a prospective longitudinal study of urinary excretion of a bone resorption marker in adolescents
TRANSCRIPT
A prospective longitudinal study of urinary excretion of a boneresorption marker in adolescents
A. M . BOLLEN
Departments of Orthodontics and Orthopaedics, University of Washington, Seattle, USA
Received 25 February 1999; revised 18 June 1999
Summary. In growing subjects, the rates of bone resorption and bone deposition aresubstantially larger than in non-growing individuals. The purpose of this study was tomeasure the urinary excretion of a speci® c bone resorption marker in function of adolescentgrowth stages in a prospective longitudinal study. A cohort of 60 adolescents (28 male and32 female) was followed for 3.4 years (range 1.7± 4.6 years). Monthly measurements ofheight, weight and urinary excretion of a bone resorption marker, collagen type I N-telo-peptides (NTx), were made. Changes in standing height were used to classify the adolescentsinto one or more of six adolescent growth stages: pre-pubertal growth (continuous moderategrowth rate), ascending growth spurt (increasing growth rate), peak growth spurt (growthrate higher than 7 cm/year for at least 6 months), descending growth spurt (continuousdecrease in growth rate), end of growth (growth rate between 0 and 2 cm/year), and nogrowth. An increase in NTx excretion from the pre-pubertal to peak growth spurt of about33% was found (44% and 27% for females and males respectively). The decreasing growthrate after the pubertal growth spurt coincided with a clear decrease in NTx excretion. Thesedi� erences were statistically signi® cant, except between the prepubertal and ascendinggrowth stage. Individual mean NTx excretion during each growth stage was correlatedwith the individual’s growth rate during that time (r ˆ 0:81). There was large inter-andintra individual variability. In non-growing adolescents (growth rate 0 cm/y) NTx excretionlevels were 4± 7 times greater than in adults. In all females, menarche was followed by adecrease in NTx excretion. In conclusion, the excretion of a speci® c bone resorption marker,NTx, was correlated with the changes in growth rate during adolescence, both for males andfemales. There were large inter and intra-individual di� erences in NTx excretion during thedi� erent growth stages. In adolescents who reached their adult height at the end of thepubertal growth spurt, bone resorption decreased dramatically but remained 4± 7 fold higherthan in adults.
1. Introduction
During growth, the rate of bone turnover is greatly increased. This has beenshown in cross-sectional studies measuring age-related changes in metabolic markersof bone deposition (Johansen, Giwercman, Hartwell et al. 1988, Trivedi, Risteli,Risteli et al. 1991, Saggese et al. 1992) and bone resorption (Beardsworth, Eyreand Dickson 1990, Bluhmson, Hannon, Wrate et al. 1994, Bollen and Eyre 1994).These cross-sectional studies indicate greater marker measurements at times ofgreater growth. Few longitudinal studies have correlated marker excretion withgrowth rate on an individual basis rather than on a population basis. One long-itudinal study in healthy infants (aged 1± 18 months) found that growth velocityduring the three months preceding the sample collection correlated with serum mar-
ANNALS OF HUMAN BIOLOGY, 2000 , VOL. 27 , NO. 2, 199 ± 211
Annals of Human Biology ISSN 0301± 4460 print/ISSN 1464± 5033 online # 2000 Taylor & Francis Ltdhttp://www.tandf.co.uk/journals/tf/03014460.html
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.
kers of bone resorption and deposition (Lieuw-A-Fa, Sierra and Specker 1995). Atwo-year study on 9-year-old girls compared three yearly measures of urinary hydro-xyproline/creatinine excretion, and found greater values at age 11 than at age 9(Wake® eld, Disney, Mason et al. 1980). Bone deposition in adolescent boys increasessharply, concomitant with the increase in serum testosterone (Riis, Krabbe,Christiansen et al. 1985). In a ® ve-year study, monthly measures of hydroxyprolineexcretion in ® ve siblings was found to follow the growth velocity curve (Togo,Morinaga and Togo 1977). A sharp decline in hydroxyproline excretion wasobserved after the pubertal growth peak and after the start of menarche (Togoand Togo 1988). Hannon and colleagues similarly showed a rapid decrease inboth resorption and deposition markers after the onset of menses when comparingfour annual samples in relation to the time of menarche (Hannon, Bluhmsohn andWrate et al. 1996). An 18-month longitudinal study of adolescent females indicated amoderate relation between bone turn-over markers and height gain (Cadogan,Bluhmsohn, Barker et al. 1998).
Standing height growth rate re¯ ects changes in cartilage growth. Cartilagegrowth is followed by changes in bone metabolism: endochondral ossi® cationand bone modeling. The postnatal standing height growth rate ¯ uctuates,depending on developmental stage. It is high during the ® rst two years of lifeand decreases to a pre-pubertal low. Some childhood growth spurts may happenin between. After the pre-pubertal low growth rate there is an increase (thepubertal growth spurt), which is followed by a decrease, and eventually thegrowth rate declines to zero. The age at which individuals experience thesedi� erent growth stages is variable, but generally earlier in females than in males.No longitudinal study has followed adolescents during the various adolescentgrowth stages to investigate the correlation between growth rate and bone metab-olism on an individual basis.
The purpose of this prospective study was to describe the urinary excretion of aspeci® c bone resorption marker in function of adolescent growth stage in a long-itudinal study of males and females. Monthly measurements of standing height andweight, and collection of a urine sample were made on a cohort of healthy adoles-cents. A urine sample was chosen over a serum sample, due to the frequent samplingand ease of collection. Since currently no urinary markers of bone deposition areavailable, bone resorption was evaluated. The bone resorption marker used in thisstudy, collagen type I N-telopeptide crosslink (NTx) is speci® c to the osteoclasticresorption process (Apone, Lee and Eyre 1997), is highly responsive to antiresorptivetherapies (Rosen, Dresener-Pollak, Rosenblatt et al. 1994) and provides a goodmeasure of systemic bone turnover activity (Calvo, Eyre and Gundberg 1997). Inaddition, the use of NTx as a marker of systemic bone turnover during growth, hasbeen validated against calcium kinetics and calcium balance measures (Weaver,Peacock, Martin et al. 1997). Due to the longitudinal nature, and the frequentsampling proposed in this study, the collection of spot urines was chosen over24 h urine samples. The reliability of measurement of bone resorption markers inspot urines in children was shown by Fujimoto and colleagues (1995) who found asigni® cant correlation between pyridinoline excretion corrected for creatinine and 24hour pyridinoline excretion corrected for body area. Similarly, Lapilonne and co-workers (1996) found a good correlation between NTx/creatinine (in spot urines)and 24 h urines in infants.
200 A. M. Bollen
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.
2. Methods
2.1. SubjectsSubjects for this prospective study were recruited among patients and their sib-
lings from the Department of Orthodontics at the University of Washington. Theseindividuals visit the clinic on a monthly basis for a period of at least 2 years.Inclusion criterea were: aged between 8 and 12 and being pre-menarcheal forfemales; aged between 10 and 14 for males; no history of metabolic or other bonedisease, and be willing and able to collect a spot urine sample on a monthly basis.Assent was obtained from the subjects, and consent from their parents. This studywas reviewed and approved by the University of Washington’s Human SubjectsInternal Review Board.
2.2. Height and weight measurementsHeight was measured every month with a Holtain Harpenden-model Stadiometer
(Holtain Limited, Crymych, Dyfed, Great Britain). The same instrument was usedfor all measurements. Standing height was recorded in millimetres using the non-stretch method (shoes removed). Body weight (clothed, but jacket and shoesremoved) was obtained to the nearest 0.5 pound on a single beam balance(Detecto Scales Inc, Brooklyn, NY, USA). Weight in pounds was converted tokilograms.
The height and weight measurements were obtained by several observers, but foreach subject the majority of measurements were by the same observer. All observerswere blinded for prior height and weight measurements to assure an unbiasedreading.
Reproducibility of height measurements: Ten subjects were measured by ® vedi� erent observers within one half-hour. Each observer positioned the subject inthe stadiometer, recorded standing height, allowed the subject to walk away fromthe stadiometer, then repositioned him/her and remeasured height (5 times).
2.3. Urine collectionsAt each monthly visit, the subjects were given a urine collection container. They
were instructed to obtain a midstream catch. The time was recorded when the urinewas collected. The majority of the appointments (95%) were in the morning between9.30 and 11.00 am. Subjects were examined at the same time each month in order tominimize the circadian variation in NTx excretion (Bollen, Martin, Leroux et al.1995). All collections took place between February 1992 and October 1997.
2.4. MenarcheFemale subjects were asked at their monthly appointments whether menstrual
bleeding had occured or not. The date of the ® rst reported menstrual bleeding wasrecorded as menarche.
2.5. Growth rate/adolescent growth eventsMonthly standing height measurements (in cm) were used for the generation of
growth rate curves (cm/year). Growth rates based on monthly changes in standingheight were unreliable since the measurement error was greater than the changesmeasured. Therefore, subsequent overlapping 6 month changes in standing heightwere used to derive growth rate. From the di� erence in height at month 6 and month1, the growth rate in cm/y was calculated for this time period. The same was done for
Bone resorption during puberty 201
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.
the time period between month 2 and month 7, month 3 and month 8, and so on.The growth rates thus calculated were used to classify the subjects into one or moregrowth stages using the following criterea (see ® gure 1):
(a) Pre-pubertal growth: growth rate is relatively low (< 7 cm/y) and stable: thereis no generalized increase or decrease in growth rate in subsequently over-lapping 6 months time periods.
(b) Ascending phase of the growth spurt: there is a continuous increase in growthrate in overlapping 6 month time periods.
(c) Peak of growth spurt: Growth rates are greater than 7 cm/y for more than 2overlapping 6 month time periods.
(d) Descending: there is a continuous decrease in growth rate over time (until itreaches 2 cm/year).
(e) End: growth rate is below 2 cm/y (but greater than 0 cm/y) for all overlapping6 month time intervals during this period.
(f) Non-growing adolescents: a growth rate of 0 cm/year which was de® ned ashaving at least 5 consecutive height measurements varying less than 2 SD ofthe measurement error (0.6 cm), during a time period longer than 6 months.
Subjects could go through one or more of these stages during the study.
2.6. Urine analysesAliquots of the urine samples were frozen until analysed (up to one month). The
bone resorption marker, NTx, was measured in the urine samples by immunoassay(Hanson, Weis, Bollen et al. 1992). Results are expressed as picomoles of type Icollagen equivalents per millilitre (BCE ˆbone collagen equivalents of NTx immu-noreactivity). Samples were analyzed in quadruplicate , and the mean value was usedin the statistical analysis. The intra-assay and inter-assay coe� cients of variationwere 7% and 10% respectively.
202 A. M. Bollen
Figure 1. Adolescent growth stages. Growth of the subjects was classi® ed into one or more of six stages:a) Pre-pubertal growth stage; b) Ascending phase of the growth spurt;c) Peak of the pubertal growthspurt (growth rate > 7 cm/year) ; d) Descending phase of the pubertal growth spurt; e) End (growth rate< 2 cm/year), and f) Non-growing (growth rate 0 cm/year).
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.
The creatinine concentration was measured using a commercially available kit(Sigma DiagnosticsR , St. Louis, MO, U.S.A.) which is based on the Ja� e reaction.The mean of duplicate measurements was used. For each of the urine samples, thecrosslink concentration was normalized to creatinine (in nmol BCE/mmol creati-nine).
2.7. Statistical analysesTo compare NTx excretion between each of the growth stages, the urinary NTx
concentration normalized to creatinine for males and females was modeled usinggeneralized estimating equations with an identity link, a normal error structure , andan exchangeable correlation structure (Zeger and Liang 1986). The response variablewas NTx concentration in urine. The primary explanatory variable was the growthdevelopment stage. Multiple comparisons were performed using the least signi® cantdi� erence approach.
To determine the correlation between growth rate and NTx excretion the mean ofthe NTx excretion for all samples collected from an individual during each speci® cgrowth stage was calculated. The growth rate was calculated from the change instanding height during each growth stage (height at last observation Ð height at the® rst observation, divided by the elapsed time in days, times 365.25), in cm/year. ThePearson correlation coe� cient was calculated.
The e� ects of orthodontic treatment on urinary NTx excretion were evaluated asfollows. For each of the subjects who underwent orthodontic treatment the start dateof active treatment (appliance insertion) was determined from the chart. All treatedsubjects had at least three urine collections prior to the start of treatment. The meanNTx excretion in these three samples was compared to the mean NTx excretion inthe three ® rst monthly samples obtained after the initiation of orthodontics using anANOVA adjusted for gender and growth stage. This was also done excluding sub-jects who were in the ascending or descending growth stages when treatment wasstarted since the large changes in NTx excretion during these stages may have con-cealed the e� ects caused by the orthodontics.
3. Results3.1. Subjects
Of the 70 children approached, 2 refused (1 male and 1 female), and 3 were willingto participate, but were unable to provide urine samples (2 males, 1 female). Fivesubjects dropped out of the study due to relocation from the area (3 males and 2females) within 1 year of the start. Their results are not included. The remaining 60subjects (28 male and 32 female) were followed over a total period of 5.6 years (meantime of observation was 3.4 years). Mean age at the start and end of the study, andnumber of measurements are included in table 1. Note that one of the females wasyounger than the inclusion criterion at the start of the study (6.1 years versus 8years). The majority were caucasian (90%), there were 2 hispanics (both male), 2asians (both female), 2 African-Americans (1 male and 1 female). There were 2 setsof twins (one female, one male), and in addition ® ve pairs of siblings (three brother-sister pairs, and two sister-sister pairs). Three of the subjects did not undergo anyorthodontic treatment during the study period, the remainder were in some type ofactive orthodontic treatment during part of the study. Two female subjects sustaineda bone fracture during the course of the study (collarbone and forearm). The data onthese subjects were not eliminated from the study since it has been shown that
Bone resorption during puberty 203
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.
fractures in young subjects do not consistently a� ect the levels of urinary NTx/creatinine (Bennell, Malcolm, Brukner et al. 1998).
3.2. Growth measurements3.2.1. In general. The initial and ® nal height and weight, and the overal growthchanges are summarized in table 1. All individuals had greater height and weightat the end of the study, however the extend of change varied. The majority of theheight and weight measures of the adolescents followed in this study were withinthe 25th and 75th percentiles reported by HANES I (Vital and Health Statistics1977) both at the beginning and at the end of the study. The SD of the heightmeasurements was 0.3 cm. This is comparable to the SD reported by Voss andcoworkers (1990).
3.2.2. Growth rate/adolescent growth stages. The growth pattern of most sub-jects could be classi® ed into one or more of the six de® ned stages, with the follow-ing exceptions. In three of the females, no clear growth peak was present. Thegrowth rate remained around 6 cm/year for at least two years without any clear in-crease or decrease. This was followed by menarche for all three females afterwhich a decrease in growth rate was observed. The absence of a spurt in femaleshas been reported by other investigators (Zacharias and Rand 1983, 1986). Thepre-menarcheal period for these three females was classi® ed as the prepubertalgrowth stage, which was followed by the descending growth stage when thegrowth rates decreased. In three males and one female, a peak growth stage fol-
204 A. M. Bollen
Table 1. Subject characteristics.
Females Males Total
Number 32 28 60
Age at start of study 10.l4 (6.1± 12.9) 12.1 (9.1± 14.3) 11.2 (6.1± 14.3)mean (range)Age at end of study 13.9 (9.5± 17.1) 15.4 (13.4± 18.1) 14.6 (9.5± 18.1)mean (range)Time in study, years 3.5 (1.7± 4.6) 3.3 (1.7± 4.6) 3.4 (1.7± 4.6)mean (range)Total number of 1025 880 1905observations 32.0/subject 31.4/subject 31.75/subject
Height at start, cm 142.2 154.4 147.9mean (range) (121.3± 157.5) (135.3± 179.5) (121.3± 179.5)Height at end, cm 160.4 173.2 166.4mean (range) (145.4± 178.5) (155.7± 183.8) (145.4± 183.8)Height change, cm + 18.2 + 18.7 + 18.5mean (range) (8.4± 26.2) (3.0± 31.1) (3.0± 31.1)Weight at start, kg 35.0 48.1 41.1mean (range) (20.0± 56.7) (29± 78.9) (20.0± 78.9)Weight at end, kg 52.0 68.4 59.7mean (range) (30.4± 69.9) (43.1± 112.0) (30.4± 112.0)Weight change, kg + 17.0 + 20.4 + 18.5mean (range) (2.5± 31.8) (7.7± 44.9) (2.5± 44.9)
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.
lowed immediately after the pre-pubertal growth stage, without an intervening as-cending growth stage. The number of observed growth stages per subject is listedin table 2.
3.2.3. Menarche. 22 of the 32 females started having menses during the study.The mean age at menarche was 12.9 years (range 10.8± 14.7 years). This is similarto other reports (Zacharias and Rand 1986 ; median for white females living inNewton, MA ˆ12.96). In four cases (18%) menarche occured during the peakgrowth stage. The majority of the girls reached menarche when in the descendinggrowth stage (n ˆ15 or 68%), and in the remainder menarche occured betweenthese two growth stages (n ˆ3 or 14%).
3.3. Urinary NT x excretion3.3.1. During each of the growth stages. Table 3 lists urinary NTx excretion(means of the individual means) by pubertal growth stage and gender. There is anincrease from prepubertal stage to the peak of the growth spurt, followed by a de-crease. For each growth stage the values are similar for males and females.
The signi® cance of the e� ect of growth stage on NTx excretion was evaluatedusing the least signi® cant di� erence approach. Table 4 lists the p-values comparingthe NTx excretion between the di� erent stages, separate for males and females. For
Bone resorption during puberty 205
Table 2. Number of observed growth stages persubject.
Number of observedgrowth stages Females Males
One 6 (19%) 9 (32%)Two 19 (59%) 13 (46%)Three 6 (19%) 5 (18%)Four 1 (3%) 1 (4%)
Total 32 (100%) 28 (100%)
Table 3. NTx excretion according to growth stage and gender.
A. Females B. Males
Pubertal Growth Growthgrowth NTx rate** NTx rate**stage n excretion* cm/y n excretion* cm/y
Prepub. 6 589 (199) 5.3 (1.0) 10 673 (117) 5.0 (1.1)Ascend. 8 751 (231) 5.7 (1.0) 4 627 (198) 6.3 (2.3)Peak 20 849 (220) 8.2 (0.9) 16 856 (202) 9.1 (1.4)Descend. 20 522 (160) 4.1 (1.1) 15 507 (118) 4.1 (1.3)End 7 226 (53) 1.0 (0.2) 5 260 (62) 1.3 (0.4)Non-grow. 5 189 (61) 0.0 (0.8) 4 183 (48) 0.6 (0.7)
Data are means of individual subject means and (SD).* NTx excretion in nmol BCE/mmol creatinine. BCE: Bone Collagen Equivalent: the amount of bonecollagen from which the resorption market (NTx) was derived.** The growth rate was calculated from the change in standing height during each of the growth stages(height at last observation ¡ height at the ® rst observation, divided by the elapsed time in days, times365.25), in cm/year.
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.
both genders the prepubertal plateau and the ascending growth phase of the growthspurt are not signi® cantly di� erent. In the females the prepubertal plateau and thedescending phase of the growth spurt are not di� erent. In the males NTx excretionbetween the ascending and descending phase of the growth curve are not di� erent.NTx excretion is signi® cally di� erent between all other growth stages.
Individual month-to-month variability in NTx excretion during the variousgrowth stages was large, as indicated by the high variance (mean variance 48,639,minimum 1 456 and maximum 321 051). The mean coe� cient of variation(CV ˆstandard deviation divided by the mean and expressed as a percentage) was30%.
NTx excretion (normalized to creatinine) in adolescents who stopped growing (5females, mean adult height 159.2 cm, and 4 males, mean adult height 179.2 cm) wascompared to adults. The reported values for adult males and females range between26 nmol BCE/mmol creatinine (Garnero, Hausherr, Chapuy et al. 1996, for pre-menopausal females) and 41 nmol BCE/mmol creatinine (Bollen et al. 1995, malesand females). Thus NTx excretion in the non-growing adolescents was 4± 7 foldgreater than in adults.
3.3.2. In relation to menarche. Menarche was reached during the study by 22 ofthe 32 females. In ® gure 2, NTx excretion in relation to menarche is illustrated.The individual results for a period of about one year prior to menarche areplotted for 21 subjects (one subject had menarche 3 months after the start of thestudy and since she did not have a full year of results prior to menarche, only herpost-menarcheal values are included). NTx excretion following menarche is plottedfor 19 of the subjects (three subjects collected only one urine sample after men-arche and are not included in the ® gure).
The mean NTx excretion prior to orthodontic treatment was not statisticallydi� erent from the mean NTx excretion after the start of treatment (p ˆ0.83), evenwhen subjects were excluded who started treatment during the ascending or descend-ing growth stages (p ˆ0.86).
206 A. M. Bollen
Table 4. p-values of the e� ect of growth stage on NTx excretion.
A. FemalesAscending Peak Descending End Non-
growing
Prepubertal 0.28 0.001 0.15 < 0.0000 < 0.0000Ascending 0.014 0.0006 < 0.0000 < 0.0000Peak < 0.0000 < 0.0000 < 0.0000Descending < 0.0000 < 0.0000End 0.04
B. MalesAscending Peak Descending End Non-
growing
Prepubertal 0.22 0.0006 0.0002 < 0.0000 < 0.0000Ascending 0.0031 0.21 0.0001 < 0.0000Peak < 0.0000 < 0.0000 < 0.0000Descending < 0.0000 < 0.0000End 0.01
p-values resulting from multiple comparisons using the least sigi® cant di� erence approach, for females andmales separate.
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.
3.3.3. Correlation with growth rate. The mean of the NTx excretion for allsamples collected from an individual during each speci® c growth stage (total num-ber of means was 120) was calculated. The growth rate was calculated from thechange in standing height during each growth stage (height at last observation Ðheight at the ® rst observation, divided by the elapsed time in days, times 365.25),in cm/year. The Pearson correlation coe� cient between NTx excretion means andgrowth rate was 0.81. Linear regression analysis, adjusted for the fact that individ-uals contributed data to more than one growth stage, indicated that for each1 cm/year increase in growth rate, the NTx excretion increased with 74 nmol BCE/mmol creatinine (intercept 222 nmol BCE/mmol creatinine).
4. Discussion
The mean values of NTx excretion mirror the adolescent growth curve, withhigher levels during periods of higher growth rate, both for males and females.These di� erences are statistically signi® cant except between the prepubertal plateauand the ascending phase of the growth spurt (both for males and females), betweenthe descending and ascending phase for the males, and between the prepubertalplateau and the descending phase for the females.
The correlation between mean NTx excretion during each growth stage and thegrowth rate during that time span was 0.81. A longitudinal study by Cadogan andcoworkers (1998) on 12-year-old females reported a correlation of 0.41 between NTxexcretion and growth rate. This study collected urine samples every 6 months for aperiod of 18 months. The lower correlation between NTx excretion and growth ratemay be related to the fewer urine samples (every six months versus every month)from which the mean NTx excretion was calculated, and the narrower age range oftheir subjects.
Bone resorption during puberty 207
Figure 2. NTx excretion in relation to menarche. NTx excretion in monthly urine samples, about oneyear prior to menarche (n ˆ21), and during the period following menarche (n ˆ19). Females whocollected only one urine sample after menarche are not included in this ® gure.
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.
NTx excretion in adolescents who reached their adult height (growth rate0 cm/year) was 4± 7 fold greater than in adults. This may be related to the continuedbone modelling after closure of the endochondral growth plates. An increase inbone mineral after the end of the growth period (bone consolidation period) hasbeen described (Parsons, Prentice, Smith et al. 1996), and the peak total bodybone mass is not reached until the age of 20 (Haapasalo, Kannus, Sievanen et al.1996) or 26 (Teegarden, Proulx, Martin et al. 1995). The high level of boneresorption after cessation of linear growth may be related to this continued processof maturation and adaptation until obtainment of peak bone mass a decade or solater.
The large intra-individual month-to-month variability in NTx excretion may berelated to methodological issues or may re¯ ect naturally occuring variability in boneresorption. Measurement of metabolites in spot urines is in¯ uenced by changes increatinine excretion and circadian rythms. However, when NTx excretion was meas-ured in complete 24hour urine samples of caged growing piglets (Bollen, McCullochand Herring 1997), the day-to-day variability (CV ˆ31%) was similar to the month-to-month variability (CV ˆ30%) suggesting that the cause for the variability may berelated to issues other than the collection of spot urines.
The month-to-month variability in NTx excretion could be related to the growthprocess. Growth has been reported to occur in bursts or mini-growth spurts (Butler,McKie and Ratcli� e 1990, Hermanussen 1988, Lampl, Veldhuis and Johnson 1992,Lampl 1993). The saltatory nature of growth could in¯ uence the level of boneresorption. However, studies in non-growing individuals have reported similarhigh intra-individual variability in markers of bone metabolism. In young athletes(aged 17± 26) the month-to-month variability for bone metabolites was 22% (serumosteocalcin) to 29% (urinary pyridinoline) (Bennell et al. 1998). This indicates thatthe observed monthly variability within each subject may be a re¯ ection of thesubstantial variability inherent of biochemical markers of bone metabolism, andmay not be related to the growth process.
Based on the growth and development measurements, the subjects in this studycan be considered typical American adolescents. However, although orthodontictreatment may be common for American adolescents, this behaviour in the majorityof the subjects could have in¯ uenced the level of bone resorption. While orthodontictooth movement most certainly elicits resorption in the alveolar bone surroundingthe teeth, no increase in NTx excretion after the start of orthodontic treatment wasfound, despite the use of a sensitive assay. It is possible that the increase in boneresorption caused by the orthodontics may have been smaller than the monthlyvariability in NTx excretion. This is in agreement with the ® nding that in youngathletes fractures do not result in an increase in NTx or other bone metabolitemeasurement (Bennell et al. 1998).
The urine collections, which occured a month after the previous orthodonticappliance adjustment, may have been at a time when bone resorption was back tothe initial level. A study on growing pigs ® tted with a mandibular protrusive appli-ance, indicated that bone resorption was signi® cantly increased during a two-weektime period following appliance insertion, but then returned to the same level as inthe control animals (Bollen et al. 1997). Therefore, the inclusion of subjects under-going orthodontic treatment, most likely did not a� ect the levels of bone resorptionobserved in this study.
208 A. M. Bollen
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.
Acknowledgements
This study was supported by National Institutes of Health grant DE 10052 andOstex International.
ReferencesAPONE, S., LEE, M . Y., and EYRE, D . R ., 1997, Osteoclasts generate cross-linked collagen N-telopeptides
(NTx) but not free pyridinolines when cultured on human bone. Bone, 21 , 129± 136.BEARDSWORTH, L. J., EYRE, D . R ., and D ICKSON, I. R ., 1990, Changes with age in the urinary excretion
of lysyl- and hydroxylysylpyridinoline , two new markers of bone collagen turnover. Journal of Boneand Mineral Research, 5, 671± 676.
BENNELL, K . L., M ALCOLM S.A., BRUKNER, P. D ., G REEN, R . M ., H OP PER, J. L., WARK, J. D ., andEBELING, P. R ., 1998, A 12-month prospective study of the relationship between stress fracturesand bone turnover in athletes. Calci® ed T issue International, 63, 80± 85.
BLUMSOHN, A., H ANNON, R ., WRATE, R ., BARTON, J., AL-D EHAIMI, A., COLW ELL, A., and EASTELL, R .,1994, Biochemical markers of bone turnover in girls during puberty. Clinical Endocrinology(Oxford), 40 , 663± 670.
BOLLEN,A.-M ., and EYRE,D . R .,1994, Bone resorption rates in children monitored by the urinary assayof collagen type I cross-linked peptides. Bone , 15 , 31± 34.
BOLLEN, A.-M ., MCCULLOCH, K . J., and H ERRING, S. W., 1997, Whole body bone resorption in thegrowing pig. Growth, Development and Aging, 61, 181± 189.
BOLLEN, A.-M ., M ARTIN, M . D ., LEROUX, B. G ., and EYRE, D . R ., 1995, Circadian variation in urinaryexcretion of bone collagen cross-links. Journal of Bone and Mineral Research, 10 , 1885± 1890.
BUTLER, G . E., MCK IE, M ., and R ATCLIFFE, S. G ., 1990, The cyclical nature of prepubertal growth.Annals of Human Biology, 17, 177± 198.
CADOGAN, J.,BLUHMSOHN,A.,BARKER,M .E., and EASTELL,R .,1998, A longitudinal study of bone gainin pubertal girls: anthropometric and biochemical correlates. Journal of Bone and Mineral Research,13 , 1602± 1612.
CALVO, M . S., EYRE, D . R ., and G UNDBERG, C. M ., 1997, Molecular basis and clinical application ofbiological markers of bone turnover. Endocrine Review, 17 , 333± 368.
F UJIMOTO, S., K UBO, T., TANAKA, H ., M IURA, M ., and SEINO, Y., 1995, Urinary pyridinoline and deox-ypyridinoline in healthy children and in children with growth hormone de® ciency. Journal ofClinical Endocrinology and Metabolism , 80 , 1922± 1928.
G ARNERO, P., H AUSHERR, E., CHAP UY, M .-C., MARCELLI, C., G RANDJEAN, H ., M ULLER, C., CORMIER,C., BREART, G ., MEUNIER, P. J., and D ELMAS, P. D ., 1996, Markers of bone resorption predicthip fracture in elderly women: the EPIDOS prospective study. Journal of Bone and MineralResearch, 11, 1531± 1538.
H AAP ASALO, H ., K ANNUS, P., SIEVANEN, H ., PASANEN, M ., U USI-RASI, K ., H EINONEN, A., O JA, P., andVUORI, I., 1996 , Development of mass, density, and estimated mechanical characteristics of bonesin caucasian females. Journal of Bone and Mineral Research, 11, 1751± 1760.
H ANNON,R .A.,BLUMSOHN,A.,WRATE,R .M .,BARTON,J., and EASTELL,R .,1996 , Longitudinal changesin bone turnover in girls during puberty. Journal of Bone and Mineral Research, 11 (Suppl 1), S468.
H ANSON, D . A., WEIS, M . A. E., BOLLEN, A.-M ., M ASLAN, S. L., SINGER, F . R ., and EYRE, D . R ., 1992,A speci® c immunoassay for monitoring human bone resorption: quantitation of type I collagencross-linked N-telopeptides in urine. Journal of Bone and Mineral Research, 7, 1251± 1258.
H ERMANUSSEN,M .G EIGER-BENOIT,K ., BURMEISTER, J., and SIPPELL,W.G ., 1988, Periodical changes ofshort term growth velocity (`mini growth spurts’) in human growth. Annals of Human Biology, 15,103± 109.
JOHANSEN, J. S., G IW ERCMAN, A., H ARTWELL, D ., N IELSEN, C. T., PRICE, P. A., CHRISTIANSEN, C., andSKAKKEBAEK, N . E., 1988, Serum bone Gla-protein as a marker of bone growth in children andadolescents: correlation with age, height, serum insulin-like growth factor I, and serum testoster-one. Journal of Clinical Endocrinology and Metabolism , 67, 273± 278.
LAMP L, M ., 1993, Evidence of saltatory growth in infancy. American Journal of Human Biology, 5, 641±652.
LAMP L,M ., VELDHUIS, J.D ., and JOHNSON,M . L.,1992, Saltation and stasis: a model of human growth.Science, 258 , 801± 803.
LAP ILONNE, A., TRAVERS, R ., D IMAIO, M ., SALLE, B. L., and G LORIEUX, F . H ., 1996, Bone remodelingassessed by urinary excretion of cross-linked N-telopeptides of type I collagen (NTx) in infantsfrom birth to one year. Journal of Bone and Mineral Research, 11 (Suppl 1), S194.
L IEUW-A-F A,M .,SIERRA,R . I., and SPECKER,B.L.,1995, Carboxy-terminal propeptide of human type Icollagen and pyridinium cross-links as markers of bone growth in infants 1 to 18 months of age.Journal of Bone and Mineral Research, 10 , 849± 853.
PARSONS, T. J., PRENTICE, A., SMITH, E. A., COLE, T. J., and COMP STON, J. E., 1996 , Bone mineral massconsolidation in young British adults. Journal of Bone and Mineral Research, 11, 264± 274.
Bone resorption during puberty 209
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.
R IIS, B. J., K RABBE, S., CHRISTIANSEN , C., CATHERWOOD, B. D ., and D EFTOS, L. J., 1985, Bone turnoverin male puberty: a longitudinal study. Calci® ed T issue International, 37, 213± 217.
ROSEN, H . N ., D RESENER-POLLAK, R ., ROSENBLATT, M ., ZEIND, A. J., CLEMENS, J. D ., and G REENSPAN,S. L., 1994, Speci® city of urinary excretion of cross-linked N-telopeptides of type I collagen as amarker of bone turnover. Calci® ed T issue International, 54, 26± 29.
SAGGESE,G .,BERTELLONI,S.,BARONCELLI,G . I., and D I N ERO,G .,1992, Serum levels of carboxyterminalpropeptide of type I procollagen in healthy children from 1st year of life to adulthood and inmetabolic bone diseases. European Journal of Pediatrics, 151, 764± 768.
TEEGARDEN, D ., PROULX, W. R ., M ARTIN, B. R ., ZHAO, J., M CCABE, G . P., LYLE, R . M ., PEACOCK, M .,SLEMENDA, C., JOHNSTON, C. C., and WEAVER, C. M ., 1995, Peak bone mass in young women.Journal of Bone and Mineral Research, 10, 711± 715.
TOGO, M ., M ORINAGA, E., and TOGO, T., 1977, Urinary hydroxyproline, creatinine and urea excretionduring the growth of ® ve siblings. Annals of Human Biology, 4, 439± 447.
TOGO,M ., and TOGO,T.,1988, Initiation time of adolescent growth spurt estimated by a certain trough intime-series analysis of monthly antropometric and urinalysis data in ® ve siblings. Human Biology,60, 223± 235.
TRIVEDI, P., R ISTELI, J., R ISTELI, L., H INDMARSH, P. C., BROOK, C. G . D ., and M OW AT, A. P., 1991,Serum concentrations of the type I and III procollagen propeptides as biochemical markers ofgrowth velocity in healthy infants and children and in children with growth disorders. PediatricResearch, 30, 276± 280.
VITAL AND H EALTH STATISTICS, 1977, NCHS growth curves for children. Birth-18 years. United States.Series 11, number 165, pp 44± 48.
VOSS, L. D ., BAILEY, B. J. R ., CUMMING, K ., W ILKIN, T. J., and BETTS, P. R ., 1990, The reliability ofheight measurement (The Wessex Growth Study). Archives of Diseases in Childhood, 65, 1340± 1344.
WAKEFIELD, T., D ISNEY, G . W., M ASON, R . L., and BEAUCHENE, R . E., 1980, Relationships amonganthropometric indices of growth and creatinine and hydroxyproline excretion in preadolescentblack and white girls. Growth , 44 , 192± 204.
WEAVER, C. M ., PEACOCK, M ., M ARTIN, B. R ., M CCABE, G . P., ZHAO, J., SMITH, D . L., and WASTNEY,M . E., 1997, Quanti® cation of biochemical markers of bone turnover by kinetic measures ofbone formation and resorption in young healthy females. Journal of Bone and Mineral Research,12, 1714± 1720.
ZACHARIAS, L., and RAND, W. M ., 1983, Adolescent growth in height and its relation to menarche incontemporary Americal girls. Annals of Human Biology, 10, 209± 222.
ZACHARIAS, L., and RAND, W. M ., 1986 , Adolescent growth in weight and its relation to menrache incontemporary American girls. Annals of Human Biology, 13 , 369± 386.
ZEGER, S. L., and L IANG, K . Y., 1986, Longitudinal data analysis for discrete and continuous outcomes.Biometrics 42, 121± 130.
Address for correspondence: Dr Anne-Marie Bollen, Department of Orthodontics, Box 357446 ,University of Washington, Seattle, WA 98195, USA. Phone: (206) 685-8765 , fax: (206) 685-8163 , email:[email protected]
Zusammenfassung. Bei wachsenden Individuen sind die Knochenresorptions- undKnochendepositionsrate wesentlich groÈ û er als bei nicht-wachsenden Individuen. Ziel der vorliegendenStudie war es, in einer prospektiven Longitudinalstudie die Ausscheidung eines spezi® schen Markersder Knochenresorption im Urin in AbhaÈ ngigkeit von adoleszenten Wachstumsstadien waÈ hrend derPubertaÈ t zu bestimmen. Die Studie wurde an einer Kohorte von 60 Jugendlichen (28 Jungen und 32MaÈ dchen) uÈ ber einen Zeitraum von 3.4 Jahren (Range: 1.7± 4.6 Jahre) durchgefuÈ hrt. In monatlichenAbstaÈ nden wurden die KoÈ rperhoÈ he, das KoÈ rpergewicht und die Ausscheidung des Knochen-resorptionsmarkers, Kollagen Typ I N-Telopeptid (NTx), gemessen. Die VeraÈ nderungen in derKoÈ rperhoÈ he wurden herangezogen, um die Jugendlichen in eine oder mehrere von sechs adoleszentenWachstumsstadien einzuordnen: praÈ -puberales Wachstum (kontinuierliche moderate Wachstumsge-schwindigkeit), beginnender Wachstumsspurt (ansteigende Wachstumsgeschwindigkeit), maximalerWachstumsspurt (Wachstumsrate groÈ û er als 7 cm/Jahr fuÈ r mindestens 6 Monate), abnehmendeWachstumsgeschwindigkeit (kontinuierliche Abnahme der Wachstumsgeschwindigkeit) , Ende desWachstums (Wachstumsrate zwischen 0 und 2 cm/Jahr) und kein Wachstum. Vom praÈ -puberalenWachstum bis zum maximalen Wachstumsspurt wurde ein Anstieg der NTx-Exkretion um ca. 33%beobachtet (44% bei MaÈ dchen und 27% bei Jungen). Die abnehmende Wachstumsrate nach dem pub-eralen Wachstumsspurt koinzidierte mit einer deutlichen Abnahme der NTx-Exkretion. Die Unterschiedewaren statistisch signi® kant, ausgenommen der Unterschied zwischen dem praÈ -puberalen und dem begin-nenden Wachstumsspurt. Die mittlere NTx-Exkretion in einem Wachstumsstadium war auf individuellerEbene signi® kant mit der entsprechenden Wachstumsrate korreliert (r ˆ 0.81). Die inter- und intraindi-viduelle VariabilitaÈ t waren groû . Bei nicht wachsenden Adoleszenten (Wachstumsrate 0 cm/Jahr) war dieNTx-Exkretion 4 bis 7 mal groÈ û er als bei Erwachsenen. Bei allen MaÈ dchen war nach dem Auftreten derMenarche eine Abnahme der NTx-Exkretion zu beobachten. Zusammenfassend laÈ û t sich festhalten, dass
210 A. M. Bollen
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.
die Exkretion eines spezi® schen Knochenresorptionsmarkers, NTx, sowohl bei maÈ nnlichen als auch beiweiblichen Jugendlichen mit VeraÈ nderungen in der Wachstumsrate waÈ hrend der Adoleszenz korreliertwar. WaÈ hrend der unterschiedlichen Wachstumsstadien lieû en sich groû e inter- und intraindividuelleUnterschiede in der NTx-Exkretion beobachten. Bei Adoleszenten, die am Ende des puberalenWachstumsspurts ihre ErwachsenenhoÈ he erreichten, nahm die Knochenresorption drastisch ab, bliebjedoch 4 bis 7 mal hoÈ her als bei Erwachsenen.
Re sume . Les taux de de poà t et de re sorption de l’os sont nettement plus forts chez les sujets en cours decroissance que chez les individus qui ne grandissent pas. Le but de cette e tude est de mesurer l’excre tionurinaire d’un marqueur spe ci® que de la re sorption osseuse, en fonction des stades de croissance aÁ l’ado-lescence, dans le cadre d’une e tude longitudinale prospective. Une cohorte de 60 adolescents (28 garcË ons et32 ® lles) fut suivie pendant 3.4 anne es (limites de variation 1,7± 4,6 ans) au cours desquelles furente� ectue es des mesures mensuelles de stature, de poids et d’excre tion urinaire, d’un marqueur de la re sorp-tion osseuse le collageÁ ne de type I N-telopeptides (NTx). Les changements en taille-assis furent utilise spour classi® er les adolescents en un ou plus d’un stade de croissance aÁ l’adolescence: croissance pre -pubertaire (maintien d’un taux de croissance mode re ), acce le ration de la croissance en phase ascendante(augmentation du taux de croissance), croissance de pic (taux de croissance plus e leve que 7cm/an pendantau moins 6 mois), acce le ration de croissance en phase descendante (de croissance continue du taux decroissance), ® n de la croissance (taux de croissance entre 0 et 2 cm/an) et absence de croissance. On atrouve un accroissement en excre tion de NTx d’environ 33% entre le stade pre -pubertaire et le stade de lacroissance de pic (respectivement 44% et 27% pour les garcË ons et pour les ® lles). La diminution du taux decroissance conse cutive au pic pubertaire correspond aÁ une nette de croissance de l’excre tion de NTx. Cesdi� e rences sont statistiquement signi® catives, sauf entre les stades pre -pubertaire et de croissance ascen-dante. L’excre tion individuelle de NTx au cours de chaque stade est corre le e avec le taux de croissanceindividuelle du moment (r ˆ 0.81). On rencontre une forte variabilite intra et inter individuelle. Chez lesadolescents qui ne grandissent pas (taux de croissance de 0 cm/an), les niveaux d’excre tion de NTx sont de4 aÁ 7 fois plus e leve s que chez les adultes. Chez toutes les ® lles, les premieÁ res reÁ gles sont suivies par unediminution de l’excre tion de NTx. En conclusion, l’excre tion d’un marqueur spe ci® que de la re sortionosseuse, NTx, est corre le e chez les garcË ons comme chez les ® lles avec les changements du taux de crois-sance pendant l’adolescence. Il existe de fortes di� e rences intra et inter individuelles dans l’excre tion deNTx pendant les stades de croissance. Chez les adolescents qui atteignent leur taille adulte aÁ la ® n de lape riode pubertaire, la re sorption osseuse de croõà t spectaculairement mais demeure de 4 aÁ 7 fois plus e leve eeque chez les adultes.
Bone resorption during puberty 211
Ann
Hum
Bio
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f W
ater
loo
on 1
0/29
/14
For
pers
onal
use
onl
y.