.3%). Therefore, also immigrants are under-represented in the final study population, mainly because we focused on adolescents!
Medium SES subjects comprise 25.8% and 26.0% of the initial and final study populations, whereas lower SES subjects comprise 8.8% and 6.5%, respectively. The population that is exempt from service, mainly ultra-orthodox Jews, minorities and older new immigrants, belong mainly to the lower socio-economic strata, and therefore lower SES subjects are over-represented in the excluded population.
Altogether, the initial and final study populations were very similar in most of their socio-demographic characteristics. The minor differences may be attributed to characteristics of populations missingmajor and essential data (BMI, profile) or not considered as adolescents (the age of medical profiling), in light of IDF's service law. Nevertheless,aunified, strict and non-biased selection approach was taken, and the large final study population properly represents the target population.
Supplemental Content 6
Description of hierarchical clusters and discussion of medical signatures
The hierarchical clustering analysis uncovered clusters of medical conditions with distinctive associations to BMI categories (Figure 3). In general, 2 main clusters were found, in which a higher risk of developing a medical condition was manifested among either the underweight BMI category (Cluster#1) or theabove-normal BMI categories (Cluster#2). Cluster#1 was further divided into 3 sub-clusters, where the higher risk of developing a medical condition was manifested among the underweight BMI of I) both genders (cardiac disorders, scoliosis/kyphosis, mental disorders), where it was accompanied by a protective effect of developing only the first two conditions among above-normal BMI categories, or II) males only (gastrointestinal disorders, allergic rhinitis, proteinuria), or III) males only (endocrine disorders, active asthma, micro-hematuria), where it was accompanied by a higher risk of developing the medical condition among the above-normal BMI categories, mainly in females. Cluster#2 was further divided into 2 sub-clusters, where the higher risk of developing a medical condition was manifested among above-normal BMI categories of both genders, and it was accompanied by either I) a higher risk (refractive errors, learning disorders, knee disorders) or II) a protective effect (hypertension type I&II, flat foot, pre-hypertension) of developing the conditions the among underweight BMI category of males only or both genders, respectively.