<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:mml="http://www.w3.org/1998/Math/MathML" article-type="research-article"><?properties manuscript?><front><journal-meta><journal-id journal-id-type="nlm-journal-id">0404024</journal-id><journal-id journal-id-type="pubmed-jr-id">584</journal-id><journal-id journal-id-type="nlm-ta">Ann Hum Biol</journal-id><journal-id journal-id-type="iso-abbrev">Ann. Hum. Biol.</journal-id><journal-title-group><journal-title>Annals of human biology</journal-title></journal-title-group><issn pub-type="ppub">0301-4460</issn><issn pub-type="epub">1464-5033</issn></journal-meta><article-meta><article-id pub-id-type="pmid">29082754</article-id><article-id pub-id-type="pmc">5765752</article-id><article-id pub-id-type="doi">10.1080/03014460.2017.1388845</article-id><article-id pub-id-type="manuscript">HHSPA917193</article-id><article-categories><subj-group subj-group-type="heading"><subject>Article</subject></subj-group></article-categories><title-group><article-title>Longitudinal changes in BMI <italic>z</italic>-scores among 45 414 2&#x02013;4-year olds with severe obesity</article-title></title-group><contrib-group><contrib contrib-type="author"><name><surname>Freedman</surname><given-names>David S.</given-names></name><xref ref-type="aff" rid="A1">a</xref></contrib><contrib contrib-type="author"><name><surname>Butte</surname><given-names>Nancy F.</given-names></name><xref ref-type="aff" rid="A2">b</xref></contrib><contrib contrib-type="author"><name><surname>Taveras</surname><given-names>Elsie M.</given-names></name><xref ref-type="aff" rid="A3">c</xref></contrib><contrib contrib-type="author"><name><surname>Goodman</surname><given-names>Alyson B.</given-names></name><xref ref-type="aff" rid="A1">a</xref></contrib><contrib contrib-type="author"><name><surname>Blanck</surname><given-names>Heidi M.</given-names></name><xref ref-type="aff" rid="A1">a</xref></contrib></contrib-group><aff id="A1">
<label>a</label>Division of Nutrition, Physical Activity and Obesity, Centers for Disease Control and Prevention, Atlanta, GA, USA</aff><aff id="A2">
<label>b</label>Baylor College of Medicine, Children&#x02019;s Nutrition Research Center, Houston, TX, USA</aff><aff id="A3">
<label>c</label>Department of Pediatrics, MassGeneral Hospital for Children, Boston, MA, USA</aff><author-notes><corresp id="FN1">CONTACT: David S. Freedman <email>dxf1@cdc.gov</email> CDC F-77, 4770 Buford Highway, Atlanta, GA 30341&#x02013;3724, USA</corresp></author-notes><pub-date pub-type="nihms-submitted"><day>9</day><month>1</month><year>2018</year></pub-date><pub-date pub-type="epub"><day>30</day><month>10</month><year>2017</year></pub-date><pub-date pub-type="ppub"><month>12</month><year>2017</year></pub-date><pub-date pub-type="pmc-release"><day>01</day><month>12</month><year>2018</year></pub-date><volume>44</volume><issue>8</issue><fpage>687</fpage><lpage>692</lpage><!--elocation-id from pubmed: 10.1080/03014460.2017.1388845--><abstract><sec id="S1"><title>Background</title><p id="P1">BMI <italic>z</italic>-scores (BMI<italic>z</italic>) based on the Centers for Disease Control and Prevention (CDC) growth charts among children do not accurately characterise BMI levels among children with very high BMIs. These limitations may be particularly relevant in longitudinal and intervention studies, as the large changes in the L (normality) and S (dispersion) parameters with age can influence BMI<italic>z</italic>.</p></sec><sec id="S2"><title>Aim</title><p id="P2">To compare longitudinal changes in BMI<italic>z</italic> with BMI expressed as a percentage of the 95th percentile (%BMIp95) and a modified <italic>z</italic>-score calculated as log(BMI/M)/S.</p></sec><sec id="S3"><title>Subjects and methods</title><p id="P3">A total of 45 414 2&#x02013;4-year-olds with severe obesity (%BMIp95 &#x02265; 120).</p></sec><sec id="S4"><title>Results</title><p id="P4">Changes in very high BMI<italic>z</italic> levels differed from the other metrics. Among severely obese 2-year-old girls, for example, the mean BMI<italic>z</italic> decreased by 0.6 SD between examinations, but there were only small changes in BMIp95 and modified BMI<italic>z</italic>. Some 2-year-old girls had BMI<italic>z</italic> decreases of &#x0003e;1 SD, even though they had large increases in BMI, %BMIp95 and modified BMI<italic>z</italic>.</p></sec><sec id="S5"><title>Conclusions</title><p id="P5">Among children with severe obesity, BMI<italic>z</italic> changes may be due to differences in the transformations used to estimate levels of BMI<italic>z</italic> rather than to changes in body size. The BMIs of these children could be expressed relative to the 95th percentile or as modified <italic>z</italic>-scores.</p></sec></abstract><kwd-group><kwd>BMI</kwd><kwd>longitudinal</kwd><kwd>children</kwd><kwd>LMS</kwd></kwd-group></article-meta></front><body><sec sec-type="intro" id="S6"><title>Introduction</title><p id="P6">The 2000 Centers for Disease Control and Prevention (CDC) growth charts (<xref rid="R25" ref-type="bibr">Kuczmarski et al., 2002</xref>) are widely used to classify obesity (BMI &#x02265;95th percentile for a child&#x02019;s sex and age) among children and adolescents in the US. These charts contain 10 smoothed percentiles (between the 3rd and 97th) of BMI (<xref rid="R15" ref-type="bibr">Flegal &#x00026; Cole, 2013</xref>; <xref rid="R25" ref-type="bibr">Kuczmarski et al., 2002</xref>). These percentiles were used to derive three parameters (L, power transformation for normality; M, median; and S, dispersion) (<xref rid="R13" ref-type="bibr">Cole &#x00026; Green, 1992</xref>) that allow the BMI of a child to be expressed as a <italic>z</italic>-score or percentile.</p><p id="P7">However, the use of these LMS parameters for BMIs above the CDC 97th percentile (<italic>z</italic> ~ 1.88) constrains the maximum <italic>z</italic>-score that can be attained at a given sex and age (<xref rid="R6" ref-type="bibr">Centers for Disease Control and Prevention (CDC), 2016</xref>; <xref rid="R41" ref-type="bibr">Woo, 2009</xref>). A wide range of very high BMIs can, therefore, map to similar <italic>z</italic>-scores, depending upon the sex and age of the child (<xref rid="R17" ref-type="bibr">Freedman et al., 2017b</xref>). To overcome these limitations, it has been suggested that very high BMIs be expressed as a percentage of the CDC 95th BMI percentile (%BMIp95) (<xref rid="R14" ref-type="bibr">Flegal et al., 2009</xref>). Severe obesity is classified as a %BMIp95 &#x02265; 120 (<xref rid="R23" ref-type="bibr">Kelly et al., 2013</xref>).</p><p id="P8">Although investigators have emphasised the drawbacks of analysing changes in BMI<italic>z</italic> (<xref rid="R2" ref-type="bibr">Berkey &#x00026; Colditz, 2007</xref>; <xref rid="R11" ref-type="bibr">Cole et al., 2005</xref>; <xref rid="R22" ref-type="bibr">Kakinami et al., 2014</xref>; <xref rid="R32" ref-type="bibr">Paluch et al., 2007</xref>), these changes continue to be widely used in longitudinal studies (<xref rid="R26" ref-type="bibr">Lo et al., 2014</xref>; <xref rid="R28" ref-type="bibr">McCormick et al., 2014</xref>; <xref rid="R30" ref-type="bibr">O&#x02019;Connor et al., 2017</xref>; <xref rid="R35" ref-type="bibr">Smego et al., 2017</xref>) and it has been suggested that a BMI <italic>z</italic>-score reduction of more than 0.20&#x02013;0.25 SDs may be clinically important (<xref rid="R36" ref-type="bibr">US Preventive Services Task Force, 2017</xref>). The limitations of the CDC BMI<italic>z</italic> values may be particularly relevant in longitudinal studies of children with very high BMIs, as the age-related changes in the L and S parameters could strongly influence observed changes in BMI<italic>z</italic>. BMI<italic>z</italic> changes among these children may not accurately reflect changes in body size and analyses focused on BMI<italic>z</italic> change could lead to incorrect conclusions.</p><p id="P9">The objective of the current study is to examine longitudinal changes in BMI<italic>z</italic>, %BMIp95 and a modified <italic>z</italic>-score that eliminates the compression of very high BMIs among 45 414 2&#x02013;4-year-olds with severe obesity.</p></sec><sec sec-type="methods|subjects" id="S7"><title>Subjects and methods</title><p id="P10">As previously described (<xref rid="R18" ref-type="bibr">Freedman et al., 2016</xref>, <xref rid="R17" ref-type="bibr">2017b</xref>), the CDC&#x02019;s Paediatric and Nutrition Surveillance Study (PedNSS) was a state-based public health surveillance system that monitored the nutritional status from birth to age 5 years of primarily low-income children. All procedures followed were in accordance with the ethical standards of CDC and approval was obtained from the relevant committee on human subjects.</p><sec id="S8"><title>Data collection and statistical analyses</title><p id="P11">Weight was recorded to the nearest 1/4 pound and height to the nearest 1/8 inch with a measuring board (<xref rid="R29" ref-type="bibr">Mei et al., 1998</xref>). There were 14.4 million records from children who were between 24.0&#x02013;59.9 months of age from 2008&#x02013;2011. Based on the state, assigned ID and birthdate, we identified children who were examined multiple times.</p><p id="P12">The initial data cleaning excluded (1) 623 000 (4%) records that were missing information on weight or height (or had a value of 0), (2) 70 000 children for whom the date of birth or sex differed across examinations for the same ID and (3) 151 000 records that had a weight, height or BMI considered to be implausible based on levels of modified <italic>z</italic>-scores for weight, height, and BMI (<xref rid="R6" ref-type="bibr">Centers for Disease Control and Prevention (CDC), 2016</xref>).</p><p id="P13">These exclusions resulted in a dataset containing 13.5 million records from 8.7 million children. Of these children, 3.6 million were examined more than once and we focus on the first two examinations for the 3 570 241 children whose examinations occurred at least 3 months apart. We further excluded 51 114 children (1.4%) whose weight and height did not change between examinations, possibly indicating a data error.</p><p id="P14">There were also several children who appeared to have unreasonable changes in weight, height or BMI. For example, 0.1% of the children had annualised decreases of more than 3.8 kg for weight, more than 10 cm for height, and more than 10 kg/m<sup>2</sup> for BMI. We therefore excluded children who were in the upper or lower 0.1% of the distribution of these characteristics. These exclusions resulted in a sample of 3 498 745 children,</p><p id="P15">Of these children, 45 814 had a BMI &#x02265;120% of the 95th percentile at their first examination and form the sample for the analyses. The mean interval between examinations was 11 (SD = 5) months.</p></sec><sec id="S9"><title>BMI metrics</title><p id="P16">Body mass index (BMI) was calculated as kg/m<sup>2</sup>. BMI<italic>z</italic> was calculated by expressing a child&#x02019;s BMI relative to children of the same sex and age in the CDC growth charts (<xref rid="R25" ref-type="bibr">Kuczmarski et al., 2002</xref>) using the L (transformation for normality) M (median) S (dispersion) method (<xref rid="R10" ref-type="bibr">Cole, 1990</xref>; <xref rid="R12" ref-type="bibr">Cole et al., 1998</xref>): 
<disp-formula id="FD1"><mml:math id="M1" display="block" overflow="scroll"><mml:mrow><mml:mi>BMI</mml:mi><mml:mi>z</mml:mi><mml:mo>=</mml:mo><mml:mrow><mml:mo>[</mml:mo><mml:mrow><mml:msup><mml:mrow><mml:mo>(</mml:mo><mml:mfrac><mml:mi>BMI</mml:mi><mml:mi mathvariant="normal">M</mml:mi></mml:mfrac><mml:mo>)</mml:mo></mml:mrow><mml:mi mathvariant="normal">L</mml:mi></mml:msup><mml:mo>&#x02212;</mml:mo><mml:mn>1</mml:mn></mml:mrow><mml:mo>]</mml:mo></mml:mrow><mml:mo>&#x000f7;</mml:mo><mml:mo stretchy="false">(</mml:mo><mml:mi mathvariant="normal">L</mml:mi><mml:mo>&#x000d7;</mml:mo><mml:mi mathvariant="normal">S</mml:mi><mml:mo stretchy="false">)</mml:mo></mml:mrow></mml:math></disp-formula></p><p id="P17">Because L is less than &#x02212;1 at all ages (the minimum is &#x02212;3.4) (CDC; National Center for Health Statistics, n.d.), when BMI is large relative to M (median), (BMI &#x000f7; M)<sup>L</sup> approaches 0 and the maximum possible BMI<italic>z</italic> is &#x02013;1 &#x000f7; (L &#x000d7; S). For example, the value of (BMI &#x000f7; M)<sup>L</sup> for a 4-year-old girl with a BMI of 30 would be ~0.1, resulting in a BMI<italic>z</italic> of 3.4 SDs; the maximum possible BMI<italic>z</italic> for that sex/age, irrespective of the actual BMI, is 3.8 SDs. Between 24&#x02013;59 months of age, the maximum possible BMI<italic>z</italic> in the CDC growth charts decreases from 11.9 to 3.5 SDs among girls, while the pattern among boys is concave, with the peak (~10 SDs) at about 39 months of age (<xref rid="R17" ref-type="bibr">Freedman et al., 2017b</xref>).</p><p id="P18">Obesity is defined as a BMI greater than or equal to the sex- and age-specific 95th percentile (BMI<italic>z</italic> &#x02265; 1.645) of the CDC growth charts (<xref rid="R25" ref-type="bibr">Kuczmarski et al., 2002</xref>; <xref rid="R31" ref-type="bibr">Ogden &#x00026; Flegal 2010</xref>) and severe obesity as a BMI &#x02265; 120% of the 95th percentile (<xref rid="R23" ref-type="bibr">Kelly et al., 2013</xref>). We refer to a BMI that is expressed as a percentage of the 95th percentile as %BMIp95. For example, the CDC 95th percentile for a 3-year-old boy is 19.3 kg/m<sup>2</sup> and a boy with a BMI of 21.3 kg/m<sup>2</sup> would have a %BMIp95 of 110% (100 &#x000d7; 21.3/19.3).</p><p id="P19">Although the L parameter in the CDC growth charts varies between &#x02212;1.0 and &#x02212;3.4, setting L equal to 0 removes the compression of very high BMIs into a narrow, bounded range of <italic>z</italic>-scores. Based on the Box-Cox transformation (<xref rid="R3" ref-type="bibr">Box &#x00026; Cox, 1964</xref>), the equation for BMI<italic>z</italic> when L = 0 is: 
<disp-formula id="FD2"><mml:math id="M2" display="block" overflow="scroll"><mml:mrow><mml:mi>log</mml:mi><mml:mrow><mml:mo>(</mml:mo><mml:mfrac><mml:mi>BMI</mml:mi><mml:mi mathvariant="normal">M</mml:mi></mml:mfrac><mml:mo>)</mml:mo></mml:mrow><mml:mo>&#x000f7;</mml:mo><mml:mi mathvariant="normal">S</mml:mi></mml:mrow></mml:math></disp-formula></p><p id="P20">This modified <italic>z</italic>-score expresses a child&#x02019;s BMI relative to the median BMI and adjusts the log of this ratio for the dispersion (S) of the BMI distribution. It should be noted that this metric is unrelated to the method (Centers for Disease Control and Prevention (CDC), n.d.) that has been proposed to identify outliers in the CDC growth charts.</p></sec><sec id="S10"><title>Statistical analyses</title><p id="P21">Analyses were performed in R (<xref rid="R33" ref-type="bibr">R Core Team, 2017</xref>). We examine the cross-classification of BMI<italic>z</italic> and %BMIp95 between the two examinations and the changes by sex and both age and BMI<italic>z</italic>. These changes are displayed using violin plots (<xref rid="R21" ref-type="bibr">Hintze &#x00026; Nelson, 1998</xref>), a technique that shows the entire distribution of the data rather than only selected percentiles, as do boxplots.</p><p id="P22">The intra-class correlation coefficient (ICC) (<xref rid="R34" ref-type="bibr">Shrout &#x00026; Fleiss, 1979</xref>; <xref rid="R40" ref-type="bibr">Wolak, 2017</xref>) was used to assess the strength of the association across the two examinations. In contrast to the typical Pearson correlation coefficient, the ICC assesses the within-person similarity of the characteristic across examinations and its magnitude would be reduced if the mean levels differ between examinations. Standard errors of differences between the ICCs were obtained by bootstrapping (<xref rid="R4" ref-type="bibr">Canty &#x00026; Ripley, 2017</xref>)</p><p id="P23">We also used time series plots to display changes in levels of the BMI metrics among the 100 girls who had an initial BMI<italic>z</italic> &#x02265; 4.75; these line plots connect the initial and final values of BMI<italic>z</italic> (<italic>y</italic>-axis) for each child.</p></sec></sec><sec sec-type="results" id="S11"><title>Results</title><p id="P24"><xref rid="T1" ref-type="table">Table 1</xref> shows levels of various characteristics at the first examination for the 45 414 children who had severe obesity at their first examination. Among these children, mean levels of BMI<italic>z</italic> were 3.9 SD (boys) and 3.4 SD (girls), and were 128&#x02013;129 for %BMIp95. These children with severe obesity also had very high levels of weight-for-age and moderately elevated levels of height-for-age. The bottom of <xref rid="T1" ref-type="table">Table 1</xref> shows the mean changes between examinations which occurred, on average, 11 months apart. Mean levels of all BMI metrics decreased with age, but, relative to the initial level, the mean changes in %BMIp95 and modified BMI<italic>z</italic> were very small. In contrast, the mean BMI<italic>z</italic> level among girls decreased by 0.5 SDs (~15% of the initial level).</p><p id="P25"><xref rid="T2" ref-type="table">Table 2</xref> shows the cross-classification of BMI<italic>z</italic> (left) and %BMIp95 (right) levels at the two examinations. If BMI<italic>z</italic> levels were stable over time, most children would fall along the main diagonal, and roughly similar numbers would have lower or higher levels at re-examination. In contrast, most children had a substantially lower BMI<italic>z</italic> value upon re-examination and this was particularly evident among girls. Among the 8745 girls who had an initial BMI<italic>z</italic> of 3.5&#x02013;4.4 SDs, for example, 73% had a BMI<italic>z</italic> of less than 3.5, while only 0.1% (<italic>n</italic> = 9) had a higher BMI<italic>z</italic>. As assessed by the intra-class correlation coefficient, the association between BMI<italic>z</italic> levels at the two examinations among girls was very weak (ICC = 0.14, 95% CI = 0.13&#x02013;0.15).</p><p id="P26">Substantially different results were seen for %BMIp95 (<xref rid="T2" ref-type="table">Table 2</xref>, right). For this analysis, cut points of %BMIp95 were selected so the sample size of each group at the initial examination would equal those for the BMI<italic>z</italic> categories. About 50% of girls with an initial %BMIp95 &#x02265; 155 (vs 3% of girls in the highest BMI<italic>z</italic> category) remained in this category at re-examination (bottom row). The results for modified BMI<italic>z</italic> (not shown) were very similar to those for %BMIp95, with 50% of the girls in the highest initial category (&#x02265;7.5 SDs) remaining in this category at re-examination. As assessed by bootstrapping, the ICCs for %BMIp95 (0.46) and modified BMI<italic>z</italic> (0.48) at the two examinations were significantly (<italic>p</italic> &#x0003c;0.0001) higher than the ICC for BMI<italic>z</italic> (0.14) among girls.</p><p id="P27"><xref rid="F1" ref-type="fig">Figure 1</xref> shows violin plots, categorised by sex and age, for the changes in BMI<italic>z</italic> (top) and %BMIp95 (bottom) between examinations. Among 2-year-olds, the median BMI<italic>z</italic> among girls decreased by &#x02212;0.6 to &#x02212;1.0 SDs in the three categories of BMI<italic>z</italic>, while comparable changes among boys were much smaller (range = &#x02212;0.2 to +0.3 SDs). The sex difference in BMI<italic>z</italic> changes was particularly striking for 331 2-year-olds who had an initial BMI<italic>z</italic> &#x02265; 4.5 SDs; 99% (328) of these girls showed a decrease (mean = &#x02212;1.0 SD) in BMI<italic>z</italic> between examinations. In contrast to BMI<italic>z</italic>, mean changes in %BMIp95 (bottom) showed little difference between boys and girls. The mean change in modified BMI<italic>z</italic> among 2-year-old girls with a BMI<italic>z</italic> &#x02265; 4.5 was +0.01 SDs (not shown).</p><p id="P28"><xref rid="F2" ref-type="fig">Figure 2</xref> shows line plots for the 100 girls who had the highest initial value of BMI<italic>z</italic> (&#x02265;4.75), with each line connecting the initial and final value of the BMI metrics; age is on the <italic>x</italic>-axis. As seen in the left panel, the BMI<italic>z</italic> level of all girls decreased (mean = &#x02212;1.1 SDs) between examinations. Although most of these girls also showed a decrease in %BMIp95 between examinations, 35 showed an increase in %BMIp95 and 55 showed an increase in modified BMI<italic>z</italic>. Among these 100 girls, the median BMI decreased by 1.7 kg/m<sup>2</sup>, but the range of BMI changes varied from &#x02212;6.3 kg/m<sup>2</sup> to +2.6 kg/m<sup>2</sup>.</p><p id="P29">Nine of these girls had a BMI<italic>z</italic> decrease by more than 1 SD, but a %BMIp95 increase of 10 or more percentage points. One girl, for example, had a BMI<italic>z</italic> decrease from 4.9 to 3.4 SDs between 26 and 55 months of age, but increases in BMI (from 28.0 to 30.6 kg/m<sup>2</sup>), %BMIp95 (from 148 to 169) and modified BMI<italic>z</italic> (from 6.5 to 8.6 SDs). Even if this girl&#x02019;s BMI had increased to 100 kg/m<sup>2</sup>, her BMI<italic>z</italic> would have decreased, because the maximum attainable BMI<italic>z</italic> in the CDC growth charts for a 55-month-old girl is 3.7 SDs.</p></sec><sec sec-type="discussion" id="S12"><title>Discussion</title><p id="P30">Although the report for the CDC growth charts (<xref rid="R25" ref-type="bibr">Kuczmarski et al., 2002</xref>) cautioned about extrapolating outside the 3rd and 97th percentiles of BMI, BMI<italic>z</italic> remains widely used in studies that include large numbers of children with very high BMIs (<xref rid="R20" ref-type="bibr">Hampl et al., 2016</xref>; <xref rid="R24" ref-type="bibr">Kreier et al., 2013</xref>; <xref rid="R28" ref-type="bibr">McCormick et al., 2014</xref>; <xref rid="R35" ref-type="bibr">Smego et al., 2017</xref>; <xref rid="R37" ref-type="bibr">Wang et al., 2015</xref>). We have shown that, among children with severe obesity, BMI<italic>z</italic> levels are only weakly associated with other measures of body size (<xref rid="R16" ref-type="bibr">Freedman et al., 2017a</xref>) and the current results highlight the drawbacks of attempting to assess longitudinal BMI<italic>z</italic> changes. Because <italic>z</italic>-scores in the CDC growth charts for very high BMIs are strongly influenced by the sex- and age-specific values of the L and S parameters, an analysis of BMI<italic>z</italic> changes among children with severe obesity can lead to paradoxical results. For example, some children in the current study had substantial (&#x02265;1 SD) decreases in BMI<italic>z</italic>, but large increases in BMI, %BMIp95 and modified BMI<italic>z</italic>. This is the result of differences in the L and S parameters at the initial and final examinations. We have found that the longitudinal tracking of BMI levels among 247 school-aged children with severe obesity is stronger for %BMIp95 than for BMI<italic>z</italic> (<italic>r</italic> = 0.61 vs 0.46) (<xref rid="R19" ref-type="bibr">Freedman &#x00026; Berenson, 2017</xref>).</p><p id="P31">We found, for example, particularly large decreases in the BMI<italic>z</italic> levels of girls who had very high BMIs at their initial examinations. This was the result of the maximum attainable BMI<italic>z</italic> among girls decreasing from 11.9 SDs (24 months) to 3.6 SDs (59 months), compressing very high BMIs into a lower range of <italic>z</italic>-scores at older ages. Although these decreases were most evident among girls with the very highest BMI<italic>z</italic> values, this compression affects a much larger group of children. For example, of the girls with severe obesity in the current study, only 13% showed an increase in BMI<italic>z</italic> between examinations, whereas 49% showed an increase in %BMIp95 and 38% showed an increase in modified BMI<italic>z</italic>.</p><p id="P32">The compression of very high BMI values into a narrow range of <italic>z</italic>-scores that differ by sex and age could be particularly problematic in intervention studies. Although the current study focused on 2&#x02013;4-year-olds, a period in which the L parameter change markedly with age, the maximum attainable BMI<italic>z</italic> varies by sex and across all ages (<xref rid="R16" ref-type="bibr">Freedman et al., 2017a</xref>). Because of these differences in L and S by age, an ineffective intervention could result in a substantial decrease in BMI<italic>z</italic> or an effective treatment could result in no change depending on the sex and age. To alleviate the effects of the compression of BMI<italic>z</italic>, we used %BMIp95 and modified BMI<italic>z</italic> in the current study. The latter metric, which was based on setting the L parameter equal to 0, also accounts for differences in the dispersion of BMI values between boys and girls and across ages. We found similar results for both %BMIp95 and modified BMI<italic>z</italic>. It should be noted that other powers of L, such as 0.5, could also be used in the construction of modified <italic>z</italic>-scores.</p><p id="P33">It has been suggested that a BMI <italic>z</italic>-score reduction of more than 0.20&#x02013;0.25 may be clinically important (<xref rid="R36" ref-type="bibr">US Preventive Services Task Force, 2017</xref>) and at least one intervention study has used a <italic>z</italic>-score decrease of &#x02265;0.2 SD to classify a successful intervention (<xref rid="R27" ref-type="bibr">Markert et al., 2014</xref>). However, this criterion, particularly among very obese children, is problematic, because the interpretation of a 0.2 SD depends upon the initial BMI<italic>z</italic> value. Among 2-year-old girls in the current study, for example, 87% of those with an initial %BMIp95 &#x02265; 130 showed a BMI<italic>z</italic> decrease of more than 0.2 SDs in the absence of any intervention. Among older children, in contrast, even large changes in BMI can be associated with virtually no change in BMI<italic>z</italic>. Based on the LMS parameters in the CDC growth charts, for example, the BMI<italic>z</italic> of a 94-month-old boy whose BMI decreased from 38 to 32 kg/m<sup>2</sup> over 1 year would decrease by only 0.16 SDs.</p><p id="P34">Many of the limitations of BMI<italic>z</italic> have been noted by others (<xref rid="R2" ref-type="bibr">Berkey &#x00026; Colditz, 2007</xref>; <xref rid="R11" ref-type="bibr">Cole et al., 2005</xref>; <xref rid="R14" ref-type="bibr">Flegal et al., 2009</xref>; <xref rid="R32" ref-type="bibr">Paluch et al., 2007</xref>; <xref rid="R41" ref-type="bibr">Woo, 2009</xref>). Although the compression of very high values of BMI<italic>z</italic> occur with all growth charts, the effects are particularly large in the CDC growth charts, because BMIs in the CDC reference population (1960&#x02013;1994) were more skewed and required larger, negative power transformations (L) (<xref rid="R41" ref-type="bibr">Woo, 2009</xref>) than in other growth charts. For example, the values of the L parameter in the WHO charts for both pre-school children (<xref rid="R39" ref-type="bibr">WHO Multicentre Growth Reference Study Group, 2006</xref>) and 5&#x02013;19-year-olds (World Health Organization, n.d.) are also substantially closer to 0 than in the CDC growth charts. Because the WHO charts also estimate <italic>z</italic>-scores for values that are more than 3 SDs from the median by extrapolating the distance between 2&#x02013;3 SDs, these growth standards do not show compression of very high BMIs.</p><p id="P35">Several limitations of our analyses should be considered. The PedNSS dataset is large, but the data were not collected for research purposes and there were many errors in the data. In addition to excluding a large number of records that were missing information on weight or height, we excluded children who had information for date of birth or sex that differed across examinations or who had a weight, height or BMI considered to be biologically implausible (<xref rid="R6" ref-type="bibr">Centers for Disease Control and Prevention (CDC), 2016</xref>; <xref rid="R18" ref-type="bibr">Freedman et al., 2016</xref>). We also excluded children from the analyses if their change in weight, height or BMI was in the upper or lower 0.1% of the distribution. Although it is likely that we did not exclude all errors and that some valid values were mistakenly excluded, any errors would be expected to have influenced changes in all BMI metrics. Finally, it should be realised that the longitudinal changes we observed in levels of BMI<italic>z</italic>, %BMIp95 and modified BMI<italic>z</italic> were based on cross-sectional growth charts and there are several problems with this approach (<xref rid="R9" ref-type="bibr">Cole, 1994</xref>; <xref rid="R43" ref-type="bibr">Zemel, 2009</xref>). For example, regression to the mean would result in subsequent values being closer to the median than is the initial level. Longitudinal velocity growth charts have been constructed to assess changes over time, but these have focused on weight and length among infants (<xref rid="R8" ref-type="bibr">Cole, 1995</xref>; <xref rid="R38" ref-type="bibr">WHO Department of Nutrition for Health and Development, 2009</xref>).</p><p id="P36">Extrapolated <italic>z</italic>-scores based on the CDC growth charts are widely used among children who have very high BMIs (<xref rid="R1" ref-type="bibr">Baughcum et al., 2015</xref>; <xref rid="R20" ref-type="bibr">Hampl et al., 2016</xref>; <xref rid="R24" ref-type="bibr">Kreier et al., 2013</xref>; <xref rid="R35" ref-type="bibr">Smego et al., 2017</xref>; <xref rid="R37" ref-type="bibr">Wang et al., 2015</xref>). However, BMI<italic>z</italic> values for extremely high BMIs can differ substantially from the empirical estimates (<xref rid="R14" ref-type="bibr">Flegal et al., 2009</xref>), have an effective upper limit (<xref rid="R41" ref-type="bibr">Woo, 2009</xref>) and are strongly influenced by sex and age. Our results indicate that the analyses of BMI<italic>z</italic> may be particularly problematic in longitudinal analyses of children with severe obesity. Among these children, changes in BMI<italic>z</italic> can simply reflect changes in various parameters of the CDC growth charts rather than changes in body size. In longitudinal studies of children who have very high BMIs, investigators should consider using another BMI metric, such as %BMIp95 or modified BMI<italic>z</italic> (log(BMI/M) &#x000f7; S), rather than BMI<italic>z</italic>.</p></sec></body><back><ack id="S13"><p><bold>Funding</bold></p><p>NB and ET were supported by funds from the Centers for Disease Control and Prevention [RFA-DP-11&#x02013;007]. ET was supported by funds from the Centers for Disease Control and Prevention [U18DP003370] and by [grant K24 DK10589] from the National Institute of Diabetes and Digestive and Kidney Diseases.</p></ack><fn-group><fn fn-type="COI-statement" id="FN2"><p><bold>Disclosure statement</bold></p><p>The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.</p></fn><fn id="FN3" fn-type="other"><p>This work was authored as part of the Contributor&#x02019;s official duties as an Employee of the United States Government and is therefore a work of the United States Government. In accordance with 17 U.S.C. 105, no copyright protection is available for such works under U.S. Law.</p></fn></fn-group><ref-list><ref id="R1"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Baughcum</surname><given-names>AE</given-names></name><name><surname>Gramling</surname><given-names>K</given-names></name><name><surname>Eneli</surname><given-names>I</given-names></name></person-group><year>2015</year><article-title>Severely obese preschoolers in a tertiary care obesity program: characteristics and management</article-title><source>Clin Pediatr</source><volume>54</volume><fpage>346</fpage><lpage>352</lpage></element-citation></ref><ref id="R2"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Berkey</surname><given-names>CS</given-names></name><name><surname>Colditz</surname><given-names>GA</given-names></name></person-group><year>2007</year><article-title>Adiposity in adolescents: change in actual BMI works better than change in BMI z score for longitudinal studies</article-title><source>Ann Epidemiol</source><volume>17</volume><fpage>44</fpage><lpage>50</lpage><pub-id pub-id-type="pmid">17140812</pub-id></element-citation></ref><ref id="R3"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Box</surname><given-names>GEP</given-names></name><name><surname>Cox</surname><given-names>DR</given-names></name></person-group><year>1964</year><article-title>An analysis of transformations</article-title><source>J R Stat Soc Series B</source><volume>26</volume><fpage>211</fpage><lpage>252</lpage></element-citation></ref><ref id="R4"><element-citation publication-type="web"><person-group person-group-type="author"><name><surname>Canty</surname><given-names>A</given-names></name><name><surname>Ripley</surname><given-names>B</given-names></name></person-group><year>2017</year><source>Boot: Bootstrap Functions (Originally by Angelo Canty for S)</source><comment>Available at: <ext-link ext-link-type="uri" xlink:href="https://cran.r-project.org/package=boot">https://cran.r-project.org/package=boot</ext-link></comment></element-citation></ref><ref id="R5"><element-citation publication-type="web"><collab>CDC</collab><source>National Center for Health Statistics, Growth Charts - Z-score Data Files</source><date-in-citation>May 10, 2017</date-in-citation><comment>Available at: <ext-link ext-link-type="uri" xlink:href="https://www.cdc.gov/growthcharts/zscore.htm">https://www.cdc.gov/growthcharts/zscore.htm</ext-link></comment></element-citation></ref><ref id="R6"><element-citation publication-type="web"><collab>Centers for Disease Control and Prevention (CDC)</collab><year>2016</year><source>A SAS Program for the 2000 CDC Growth Charts</source><comment>Available at: <ext-link ext-link-type="uri" xlink:href="https://www.cdc.gov/nccdphp/dnpao/growthcharts/resources/sas.htm">https://www.cdc.gov/nccdphp/dnpao/growthcharts/resources/sas.htm</ext-link></comment></element-citation></ref><ref id="R7"><element-citation publication-type="web"><collab>Centers for Disease Control and Prevention (CDC)</collab><year>nd</year><source>Modified z-scores in the CDC growth charts</source><date-in-citation>April 25, 2017</date-in-citation><comment>Available at: <ext-link ext-link-type="uri" xlink:href="https://www.cdc.gov/nccdphp/dnpao/growthcharts/resources/biv-cutoffs.pdf">https://www.cdc.gov/nccdphp/dnpao/growthcharts/resources/biv-cutoffs.pdf</ext-link></comment></element-citation></ref><ref id="R8"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cole</surname><given-names>TJ</given-names></name></person-group><year>1995</year><article-title>Conditional reference charts to assess weight gain in British infants</article-title><source>Arch Dis Child</source><volume>73</volume><fpage>8</fpage><lpage>16</lpage><pub-id pub-id-type="pmid">7639558</pub-id></element-citation></ref><ref id="R9"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cole</surname><given-names>TJ</given-names></name></person-group><year>1994</year><article-title>Growth charts for both cross-sectional and longitudinal data</article-title><source>Stat Med</source><volume>13</volume><fpage>2477</fpage><lpage>2492</lpage><pub-id pub-id-type="pmid">7701148</pub-id></element-citation></ref><ref id="R10"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cole</surname><given-names>TJ</given-names></name></person-group><year>1990</year><article-title>The LMS method for constructing normalized growth standards</article-title><source>Eur J Clin Nutr</source><volume>44</volume><fpage>45</fpage><lpage>60</lpage></element-citation></ref><ref id="R11"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cole</surname><given-names>TJ</given-names></name><name><surname>Faith</surname><given-names>MS</given-names></name><name><surname>Pietrobelli</surname><given-names>A</given-names></name><name><surname>Heo</surname><given-names>M</given-names></name></person-group><year>2005</year><article-title>What is the best measure of adiposity change in growing children: BMI, BMI%, BMI z-score or BMI centile?</article-title><source>Eur J Clin Nutr</source><volume>59</volume><fpage>419</fpage><lpage>425</lpage><pub-id pub-id-type="pmid">15674315</pub-id></element-citation></ref><ref id="R12"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cole</surname><given-names>TJ</given-names></name><name><surname>Freeman</surname><given-names>JV</given-names></name><name><surname>Preece</surname><given-names>MA</given-names></name></person-group><year>1998</year><article-title>British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood</article-title><source>Stat Med</source><volume>17</volume><fpage>407</fpage><lpage>429</lpage><pub-id pub-id-type="pmid">9496720</pub-id></element-citation></ref><ref id="R13"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cole</surname><given-names>TJ</given-names></name><name><surname>Green</surname><given-names>PJ</given-names></name></person-group><year>1992</year><article-title>Smoothing reference centile curves: the LMS method and penalized likelihood</article-title><source>Stat Med</source><volume>11</volume><fpage>1305</fpage><lpage>1319</lpage><pub-id pub-id-type="pmid">1518992</pub-id></element-citation></ref><ref id="R14"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Flegal</surname><given-names>KM</given-names></name><name><surname>Wei</surname><given-names>R</given-names></name><name><surname>Ogden</surname><given-names>CL</given-names></name><name><surname>Freedman</surname><given-names>DS</given-names></name><name><surname>Johnson</surname><given-names>CL</given-names></name><name><surname>Curtin</surname><given-names>LR</given-names></name></person-group><year>2009</year><article-title>Characterizing extreme values of body mass index-for-age by using the 2000 Centers for Disease Control and Prevention growth charts</article-title><source>Am J Clin Nutr</source><volume>90</volume><fpage>1314</fpage><lpage>1320</lpage><pub-id pub-id-type="pmid">19776142</pub-id></element-citation></ref><ref id="R15"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Flegal</surname><given-names>KM</given-names></name><name><surname>Cole</surname><given-names>TJ</given-names></name></person-group><year>2013</year><article-title>Construction of LMS parameters for the Centers for Disease Control and Prevention 2000 Growth Charts</article-title><source>Natl Health Stat Report</source><volume>9</volume><fpage>1</fpage><lpage>3</lpage></element-citation></ref><ref id="R16"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Freedman</surname><given-names>DS</given-names></name><name><surname>Butte</surname><given-names>NF</given-names></name><name><surname>Taveras</surname><given-names>EM</given-names></name><name><surname>Lundeen</surname><given-names>EA</given-names></name><name><surname>Blanck</surname><given-names>HM</given-names></name><name><surname>Goodman</surname><given-names>AB</given-names></name><name><surname>Ogden</surname><given-names>CL</given-names></name></person-group><year>2017a</year><article-title>BMI z-Scores are a poor indicator of adiposity among 2- to 19-year-olds with very high BMIs, NHANES 1999-2000 to 2013-2014</article-title><source>Obesity (Silver Spring, Md)</source><volume>25</volume><fpage>739</fpage><lpage>746</lpage></element-citation></ref><ref id="R17"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Freedman</surname><given-names>DS</given-names></name><name><surname>Butte</surname><given-names>NF</given-names></name><name><surname>Taveras</surname><given-names>EM</given-names></name><name><surname>Goodman</surname><given-names>AB</given-names></name><name><surname>Ogden</surname><given-names>CL</given-names></name><name><surname>Blanck</surname><given-names>HM</given-names></name></person-group><year>2017b</year><article-title>The limitations of transforming very high body mass indexes into z-scores among 8.7 million 2- to 4-year-old children</article-title><source>J Pediatr</source><volume>188</volume><fpage>50</fpage><lpage>56.e1</lpage><pub-id pub-id-type="pmid">28433203</pub-id></element-citation></ref><ref id="R18"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Freedman</surname><given-names>DS</given-names></name><name><surname>Lawman</surname><given-names>HG</given-names></name><name><surname>Pan</surname><given-names>L</given-names></name><name><surname>Skinner</surname><given-names>AC</given-names></name><name><surname>Allison</surname><given-names>DB</given-names></name><name><surname>McGuire</surname><given-names>LC</given-names></name><name><surname>Blanck</surname><given-names>HM</given-names></name></person-group><year>2016</year><article-title>The prevalence and validity of high, biologically implausible values of weight, height, and BMI among 8.8 million children</article-title><source>Obesity (Silver Spring, Md)</source><volume>24</volume><fpage>1132</fpage><lpage>1139</lpage></element-citation></ref><ref id="R19"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Freedman</surname><given-names>DS</given-names></name><name><surname>Berenson</surname><given-names>GS</given-names></name></person-group><year>2017</year><article-title>The longitudinal tracking of BMI z-scores and other BMI metrics among 2- to 17-year-olds</article-title><source>Pediatrics</source><pub-id pub-id-type="doi">10.1542/peds.2017-1072</pub-id><comment>In Press</comment></element-citation></ref><ref id="R20"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Hampl</surname><given-names>S</given-names></name><name><surname>Odar Stough</surname><given-names>C</given-names></name><name><surname>Poppert Cordts</surname><given-names>K</given-names></name><name><surname>Best</surname><given-names>C</given-names></name><name><surname>Blackburn</surname><given-names>K</given-names></name><name><surname>Dreyer Gillette</surname><given-names>ML</given-names></name></person-group><year>2016</year><article-title>Effectiveness of a hospital-based multidisciplinary pediatric weight management program: Two-year outcomes of PHIT Kids</article-title><source>Child Obes</source><volume>12</volume><fpage>20</fpage><lpage>25</lpage><pub-id pub-id-type="pmid">26790094</pub-id></element-citation></ref><ref id="R21"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Hintze</surname><given-names>JL</given-names></name><name><surname>Nelson</surname><given-names>RD</given-names></name></person-group><year>1998</year><article-title>Violin plots: a box plot-density trace synergism</article-title><source>Am Stat</source><volume>52</volume><fpage>181</fpage><lpage>184</lpage></element-citation></ref><ref id="R22"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Kakinami</surname><given-names>L</given-names></name><name><surname>Henderson</surname><given-names>M</given-names></name><name><surname>Chiolero</surname><given-names>A</given-names></name><name><surname>Cole</surname><given-names>TJ</given-names></name><name><surname>Paradis</surname><given-names>G</given-names></name></person-group><year>2014</year><article-title>Identifying the best body mass index metric to assess adiposity change in children</article-title><source>Arch Dis Child</source><volume>99</volume><fpage>1020</fpage><lpage>1024</lpage><pub-id pub-id-type="pmid">24842797</pub-id></element-citation></ref><ref id="R23"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Kelly</surname><given-names>AS</given-names></name><name><surname>Barlow</surname><given-names>SE</given-names></name><name><surname>Rao</surname><given-names>G</given-names></name><name><surname>Inge</surname><given-names>TH</given-names></name><name><surname>Hayman</surname><given-names>LL</given-names></name><name><surname>Steinberger</surname><given-names>J</given-names></name><name><surname>Urbina</surname><given-names>EM</given-names></name><etal/></person-group><year>2013</year><article-title>Severe obesity in children and adolescents: identification, associated health risks, and treatment approaches: a Scientific Statement from the American Heart Association</article-title><source>Circulation</source><volume>128</volume><fpage>1689</fpage><lpage>1712</lpage><pub-id pub-id-type="pmid">24016455</pub-id></element-citation></ref><ref id="R24"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Kreier</surname><given-names>F</given-names></name><name><surname>Genco</surname><given-names>&#x0015e;M</given-names></name><name><surname>Boreel</surname><given-names>M</given-names></name><name><surname>Langkemper</surname><given-names>MP</given-names></name><name><surname>Nugteren</surname><given-names>IC</given-names></name><name><surname>Rijnveld</surname><given-names>V</given-names></name><name><surname>Thissen</surname><given-names>V</given-names></name><etal/></person-group><year>2013</year><article-title>An individual, community-based treatment for obese children and their families: the solution-focused approach</article-title><source>Obesity Facts</source><volume>6</volume><fpage>424</fpage><lpage>432</lpage><pub-id pub-id-type="pmid">24107796</pub-id></element-citation></ref><ref id="R25"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Kuczmarski</surname><given-names>RJ</given-names></name><name><surname>Ogden</surname><given-names>CL</given-names></name><name><surname>Guo</surname><given-names>SS</given-names></name><name><surname>Grummer-Strawn</surname><given-names>LM</given-names></name><name><surname>Flegal</surname><given-names>KM</given-names></name><name><surname>Mei</surname><given-names>Z</given-names></name><name><surname>Wei</surname><given-names>R</given-names></name><etal/></person-group><year>2002</year><article-title>2000 CDC Growth Charts for the United States: methods and development</article-title><source>Vital Health Stat</source><volume>11</volume><fpage>1</fpage><lpage>190</lpage></element-citation></ref><ref id="R26"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Lo</surname><given-names>JC</given-names></name><name><surname>Chandra</surname><given-names>M</given-names></name><name><surname>Sinaiko</surname><given-names>A</given-names></name><name><surname>Daniels</surname><given-names>SR</given-names></name><name><surname>Prineas</surname><given-names>RJ</given-names></name><name><surname>Maring</surname><given-names>B</given-names></name><name><surname>Parker</surname><given-names>ED</given-names></name><etal/></person-group><year>2014</year><article-title>Severe obesity in children: prevalence, persistence and relation to hypertension</article-title><source>Int J Pediatr Endocrinol</source><volume>2014</volume><fpage>3</fpage><pub-id pub-id-type="pmid">24580759</pub-id></element-citation></ref><ref id="R27"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Markert</surname><given-names>J</given-names></name><name><surname>Herget</surname><given-names>S</given-names></name><name><surname>Petroff</surname><given-names>D</given-names></name><name><surname>Gausche</surname><given-names>R</given-names></name><name><surname>Grimm</surname><given-names>A</given-names></name><name><surname>Kiess</surname><given-names>W</given-names></name><name><surname>Bl&#x000fc;her</surname><given-names>S</given-names></name></person-group><year>2014</year><article-title>Telephone-based adiposity prevention for families with overweight children (T.A.F.F.-Study): one year outcome of a randomized, controlled trial</article-title><source>Int J Environ Res Public Health</source><volume>11</volume><fpage>10327</fpage><lpage>10344</lpage><pub-id pub-id-type="pmid">25286167</pub-id></element-citation></ref><ref id="R28"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>McCormick</surname><given-names>EV</given-names></name><name><surname>Dickinson</surname><given-names>LM</given-names></name><name><surname>Haemer</surname><given-names>MA</given-names></name><name><surname>Knierim</surname><given-names>SD</given-names></name><name><surname>Hambidge</surname><given-names>SJ</given-names></name><name><surname>Davidson</surname><given-names>AJ</given-names></name></person-group><year>2014</year><article-title>What can providers learn from childhood body mass index trajectories: a study of a large, safety-net clinical population</article-title><source>Acad Pediatr</source><volume>14</volume><fpage>639</fpage><lpage>645</lpage><pub-id pub-id-type="pmid">25129568</pub-id></element-citation></ref><ref id="R29"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Mei</surname><given-names>Z</given-names></name><name><surname>Scanlon</surname><given-names>KS</given-names></name><name><surname>Grummer-Strawn</surname><given-names>LM</given-names></name><name><surname>Freedman</surname><given-names>DS</given-names></name><name><surname>Yip</surname><given-names>R</given-names></name><name><surname>Trowbridge</surname><given-names>FL</given-names></name></person-group><year>1998</year><article-title>Increasing prevalence of overweight among US low-income preschool children: the Centers for Disease Control and Prevention Pediatric Nutrition Surveillance, 1983 to 1995</article-title><source>Pediatrics</source><volume>101</volume><fpage>E12</fpage></element-citation></ref><ref id="R30"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>O&#x02019;Connor</surname><given-names>EA</given-names></name><etal/></person-group><year>2017</year><article-title>Screening for Obesity and Intervention for Weight Management in Children and Adolescents. Evidence Report and Systematic Review for the US Preventive Services Task Force (USPSTF)</article-title><source>JAMA</source><volume>317</volume><fpage>2427</fpage><pub-id pub-id-type="pmid">28632873</pub-id></element-citation></ref><ref id="R31"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Ogden</surname><given-names>CL</given-names></name><name><surname>Flegal</surname><given-names>KM</given-names></name></person-group><year>2010</year><article-title>Changes in terminology for childhood overweight and obesity</article-title><source>Natl Health Stat Report</source><volume>25</volume><fpage>1</fpage><lpage>5</lpage></element-citation></ref><ref id="R32"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Paluch</surname><given-names>RA</given-names></name><name><surname>Epstein</surname><given-names>LH</given-names></name><name><surname>Roemmich</surname><given-names>JN</given-names></name></person-group><year>2007</year><article-title>Comparison of methods to evaluate changes in relative body mass index in pediatric weight control</article-title><source>Am J Hum Biol</source><volume>19</volume><fpage>487</fpage><lpage>494</lpage><pub-id pub-id-type="pmid">17546615</pub-id></element-citation></ref><ref id="R33"><element-citation publication-type="book"><collab>R Core Team</collab><year>2017</year><source>R: A language and environment for statistical computing</source><publisher-loc>Vienna, Austria</publisher-loc><publisher-name>R Foundation for Statistical Computing</publisher-name><comment>Available at: <ext-link ext-link-type="uri" xlink:href="http://www.r-project.org/">http://www.r-project.org/</ext-link></comment></element-citation></ref><ref id="R34"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Shrout</surname><given-names>PE</given-names></name><name><surname>Fleiss</surname><given-names>JL</given-names></name></person-group><year>1979</year><article-title>Intraclass correlations: uses in assessing rater reliability</article-title><source>Psychol Bull</source><volume>86</volume><fpage>420</fpage><lpage>428</lpage><pub-id pub-id-type="pmid">18839484</pub-id></element-citation></ref><ref id="R35"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Smego</surname><given-names>A</given-names></name><name><surname>Woo</surname><given-names>JG</given-names></name><name><surname>Klein</surname><given-names>J</given-names></name><name><surname>Suh</surname><given-names>C</given-names></name><name><surname>Bansal</surname><given-names>D</given-names></name><name><surname>Bliss</surname><given-names>S</given-names></name><name><surname>Daniels</surname><given-names>SR</given-names></name><etal/></person-group><year>2017</year><article-title>High body mass index in infancy may predict severe obesity in early childhood</article-title><source>J Pediatr</source><volume>183</volume><fpage>87</fpage><lpage>93.e1</lpage><pub-id pub-id-type="pmid">27916426</pub-id></element-citation></ref><ref id="R36"><element-citation publication-type="journal"><collab>US Preventive Services Task Force</collab><year>2017</year><article-title>Screening for obesity in children and adolescents: US Preventive Services Task Force Recommendation Statement</article-title><source>JAMA</source><volume>317</volume><fpage>2417</fpage><lpage>2426</lpage><pub-id pub-id-type="pmid">28632874</pub-id></element-citation></ref><ref id="R37"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Wang</surname><given-names>Y</given-names></name><name><surname>Cai</surname><given-names>L</given-names></name><name><surname>Wu</surname><given-names>Y</given-names></name><name><surname>Wilson</surname><given-names>RF</given-names></name><name><surname>Weston</surname><given-names>C</given-names></name><name><surname>Fawole</surname><given-names>O</given-names></name><name><surname>Bleich</surname><given-names>SN</given-names></name><etal/></person-group><year>2015</year><article-title>What childhood obesity prevention programmes work? A systematic review and meta-analysis</article-title><source>Obes Rev</source><volume>16</volume><fpage>547</fpage><lpage>565</lpage><pub-id pub-id-type="pmid">25893796</pub-id></element-citation></ref><ref id="R38"><element-citation publication-type="web"><collab>WHO Department of Nutrition for Health and Development</collab><year>2009</year><article-title>WHO child growth standards. Growth velocity based on weight, length and head circumference</article-title><source>Methods and development</source><date-in-citation>May 17, 2017</date-in-citation><comment>Available at: <ext-link ext-link-type="uri" xlink:href="http://www.who.int/childgrowth/standards/velocity/tr3_velocity_report.pdf">http://www.who.int/childgrowth/standards/velocity/tr3_velocity_report.pdf</ext-link></comment></element-citation></ref><ref id="R39"><element-citation publication-type="book"><collab>WHO Multicentre Growth Reference Study Group</collab><year>2006</year><source>WHO Child Growth Standards: Length/height-for-age, weight-for- age, weight-for-length, weight-for-height and body mass index-for-age: Methods and development</source><publisher-loc>Geneva</publisher-loc><publisher-name>World Health Organization</publisher-name></element-citation></ref><ref id="R40"><element-citation publication-type="web"><person-group person-group-type="author"><name><surname>Wolak</surname><given-names>M</given-names></name></person-group><year>2017</year><source>ICC: Facilitating Estimation of the Intraclass CorrelationCoefficient [R package ICC version 2.3.0]</source><date-in-citation>February 22, 2017</date-in-citation><comment>Available at: <ext-link ext-link-type="uri" xlink:href="https://cran.r-project.org/package=ICC">https://cran.r-project.org/package=ICC</ext-link></comment></element-citation></ref><ref id="R41"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Woo</surname><given-names>JG</given-names></name></person-group><year>2009</year><article-title>Using body mass index Z-score among severely obese adolescents: a cautionary note</article-title><source>Int J Pediatr Obes</source><volume>4</volume><fpage>405</fpage><lpage>410</lpage><pub-id pub-id-type="pmid">19922058</pub-id></element-citation></ref><ref id="R42"><element-citation publication-type="book"><collab>World Health Organization</collab><year>n.d</year><source>Growth reference 5-19 years WHO | BMI-for-age (5-19 years)</source><publisher-name>WHO</publisher-name><date-in-citation>August 2, 2017</date-in-citation><comment>Available at: <ext-link ext-link-type="uri" xlink:href="http://www.who.int/growthref/who2007_bmi_for_age/en/">http://www.who.int/growthref/who2007_bmi_for_age/en/</ext-link></comment></element-citation></ref><ref id="R43"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Zemel</surname><given-names>BS</given-names></name></person-group><year>2009</year><article-title>A commentary on the construction of weight velocity charts</article-title><source>Nutr Clin Pract</source><volume>24</volume><fpage>651</fpage><lpage>653</lpage><pub-id pub-id-type="pmid">19841252</pub-id></element-citation></ref></ref-list></back><floats-group><fig id="F1" orientation="portrait" position="float"><label>Figure 1</label><caption><p>Change in BMI<italic>z</italic> (top) and %BMIp95 (bottom) between the two examinations among children who were severely obese at their initial examination. Results are stratified by sex and initial year of age. The violin plots represent the density distribution of the observed changes within each BMI<italic>z</italic> group and the horizontal lines represent the 10th, 50th and 90th percentiles.</p></caption><graphic xlink:href="nihms917193f1"/></fig><fig id="F2" orientation="portrait" position="float"><label>Figure 2</label><caption><p>Line plots showing changes in BMI<italic>z</italic> and %BMIp95 among the 100 girls with the highest BMI<italic>z</italic> values (&#x02265;4.75) at the initial examination. Each line represents the value of the BMI metric at the two examinations for a single girl. Age is shown on the <italic>x</italic>-axis.</p></caption><graphic xlink:href="nihms917193f2"/></fig><table-wrap id="T1" position="float" orientation="portrait"><label>Table 1</label><caption><p>Levels of various characteristics among the 45 414 children who had severe obesity at their first examinations.</p></caption><table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="1" colspan="1"/><th align="center" rowspan="1" colspan="1">Boys</th><th align="center" rowspan="1" colspan="1">Girls</th></tr></thead><tbody><tr><td align="left" rowspan="1" colspan="1"><italic>n</italic></td><td align="center" rowspan="1" colspan="1">21 007</td><td align="center" rowspan="1" colspan="1">24 407</td></tr><tr><td align="left" rowspan="1" colspan="1">Age in months</td><td align="center" rowspan="1" colspan="1">37.4 (24.4, 51.9)<xref rid="TFN1" ref-type="table-fn">a</xref></td><td align="center" rowspan="1" colspan="1">36.9 (24.3, 51.5)</td></tr><tr><td align="left" rowspan="1" colspan="1">BMI (kg/m<sup>2</sup>)</td><td align="center" rowspan="1" colspan="1">23.5 (21.6, 26.6)</td><td align="center" rowspan="1" colspan="1">23.7 (21.8, 27.7)</td></tr><tr><td align="left" rowspan="1" colspan="1">BMI<italic>z</italic> (SDs)</td><td align="center" rowspan="1" colspan="1">3.9 (3.2, 4.8)</td><td align="center" rowspan="1" colspan="1">3.4 (2.8, 4.4)</td></tr><tr><td align="left" rowspan="1" colspan="1">%BMIp95</td><td align="center" rowspan="1" colspan="1">128.2 (120.2, 145.7)</td><td align="center" rowspan="1" colspan="1">129.3 (120.3, 151.3)</td></tr><tr><td align="left" rowspan="1" colspan="1">Modified BMI<italic>z</italic> (SDs)</td><td align="center" rowspan="1" colspan="1">5.2 (4.3, 7.0)</td><td align="center" rowspan="1" colspan="1">5.1 (4.1, 7.2)</td></tr><tr><td align="left" rowspan="1" colspan="1">Weight-for-age <italic>z</italic>-score</td><td align="center" rowspan="1" colspan="1">3.4 (1.3, 5.1)</td><td align="center" rowspan="1" colspan="1">3.1 (1.4, 4.4)</td></tr><tr><td align="left" rowspan="1" colspan="1">Height-for age <italic>z</italic>-score</td><td align="center" rowspan="1" colspan="1">0.8 (&#x02013;2.6, 3.1)</td><td align="center" rowspan="1" colspan="1">0.9 (&#x02013;2.2, 3.2)</td></tr><tr><td align="left" rowspan="1" colspan="1">Changes in the BMI metrics<xref rid="TFN2" ref-type="table-fn">b</xref></td><td align="center" rowspan="1" colspan="1"/><td align="center" rowspan="1" colspan="1"/></tr><tr><td align="left" rowspan="1" colspan="1">BMI (kg/m<sup>2</sup>)</td><td align="center" rowspan="1" colspan="1">&#x02212;0.7 (&#x02212;5.9, 3.2)</td><td align="center" rowspan="1" colspan="1">&#x02212;0.5 (&#x02212;5.6, 3.8)</td></tr><tr><td align="left" rowspan="1" colspan="1">BMI<italic>z</italic> (SD)</td><td align="center" rowspan="1" colspan="1">&#x02212;0.3 (&#x02212;2.3, 1.2)</td><td align="center" rowspan="1" colspan="1">&#x02212;0.5 (&#x02212;2.0, 0.3)</td></tr><tr><td align="left" rowspan="1" colspan="1">%BMIp95</td><td align="center" rowspan="1" colspan="1">&#x02212;0.9 (&#x02212;28.7, 20.7)</td><td align="center" rowspan="1" colspan="1">&#x02212;0.8 (&#x02212;27.9, 23.0)</td></tr><tr><td align="left" rowspan="1" colspan="1">Modified BMI<italic>z</italic></td><td align="center" rowspan="1" colspan="1">&#x02212;0.1 (&#x02212;3.4, 2.2)</td><td align="center" rowspan="1" colspan="1">0.0 (&#x02212;2.9, 2.3)</td></tr></tbody></table><table-wrap-foot><fn id="TFN1"><label>a</label><p>Values in parentheses include 95% of the values (2.5 percentile to 97.5 percentile).</p></fn><fn id="TFN2"><label>b</label><p>Calculated as Exam 2 value &#x02013; Exam 1 value. The mean interval between examinations varied was 10.8 months.</p></fn></table-wrap-foot></table-wrap><table-wrap id="T2" position="float" orientation="landscape"><label>Table 2</label><caption><p>Cross-classification of BMI<italic>z</italic> levels at the first and second examinations among children who were severely obese (<italic>n</italic> =45 414) at their initial examination</p></caption><table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="1" colspan="1"/><th align="left" rowspan="1" colspan="1"/><th align="left" rowspan="1" colspan="1"/><th align="center" colspan="3" valign="bottom" rowspan="1">BMI<italic>z</italic>
<hr/></th><th align="center" colspan="4" valign="bottom" rowspan="1">%BMIp95<xref rid="TFN3" ref-type="table-fn">a</xref>
<hr/></th></tr><tr><th align="left" rowspan="1" colspan="1"/><th align="left" rowspan="1" colspan="1"/><th align="left" rowspan="1" colspan="1"/><th align="center" colspan="3" valign="bottom" rowspan="1">Second examination
<hr/></th><th align="center" rowspan="1" colspan="1"/><th align="center" colspan="3" valign="bottom" rowspan="1">Second examination
<hr/></th></tr><tr><th align="left" rowspan="1" colspan="1"/><th align="right" rowspan="1" colspan="1"><italic>n</italic></th><th align="center" rowspan="1" colspan="1">Initial category</th><th align="right" rowspan="1" colspan="1">&#x0003c;3.5</th><th align="right" rowspan="1" colspan="1">3.5&#x02013;4.4</th><th align="right" rowspan="1" colspan="1">&#x02265;4.5</th><th align="right" rowspan="1" colspan="1">Initial category</th><th align="right" rowspan="1" colspan="1">Low</th><th align="right" rowspan="1" colspan="1">Middle</th><th align="right" rowspan="1" colspan="1">High</th></tr></thead><tbody><tr><td align="left" rowspan="1" colspan="1">Boys</td><td align="right" rowspan="1" colspan="1">4 202</td><td align="center" rowspan="1" colspan="1">&#x0003c;3.5</td><td align="right" rowspan="1" colspan="1">52%<xref rid="TFN4" ref-type="table-fn">b</xref></td><td align="right" rowspan="1" colspan="1">37%</td><td align="right" rowspan="1" colspan="1">11%</td><td align="right" rowspan="1" colspan="1">120&#x02013;121</td><td align="right" rowspan="1" colspan="1">52%</td><td align="right" rowspan="1" colspan="1">40%</td><td align="right" rowspan="1" colspan="1">8%</td></tr><tr><td align="left" rowspan="1" colspan="1"/><td align="right" rowspan="1" colspan="1">14 690</td><td align="center" rowspan="1" colspan="1">3.5&#x02013;4.4</td><td align="right" rowspan="1" colspan="1">38%</td><td align="right" rowspan="1" colspan="1">50%</td><td align="right" rowspan="1" colspan="1">12%</td><td align="right" rowspan="1" colspan="1">122&#x02013;138</td><td align="right" rowspan="1" colspan="1">31%</td><td align="right" rowspan="1" colspan="1">48%</td><td align="right" rowspan="1" colspan="1">20%</td></tr><tr><td align="left" rowspan="1" colspan="1"/><td align="right" rowspan="1" colspan="1">2 115</td><td align="center" rowspan="1" colspan="1">&#x02265;4.5</td><td align="right" rowspan="1" colspan="1">10%</td><td align="right" rowspan="1" colspan="1">66%</td><td align="right" rowspan="1" colspan="1">24%</td><td align="right" rowspan="1" colspan="1">&#x02265;139</td><td align="right" rowspan="1" colspan="1">6%</td><td align="right" rowspan="1" colspan="1">35%</td><td align="right" rowspan="1" colspan="1">59%</td></tr><tr><td align="left" rowspan="1" colspan="1"/><td align="right" rowspan="1" colspan="1">15 331</td><td align="center" rowspan="1" colspan="1">&#x0003c;3.5</td><td align="right" rowspan="1" colspan="1">96%</td><td align="right" rowspan="1" colspan="1">4%</td><td align="right" rowspan="1" colspan="1">0.01%</td><td align="right" rowspan="1" colspan="1">120&#x02013;129</td><td align="right" rowspan="1" colspan="1">70%</td><td align="right" rowspan="1" colspan="1">29%</td><td align="right" rowspan="1" colspan="1">0.9%</td></tr><tr><td align="left" rowspan="1" colspan="1">Girls</td><td align="right" rowspan="1" colspan="1">8 745</td><td align="center" rowspan="1" colspan="1">3.5&#x02013;4.4</td><td align="right" rowspan="1" colspan="1">73%</td><td align="right" rowspan="1" colspan="1">27%</td><td align="right" rowspan="1" colspan="1">0.1%</td><td align="right" rowspan="1" colspan="1">130&#x02013;154</td><td align="right" rowspan="1" colspan="1">28%</td><td align="right" rowspan="1" colspan="1">63%</td><td align="right" rowspan="1" colspan="1">9%</td></tr><tr><td align="left" rowspan="1" colspan="1"/><td align="right" rowspan="1" colspan="1">331</td><td align="center" rowspan="1" colspan="1">&#x02265;4.5</td><td align="right" rowspan="1" colspan="1">39%</td><td align="right" rowspan="1" colspan="1">58%</td><td align="right" rowspan="1" colspan="1">3%</td><td align="right" rowspan="1" colspan="1">&#x02265;155</td><td align="right" rowspan="1" colspan="1">2%</td><td align="right" rowspan="1" colspan="1">49%</td><td align="right" rowspan="1" colspan="1">50%</td></tr></tbody></table><table-wrap-foot><fn id="TFN3"><label>a</label><p>The sex-specific cut points for %BMIp95 at the first examination were chosen to make the number of children in each group equal to that in each BMI<italic>z</italic> group. Cut-points for %BMIp95 at the second examination correspond to those at the first examination: &#x0003c;122, 122&#x02013;138 and &#x02265;139 for boys, and &#x0003c;130, 130&#x02013;154 and &#x02265;155 for girls.</p></fn><fn id="TFN4"><label>b</label><p>Values represent the percentage of children in each initial category who are in the specified category at the second examination.</p></fn></table-wrap-foot></table-wrap></floats-group></article>