Timing of Spermarche and Menarche Are Associated With

PHRP346_proof - 19 May 2016 - 1/7 Osong Public Health Res Perspect 2016;--(-):--e-- http://dx.doi.org/10.1016/j.phrp.2016.04.010 pISSN 2210-9099 eISSN 2233-6052

64 65 1 66 2 67 3 - ORIGINAL ARTICLE - 68 4 69 5 70 6 71 7 72 8 73 9 Timing of Spermarche and Menarche are 74 10 75 11 Associated with Physical Activity and Sedentary 76 12 77 13 Behavior Among Korean Adolescents 78 14 79 15 80 16 a, b,c c 81 Q10 17 Eun-Young Lee *, Roman Pabayo , Ichiro Kawachi 82 18 83 19 aSedentary Living Laboratory, Faculty of Physical Education and Recreation, University of Alberta, 84 20 85 21 Edmonton, AB, Canada. b 86 22 School of Community and Health Sciences, University of Nevada, Reno, NV, USA. c 87 23 Harvard T. H. Chan School of Public Health, Department of Social and Behavioral Sciences, Boston, MA, 88 24 USA. 89 25 90 26 91 27 92 28 Received: December 4, Abstract 93 29 2015 Objectives: This study examined the timing of menarche and spermarche and 94 30 Revised: March 8, 2016 their associations with physical activity (PA) and sedentary behavior (SB) after 95 31 Accepted: April 23, controlling for body mass index (BMI). 96 32 2016 Methods: Multiple logistic regression analyses were conducted to determine 97 33 whether the timing of menarche in girls and spermarche in boys is associated 98 34 with PA and SB independent of BMI in a nationally representative sample of 99 35 KEYWORDS: Korean adolescents (13e18 years; N Z 74,186). 100 36 Korea Youth Risk Behavior Results: After controlling for age, family economic status, and BMI, early timing 101 37 Web-based Survey, of spermarche among boys was associated with a higher likelihood of engaging in 102 38 Korean adolescents, PA and a lower likelihood of engaging in SB for <2 hours during weekdays. By 103 39 physical activity, 104 40 contrast, boys with late timing of spermarche were less likely to engage in PA and pubertal timing, more likely to engage in SB for <2 hours. Among girls, early or late timing of 105 41 106 sedentary behavior menarche was associated with a higher likelihood of engaging in PA and a lower 42 107 likelihood of engaging in SB. 43 108 Conclusion: Timing of menarche in girls and spermarche in boys could be a 44 109 45 marker for PA and SB among Korean adolescents. To promote PA and discourage 110 46 SB among Korean adolescents, school-based, grade-speciﬁc interventions can be 111 47 tailored by the absence or presence of menarche/spermarche. 112 48 113 49 114 50 115 51 phase of adolescents, timing of puberty varies by genetic 116 52 1. Introduction 117 (e.g., sex, race, parental inﬂuence), environmental (e.g., 53 118 54 Puberty is a complex transition that involves dra- body fatness, nutrition), and socioeconomic/cultural 119 55 matic changes in several domains of human develop- factors (e.g., immigration status) [2]. For example, girls 120 56 ment including biological, physical, psychological, and generally experience puberty 18e24 months earlier than 121 57 social development [1]. Although puberty is a universal boys, and overweight/obese status in early childhood is 122 58 123 59 124 60 125 61 *Corresponding author. E-mail: [email protected] (E.-Y. Lee). 126 62 127 ª 63 Copyright 2016 Korea Centers for Disease Control and Prevention. Published by Elsevier Korea LLC. This is an open access article 128 under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article in press as: Lee E-Y, et al., Timing of Spermarche and Menarche are Associated with Physical Activity and Sedentary Behavior Among Korean Adolescents, Osong Public Health Res Perspect (2016), http://dx.doi.org/10.1016/j.phrp.2016.04.010 PHRP346_proof - 19 May 2016 - 2/7

2 E.-Y. Lee, et al

1 associated with advanced pubertal maturation [2,3]. The While much of the research is conducted in the Eu- 66 2 individual variations in puberty can be viewed by three ropean and North American contexts, studies among 67 3 aspects, namely, timing, status, and tempo [4,5]. Timing Korean adolescents are limited. In Korea, health-care 68 4 refers to relative or expected pubertal maturation at a costs associated with precocious puberty (i.e., the onset 69 5 e 70 6 given chronological age or within specific reference of signs of puberty before the age of 7 8 years in girls 71 7 groups such as school class, whereas status refers to the and 9 years in boys) have increased remarkably since the 72 8 level of development reached by an individual in terms past decade. The total cost of health care for precocious 73 9 of physical changes at a given time. Tempo of puberty puberty was approximately US $2.3 million in 2006 and 74 10 describes how quickly or slowly individuals progress US $17.9 million in 2010 [22]. This trend overlaps with 75 11 toward full sexual maturity. Among these concepts, the the increasing trend of childhood obesity in Korea, the 76 12 majority of literature that examined the relationship prevalence of which doubled from 5.4% in 1998 to 77 13 78 14 between puberty and physical and/or psychological 10.8% in 2008 [23,24]. Greater understanding of the 79 15 development used pubertal timing as an indicator of associations between pubertal timing, PA, and SB, in- 80 16 biological maturation [5]. dependent of weight status among Korean adolescents 81 17 Individual variations in the timing of puberty may may help researchers and policy makers to develop 82 18 influence the adoption of unhealthy behaviors. Specif- health-promotion strategies for adolescents during these 83 19 ically, a comparison between those who mature on time formative years (i.e., adolescence). Therefore, the pur- 84 20 and those who mature late shows that adolescents who pose of this study was to examine the associations be- 85 21 86 22 mature early are at a higher risk of exposure to several tween pubertal timing, PA, and SB after controlling for 87 23 psychological, social, and health disadvantages [6]. For BMI. It is hypothesized that adolescents who experience 88 24 example, girls who experience puberty earlier than their menarche and spermarche earlier or later than their 89 25 counterparts are more susceptible to adverse health be- peers will show negative outcomes. 90 26 haviors such as earlier alcohol use [7], cigarette smoking 91 27 [8], and eating disorder [9]. Furthermore, evidence 92 28 suggests that early timing of puberty among girls is also 93 29 2. Materials and methods 94 30 associated with decline in physical activity (PA) 95 e 31 [10 13], and an increased time spent in sedentary 2.1. Data/sample 96 32 behavior (SB) [14]. Although supporting evidence for Data collected from the eighth Korea Youth Risk 97 33 boys is lacking, Cumming and his colleagues [15] pro- Behavior Web-based Survey (KYRBS) in 2012 were 98 34 posed a biocultural model of maturity, which suggests used for the analysis. KYRBS is an annual, cross- 99 35 that antecedent biological variables (e.g., sexual matu- sectional, nationwide school-based web survey that 100 36 ration, pubertal timing, changes in body composition) monitors health risk behavior among Korean adoles- 101 37 e 102 38 influence different contexts of PA (e.g., energy expen- cents in Grades 7 12; respondents were recruited using 103 39 diture, health-related fitness, skill proficiency, sport a stratified multistage probability sampling design [25]. 104 40 participation, and performance) directly and indirectly Before survey administration, consent was obtained 105 41 via psychosocial variables (e.g., physical self-concept, from the participating school boards, individual schools, 106 42 body image dissatisfaction, self-esteem). and teachers. In June 2012, students completed a self- 107 43 One of the potential mechanisms explaining the as- administered, 129-item questionnaire in a computer 108 44 sociation between high endogenous sex hormone (i.e., laboratory under the supervision of teachers assigned by 109 45 110 46 estrogen/testosterone) and early puberty is overnutrition principals in each school during regular school hours. 111 47 in early childhood [2,16]. It is particularly well- Before beginning the online survey, students were asked 112 48 documented among girls that increased adiposity may to read the research information letter, which indicated 113 49 trigger estrogen production and lead to the early onset of that the participation in this survey is anonymous and 114 50 menarche [2,3]. Only a few studies have examined the voluntary. The survey is designed to take approximately 115 51 relationship between puberty and body mass index 40e45 minutes to complete. All surveys included a set 116 52 (BMI) among boys by using different measures [e.g., of questions that were supplemented with additional 117 53 118 54 voice break, age at onset of pubertal growth spurt, peak focus questions to gain further information on specific 119 55 height velocity, public hair growth, testicular volume, issues. The core information collected includes de- 120 56 and/or penis length], and the results reported have been mographic background (e.g., age, sex), health behaviors 121 57 inconsistent [17e21]. However, a recent study exam- (e.g., smoking, alcohol use, PA), and health outcomes 122 58 ining the trend of age at spermarche and its association (e.g., self-reported health, obesity). A total of 76,980 123 59 with BMI among Chinese school boys found that a students from 400 middle and 400 high schools partic- 124 60 higher BMI or BMI-for-Age z-score was associated with ipated (response rate: 96.4%). In this study, 74,186 125 61 126 62 an increased likelihood of having reached spermarche, students (48.5% of girls) who completed the survey 127 63 indicating the overlapping trend of earlier age at sper- were included in the analysis after excluding those with 128 64 marche with increase in BMI over the past 15 years missing scores in height and weight, and those aged < 129 65 among Chinese boys [3]. 12 years or aged > 19 years (n Z 2,794). Weights were 130

Q1 Pubertal timing and health behavior 3

1 assigned to each respondent to give an equal probability chatting with friends during the last 5 weekdays (2) and 66 2 of being sampled from the entire Korean population on weekends?” Response options were ranged from 1 67 3 aged between 12 years and 19 years. The survey pro- (<1 h/d) to 6 (>4 h/d). 68 4 tocol was approved by Korea Centers for Diseases 69 5 70 6 Control and Prevention (approval number: 11758) [25]. 2.2.4. Covariates 71 7 Family socioeconomic status and chronological age 72 8 2.2. Measures were included as covariates. Family socioeconomic 73 9 2.2.1. Pubertal timing status was measured by asking participants to rate their 74 10 Individual menarcheal and spermarcheal data were economic status from 1 (low) to 4 (high). 75 11 collected by the status quo method. The question was 76 12 “When did you experience your first menstruation (i.e., 2.3. Statistical analysis 77 13 78 menarche)/ejaculation (i.e., spermarche)?” Response 14 All analyses were performed for boys and girls 79 15 options were ranged from 1 (have not yet experienced) separately. The population weight provided by KYRBS 80 16 to 14 (Grade 12). For data analysis, the average of pu- was applied to estimate a representation of the target 81 17 bertal timing was calculated except for those who have population. Descriptive statistics were calculated and 82 18 not yet experienced menarche/spermarche; the average expressed as means and standard deviations or as per- 83 19 grade of puberty was Grade 7 (8.13 1.76) for boys and centages. To examine the associations between relative 84 20 85 Grade 6 (7.27 1.27) for girls. Students were then pubertal timing, BMI, and PA/SB, a series of logistic 21 86 categorized into four pubertal timing groups by sex: regression models stratified by sex were conducted. All 22 e 87 23 Group 1 (menarche in Grades 1 5 for girls; spermarche analyses were controlled for age and family socioeco- 88 24 in Grades 1e6 for boys), Group 2 (average: menarche in nomic status. Models examining the associations be- 89 25 Grade 6 for girls; spermarche in Grade 7 for boys), tween pubertal timing and health behavior were further 90 26 Group 3 (menarche in Grades 7e8 for girls; spermarche adjusted for BMI. Analyses were conducted using IBM 91 27 in Grades 8e9 for boys), and Group 4 (have not yet SPSS 21.0 and the results were reported as predicted Q2 92 28 93 experienced and maturation in or over Grade 9 for girls; odds ratio (OR) and 95% confidence interval (95% CIs). 29 94 have not yet experienced and maturation in or over 30 The alpha level was set at 0.05. 95 31 Grade 10 for boys). Menarche is a distinctive and 96 32 commonly used measure of pubertal timing among girls 97 33 [3]. First ejaculation, also known as spermarche, has 3. Results 98 34 been used to indicate pubertal timing in previous studies 99 35 [3,26]. The average grades of menarche and spermarche The sample characteristics of the study population 100 36 were similar to those reported previously [1,26]. by sex are presented in Table 1. Of the 74,186 re- 101 37 102 38 spondents, average age of all respondents was 14.94 103 39 2.2.2. Body mass index years (standard deviation Z 1.75). Relative pubertal 104 2 40 BMI (kg/m ) was calculated by self-reported height timing among respondents was described; 15.4% of 105 41 and weight. Each BMI score was then categorized into boys and 23.8% of girls experienced puberty earlier 106 42 nonoverweight (BMI percentile 0e84.9th) and over- than their peers. Boys were more physically active than 107 43 weight (BMI percentile 85.0th) based on the age- and girls. The proportion of students engaging in MVPA >5 108 44 sex-specific BMI reference data for Korean children and times/wk was 17.3% among boys and 6.1% among 109 45 110 46 adolescents [27]. girls. More girls engaged in SB 4 h/d than boys 111 47 during weekdays (16% vs. 11.9%) and on weekends 112 48 2.2.3. Physical activity/sedentary behavior (30.4% vs. 24.8%). 113 49 Moderate-to-vigorous PA (MVPA) and SB were Table 2 shows the associations between relative pu- 114 50 measured using questions adopted from the International bertal timing, PA, and SB among Korean adolescent 115 51 Physical Activity Questionnaire [28]. MVPA was boys after controlling for age, family economic status, 116 52 measured by asking participants the following ques- and BMI. Compared to boys with average pubertal 117 53 118 54 tions: “For the past seven days: (1) How many days did timing (referent group), those with early pubertal timing 119 55 you engage in activities resulting in an increase of heart were more likely to engage in MVPA for 60 minutes for 120 56 rate or experiencing shortness of breath (moderate-to- >5 times/wk (OR Z 1.04, 95% CI Z 1.02e1.05) and 121 57 vigorous) for more than 60 minutes?; (2) How many muscular strengthening exercises (OR Z 1.06, 95% 122 58 days did you engage in weight-bearing exercise?” CI Z 1.05e1.07). In contrast, boys with late pubertal 123 59 Questions were scaled from 1 (no participation at all) to timing were less likely to engage in MVPA (OR Z 0.99, 124 60 8 (7 d/wk). PA was categorized into four groups (i.e., 95% CI Z 0.97e1.00 for Group 3; OR Z 0.95, 95% 125 61 e e > Z e 126 62 none, 1 2 times, 3 4 times, 5 times). To measure SB, CI 0.94 0.96 for Group 4) and muscular strength- 127 63 the following questions were asked. “How many hours a ening exercise (OR Z 0.91, 95% CI Z 0.90e0.92 for 128 64 day: (1) did you spend your leisure time watching Group 3; OR Z 0.85, 95% CI Z 0.84e0.86). During 129 65 television, playing video game, Internet browsing, or weekdays, boys with early pubertal timing were less 130

4 E.-Y. Lee, et al

1 Z d 66 2 Table 1. Sociodemographic characteristics of the sample (N 74,186) 2012 Korea Youth Risk Behavior Web-based 67 Q9 3 Survey. 68 4 69 Total (N Z 74,186) Boys (n Z 38,221) Girls (n Z 35,965) 5 70 6 Age (y), mean SD 14.94 (1.75) 14.96 (1.75) 14.92 (1.74) 71 2 7 BMI (kg/m ), mean SD* 20.60 (3.02) 20.90 (3.28) 20.26 (7.93) 72 8 Nonoverweight (0e85%) 89.0 85.4 92.6 73 9 Overweight (85%) 11.0 14.6 7.4 74 10 Family economic status (%) 75 11 Low 30.7 33.4 27.7 76 12 Middle 46.7 44.6 49.0 77 13 High 22.6 22.0 23.3 78 14 Pubertal timing (%)y 79 15 1 19.4 15.4 23.8 80 16 2 (average) 23.3 14.8 32.6 81 17 3 25.7 18.1 34.1 82 18 4 31.6 51.7 9.5 83 19 Physical activity (%) 84 20 Moderate-to-vigorous physical activity 60 min/d 85 21 86 None 36.9 27.9 46.8 22 87 1e2 times/wk 32.6 31.7 33.5 23 88 3e4 times/wk 18.5 23.1 13.5 24 89 25 5 times/wk 12.0 17.3 6.1 90 26 Muscular strengthening exercises 91 27 None 51.7 37.1 67.8 92 e 28 1 2 times/wk 29.1 34.7 23.0 93 29 3e4 times/wk 11.4 16.3 6.0 94 30 5 times/wk 7.8 11.9 3.2 95 31 Sedentary behavior (%) 96 32 Weekdays 97 33 >2 h 52.1 54.6 49.3 98 34 2e3 h 23.0 23.3 22.7 99 35 3e4 h 11.0 10.2 12.0 100 36 4 h 13.9 11.9 16.0 101 37 Weekends 102 38 >2 h 29.9 32.3 27.2 103 39 2e3 h 24.7 25.8 23.5 104 40 3e4 h 18.0 17.1 18.9 105 41 4 h 27.5 24.8 30.4 106 42 107 *Body mass index: calculated based on the 2007 BMI-for-Age in Korean children and adolescents [27]; yPubertal timing: 1 (menarche in Grades 1e5; 43 108 spermarche in Grades 1e6); 2 (average: menarche in Grade 6; spermarche in Grade 7); 3 (menarche in Grades 7e8; spermarche in Grades 8e9); and 4 44 109 (menarche in or over Grade 9 or have not experienced yet; spermarche in or over Grade 10 or have not experienced yet). BMI Z body mass index; 45 SD Z standard deviation. 110 46 111 47 112 48 113 49 likely to engage in SB for <2 h/d (OR Z 0.94, 95% odds of reporting the participation in MVPA for >60 114 50 CI Z 0.93e0.95), whereas those with late pubertal minutes and 5 d/wk (OR Z 1.05, 95% CI Z 1.04e1.07 115 51 timing reported the opposite (OR Z 1.10, 95% CI for early timing; OR Z 1.07, 95% CI Z 1.06e1.09 for 116 52 1.09e1.11 for late timing; OR Z 1.01, 95% late timing; OR Z 1.29, 95% CI Z 1.26e1.32 for 117 53 CI Z 1.00e1.02 for delayed timing) compared with the delayed timing), muscular strengthening exercises for 118 54 > Z Z e 119 55 referent group. On weekends, boys with either early 3 d/wk (OR 1.13, 95% CI 1.11 1.15 for early 120 56 (OR Z 1.02, 95% CI Z 1.01e1.04) or late timing of timing; OR Z 1.12, 95% CI Z 1.11e1.14 for late 121 57 puberty (Group 4: OR Z 1.06, 95% CI Z 1.05e1.07) timing group 3; OR Z 1.26, 95% CI Z 1.24e1.28 for 122 58 were more likely to have SB for <2 hours compared late timing group 4), and SB for <2 h/d during week- 123 59 with the referent group. days (OR Z 1.04, 95% CI Z 1.03e1.05 for late timing; 124 60 The associations between relative pubertal timing, OR Z 1.28, 95% CI Z 1.16e1.29 for delayed timing) 125 61 PA, and SB, independent of BMI among Korean girls and on weekends (OR Z 1.04, 95% CI Z 1.03e1.05 for 126 62 Z Z e 127 63 are shown in Table 3. Among girls, early, late, and early timing; OR 1.01, 95% CI 1.00 1.02 for late 128 64 delayed timing of puberty, compared to girls with timing group 3; OR Z 1.34, 95% CI Z 1.32e1.36 for 129 65 average pubertal timing, were associated with higher late timing group 4). 130

Pubertal timing and health behavior 5

1 66 2 Table 2. Associations (OR and 95% CI) between relative pubertal timing, physical activity, and sedentary behavior after 67 * Z d 3 controlling for BMI among Korean adolescent boys (N 38,221) 2012 Korea Youth Risk Behavior Web-based 68 4 Survey. 69 5 70 1y 2 (average)y 3y 4y 6 71 7 OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) 72 8 Moderate-to-vigorous physical 5 times/wk 1.04 (1.02e1.05) 1 0.99 (0.97e1.00) 0.95 (0.94e0.96) 73 9 activity 60 min/d 74 e e e 10 Muscular strengthening 3 times/wk 1.06 (1.05 1.07) 1 0.91 (0.90 0.92) 0.85 (0.84 0.86) 75 11 exercise 76 < e e e 12 Sedentary behavior (weekdays) 2 h/d 0.94 (0.93 0.95) 1 1.10 (1.09 1.11) 1.01 (1.00 1.02) 77 13 Sedentary behavior (weekends) <2 h/d 1.02 (1.01e1.04) 1 0.99 (0.98e1.00) 1.06 (1.05e1.07) 78 14 *Body Mass Index (BMI): calculated based on the 2007 BMI-for-Age in Korean children and adolescents [27]; yPubertal timing: 1 (spermarche in Grades 79 15 1e6), 2 (average: spermarche in Grade 7), 3 (spermarche in Grades 8e9), and 4 (spermarche in or over Grade 10 or have not experienced yet). All analyses 80 16 are adjusted for chronological age, family economic status, and BMI. p < 0.05. CI Z confidence interval; OR Z odds ratio. 81 17 82 18 83 19 84 4. Discussion likely to engage in MVPA and strengthening exercises, 20 > 85 21 they were more likely to spend 2 h/d in SB during 86 22 This study examined the associations between pu- weekdays (i.e., high PA and high sedentary). By contrast, 87 23 bertal timing, PA, and SB, independent of BMI among boys with late pubertal timing were less likely to engage in 88 24 Korean adolescents. This study offers a better under- PA, but more likely to engage in SB for <2 hours (i.e., low 89 25 standing of the relationship between pubertal timing and PA and low sedentary). Given that PA and SB are likely to 90 26 91 health behavior among adolescents. Consistent with the have combined effects on health among young people 27 92 [29], Korean adolescent boys who experience puberty 28 previous findings, our results suggest that pubertal 93 29 timing is a potentially relevant marker associated with early or late may be predisposed to negative health con- 94 30 PA and SB, independent of BMI, in a representative ditions. Regardless of individual variations in weight sta- 95 31 sample of Korean adolescents. A previous study indi- tus, promoting PA among boys with late timing of puberty 96 32 cated that pubertal timing predicts psychological out- and discouraging SB among boys with early timing of 97 33 comes among girls but not among boys [5]. The results puberty may positively contribute to their health. 98 34 99 of our study add to the current literature that pubertal Early or late timing of puberty among girls was 35 100 associated with physically active lifestyle (i.e., high PA 36 timing is also an important predictor for behavioral 101 37 outcomes among both boys and girls. and low SB) relative to their peers with average timing 102 38 Specifically, early timing of puberty among boys was of puberty after controlling for weight status. This is 103 39 associated with a higher likelihood of engaging in PA, somewhat inconsistent with previous findings. For 104 40 whereas a decreased likelihood of engaging in PA was example, previous studies examining the relationship 105 41 found among boys with late pubertal timing. By contrast, between pubertal timing and PA after controlling for age 106 42 107 early timing of puberty was associated with a higher and BMI reported nonsignificant associations [30,31]. 43 108 > By contrast, in a longitudinal examination by Davison 44 likelihood of spending time in SB for 2 hours during 109 45 weekdays, whereas boys with late timing of puberty and colleagues [12], advanced maturation at 11 years of 110 46 showed the opposite (i.e., less likely to spend >2 hours in age was associated with low levels of MVPA at 13 years 111 47 SB). Although boys with early pubertal timing are more of age among 178 American girls after controlling for 112 48 113 49 114 50 115 51 Table 3. Associations (OR and 95% CI) between relative pubertal timing, physical activity, and sedentary behavior after 116 52 controlling for BMI * among Korean adolescent girls (N Z 35,965)d2012 Korea Youth Risk Behavior Web-based 117 53 Survey. 118 54 119 55 1 y 2 (average) y 3 y 4 y 120 56 OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) 121 57 122 Moderate-to-vigorous physical 5 times/wk 1.05 (1.04e1.07) 1 1.07 (1.06e1.09) 1.29 (1.26e1.32) 58 123 activity 60 min/d 59 Muscular strengthening exercise 3 times/wk 1.13 (1.11e1.15) 1 1.12 (1.11e1.14) 1.26 (1.24e1.28) 124 60 Sedentary behavior (weekdays) <2 h/d 1.00 (0.99e1.01) 1 1.04 (1.03e1.05) 1.28 (1.16e1.29) 125 61 Sedentary behavior (weekends) <2 h/d 1.04 (1.03e1.05) 1 1.01 (1.00e1.02) 1.34 (1.32e1.36) 126 62 y 127 63 *Body mass index (BMI): calculated based on the 2007 BMI-for-Age in Korean children and adolescents [27]; Pubertal timing: 1 (spermarche in Grades 128 e e 64 1 6), 2 (average: spermarche in Grade 7), 3 (spermarche in Grades 8 9), and 4 (spermarche in or over Grade 10 or have not experienced yet). All analyses 129 are adjusted for chronological age, family economic status, and BMI. p < 0.05. CI Z confidence interval; OR Z odds ratio. 65 130

6 E.-Y. Lee, et al

1 covariates including body fat, and this relationship be- the average timing of menarche/spermarche among our 66 2 tween maturation and PA was mediated by global self- sample was similar to those reported in previous studies 67 3 worth, depression, and maturity fears. It is likely that [1,3], misclassification error might have occurred when 68 4 the differences in findings across studies are due to categorizing pubertal timing into four groups. To 69 5 70 6 different measures used to measure pubertal timing and/ decrease the validity of these threats in future work, it is 71 7 or health behavior (i.e., PA and SB). Nonetheless, recommended that KYRBS should adopt more robust, 72 8 perhaps there are mechanisms explaining the relation- yet noninvasive measures of pubertal maturation (e.g., 73 9 ship between average timing of puberty and unhealthy secondary sexual characteristics, PDS, or both). 74 10 behaviors (i.e., low PA and high SB) relative to their Despite such limitations, we found that pubertal 75 11 off-time peers among Korean girls. A recent review timing was a potential relevant marker associated with 76 12 examining the correlates of PA among Korean girls PA and SB, independent of BMI among Korean ado- 77 13 78 14 indicated that traditional gender roles and gender lescents. Our results have implications for public health 79 15 inequality in physical education class may prevent girls policy. Furthermore, interventions promoting active 80 16 from being physically active [32]. Indeed, boys who living among adolescents should be tailored based on 81 17 engage in MVPA for >5 times/wk were almost three- different maturational timing with other sociodemo- 82 18 fold of girls in our study (17.3% vs. 6.1%). Girls with graphic characteristics; premenarcheal/spermarcheal 83 19 average pubertal timing may have more friends who are stage may be a critical period for health behavior for- 84 20 not physically active compared to girls with early or late mation. Furthermore, physicians and health pro- 85 21 86 22 timing of puberty. fessionals are suggested to use early or late timing of 87 23 Our study has several limitations to be acknowl- puberty as a marker for overweight and physically 88 24 edged. We used a self-administered questionnaire, inactive lifestyle among Korean adolescent boys. It is 89 25 which has been noted as a weakness in comparison to recommended to promote adopting physically active 90 26 direct or objective measures; however, the large sample lifestyle among boys with early or late timing of pu- 91 27 size provides representative estimates of Korean ado- berty. Specifically, discouraging overall SB should be 92 28 lescents. Future studies may benefit from incorporating focused on boys with early pubertal timing and pro- 93 29 94 30 the objective measure of PA and SB such as acceler- moting PA should target boys with late pubertal timing. 95 31 ometer. The study also relied on a single-item, self- In addition, it may help researchers to better understand 96 32 report to assess pubertal timing. Grade at menarche/ how relative pubertal timing influences health behavior 97 33 spermarche refers to either the past or to an event that by incorporating mediating or moderating variables (i.e., 98 34 one has not experienced yet. Using other measures of psychosocial factors associated with puberty) in future 99 35 pubertal timing (e.g., subjective pubertal timing, sec- studies. Longitudinal tracking is also required for future 100 36 ondary sex characteristics) or other dimensions of pu- studies. It is still unclear whether pubertal timing in- 101 37 102 38 berty [e.g., pubertal status using age at peak height fluences the decline in PA and increase in SB. 103 39 velocity, tanner stage, Pubertal Development Scale Early or late timing of puberty could be a relevant 104 40 (PDS)] may provide more insight on this topic than marker for physically inactive lifestyle among Korean 105 41 direct inquiry about the onset of one event. In addition, adolescent boys. The results of our study has some 106 42 although chronological age can be a better measure than consistent and inconsistent findings compared with 107 43 school grade for pubertal timing, KYRBS does not existing literature mostly from Western countries; 108 44 provide full information in age (i.e., year, month, and nonetheless, the present study makes a unique contri- 109 45 110 46 date or birth), and therefore, it was impossible to provide bution to the limited existing research examining pu- 111 47 age-specific data. However, school grade can be a better bertal timing and health behavior, particularly among 112 48 predictor of PA and SB in the Korean context given the adolescent boys. School-based interventions to promote 113 49 Asian age reckoning and school system. Specifically, PA and discourage SB among Korean adolescents 114 50 newborns start at the age of 1 year, and commonly go to should take individual differences in pubertal timing 115 51 elementary school at the age of 8 years in Asian age into consideration. 116 52 reckoning. As a result, Korean students spend most of 117 53 118 54 their time with their peers in the same grade, rather than 119 55 with peers in the same chronological age per se. Conflicts of interest 120 56 Furthermore, previous studies have reported that girls 121 57 and boys may respond differently to the onset of All authors have no conflicts of interest to declare. 122 58 menarche and spermarche based on how prepared they 123 59 are for the event [33,34]. For instance, boys tend to 124 60 report that they have experienced spermarche when they 125 61 Acknowledgments 126 62 have not because it reflects muscularity. Nonetheless, 127 63 the only available datum regarding pubertal timing was 128 64 menarche for girls and spermarche for boys in the self- The authors thank the Korea Centers for Disease 129 65 reported data from the KYRBS. In addition, although Control and Prevention, Ministry of Health and Welfare, 130

Pubertal timing and health behavior 7

1 and Ministry of Education, Science, and Technology for 19. De Leonibus C, Marcovecchio ML, Chiavaroli V, et al. Timing of 61 2 releasing and providing the data. puberty and physical growth in obese children: a longitudinal 62 3 study in boys and girls. Pediatr Obes 2014 Aug;9(4):292e9. 63 20. Kryst q, Kowal M, Woronkowicz A, et al. Secular changes in 4 References 64 5 height, body weight, body mass index and pubertal development in 65 6 male children and adolescents in Krakow, Poland. J Biosoc Sci 66 e 7 1. Patton GC, Viner R. Pubertal transitions in health. Lancet 2007 2012 Jul;44(4):495 507. 67 Mar;369(9567):1130e9. 21. Sørensen K, Aksglaede L, Petersen JH, et al. Recent changes in 8 68 2. Kaplowitz PB. Link between body fat and the timing of puberty. pubertal timing in healthy Danish boys: associations with body 9 69 Pediatrics 2008 Feb;121(3):S208e17. mass index. J Clin Endocrinol Metab 2010;95(1):263e70. 10 70 3. Song Y, Ma J, Wang HJ, et al. Age at spermarche: 15-year trend 22. Health Insurance Review and Assessment Service. Precocious 11 71 and its association with body mass index in Chinese school-aged puberty, early detection important, maybe my child? [Internet]. 12 Q3 boys. Pediatr Obes; 2016. in press. Seoul, Korea: Health Insurance Review and Assessment Service; 72 13 4. Coelho-e-Silva MJ, Valente-dos-Santos J, Figueiredo AJ, et al. 2015. Available from: http://www.hira.or.kr/dummy.do? 73 14 Pubertal Status: Assessment, interpretation, analysis. J Sports Med pgmidZHIRAA020041000000&cmsurlZ/cms/notice/02/ 74 15 Q4 Doping Stud 2013;3:1e2. 1205871_13390.html [accessed 10.11.15]. 75 16 5. Marceau K, Ram N, Houts RM, et al. Individual differences in 23. Kim HM, Park J, Kim HS, et al. Obesity and cardiovascular risk 76 17 boys’ and girls’ timing and tempo of puberty: modeling devel- factors in Korean children and adolescents aged 10e18 years from 77 18 opment with nonlinear growth models. Dev Psychol 2011 Sep; the Korean National Health and Nutrition Examination Survey, 78 19 47(5):1389e409. 1998 and 2001. Am J Epidemiol 2006 Oct;164(8):787e93. 79 20 6. Kaplowitz PB, Slora EJ, Wasserman RC, et al. Earlier onset of 24. Korea Centers for Disease Control and Prevention. National health 80 21 puberty in girls: relation to increased body mass index and race. statistics in 2008 based on the Fourth Korea National Health and 81 22 Pediatrics 2001 Aug;108(2):347e53. Nutrition Examination Survey (KNHANES V) [Internet]; 2011. Q6 82 23 7. Magnusson D, Stattin H, Allen VL. Differential maturation among 25. Korea Centers for Disease Control and Prevention. Reports on the 83 d 24 girls and its relations to social-adjustment a longitudinal 8th (2012) Korea Youth Risk Behavior Web-based Survey. Seoul: 84 Q5 e 25 perspective. Life-Span Dev Behav 1986;7:135 73. Korea Centers for Disease Control and Prevention; 2012. 85 26 8. Westling E, Andrews JA, Peterson M. Gender differences in pu- 26. World Health Organization. The sexual and reproductive health of 86 27 bertal timing, social competence, and cigarette use: a test of the younger adolescents: research issues in developing countries 87 early maturation hypothesis. J Adolesc Health 2012 Aug;51(2): [Internet]. Geneva (Switzerland): World Health Organization. 28 88 150e5. Available from: http://www.corhaethiopia.org/docs/AYRH/The% 29 89 9. Harden KP, Mendle J, Kretsch N. Environmental and genetic 20Sexual%20and%20Reproductive%20Health%20of%20Younger 30 90 pathways between early pubertal timing and dieting in adoles- %20Adolescents-Research%20Issues%20in%20Developing% 31 91 cence: distinguishing between objective and subjective timing. 20Countries.pdf [accessed 1.11.15]. Q7 32 Psychol Med 2012 Jan;42(1):183e93. 27. Moon JS, Lee SY, Nam CM, et al. 2007 Korean National Growth 92 33 10. Baker BL, Birch LL, Trost SG, et al. Advanced pubertal status at Charts: review of developmental process and an outlook. Korean J 93 34 age 11 and lower physical activity in adolescent girls. J Pediatr Pediatr 2008;51:1e25. 94 35 2007 Nov;151(5):488e93. 28. Craig CL, Marshall AL, Sjo¨stro¨m M, et al. International physical 95 36 11. Cumming SP, Standage M, Gillison F, et al. Sex differences in activity questionnaire: 12-country reliability and validity. Med Sci 96 37 exercise behavior during adolescence: is biological maturation a Sports Exerc 2003 Aug;35(8):1381e95. 97 38 confounding factor? J Adolesc Health 2008 May;42(5):480e5. 29. Chastin SF, Palarea-Albaladejo J, Dontje ML, et al. Combined 98 39 12. Davison KK, Werder JL, Trost SG, et al. Why are early maturing effects of time spent in physical activity, sedentary behaviors and 99 40 girls less active? Links between pubertal development, psycho- sleep on obesity and cardio-metabolic health markers: a novel 100 41 logical well-being, and physical activity among girls at ages 11 compositional data analysis approach. PLoS One 2015 Oct 13; 101 42 and 13. Soc Sci Med 2007 Jun;64(12):2391e404. 10(10):e0139984. 102 43 13. Sherar LB, Esliger DW, Baxter-Jones A, et al. Age and gender 30. Drenowatz C, Eisenmann JC, Pfeiffer KA. Influence of socio- 103 44 differences in youth physical activity: does physical maturity economic status on habitual physical activity and sedentary 104 e 45 matter? Med Sci Sports Exerc 2007 May;39(5):830 5. behavior in 8- to 11-year old children. BMC Public Health 2010 105 46 14. Bradley CB, McMurray RG, Harrell JS, et al. Changes in common Apr;10:214. 106 47 activities of 3rd through 10th graders: the CHIC study. Med Sci 31. Bradley RH, McRitchie S, Houts RM, et al. Parenting and the 107 Sports Exerc 2000 Dec;32(12):2071e8. decline of physical activity from age 9 to 15. Int J Behav Nutr 48 108 15. Cumming SP, Sherar LB, Pindus DM, et al. A biocultural model of Phys Act 2011 Apr;8:33. 49 109 maturity-associated variance in adolescent physical activity. Int 32. Lee EY, Spence JC, Song YK, et al. A systematic review of the 50 110 Rev Sport Exerc Psychol 2012;5(1):23e43. correlates of Korean girls’ physical activity: an ecological 51 111 16. Davison KK, Susman EJ, Birch LL. Percent body fat at age 5 perspective. Health Soc Sci; 2016. in press. Q8 52 predicts earlier pubertal development among girls at age 9. Pedi- 33. Ruble DN, Brooks-Gunn J. The experience of menarche. Child 112 53 atrics 2003 Apr;111(4 Pt 1):815e21. Dev 1982;53(6):1557e66. 113 54 17. Wagner IV, Sabin MA, Pfa¨ffle RW, et al. Effects of obesity on 34. Laron Z. Age at first ejaculation (spermarche)dthe overlooked 114 55 human sexual development. Nat Rev Endocrinol 2012 Jan;8(4): milestone in male development. Pediatr Endocrinol Rev 2010 115 56 246e54. MareApr;7(3):256e7. 116 57 18. Aksglaede L, Juul A, Olsen LW, et al. Age at puberty and the 117 58 emerging obesity epidemic. PLoS One 2009 Dec;4(12):e8450. 118 59 119 60 120