doi: 10.1111/ppe.

12377 1

Differences in growth of Canadian children compared to the WHO

2006 Child Growth Standards
Alison L. Park,a,b Karen Tu,b,c Joel G. Ray,a,b,c,d for the Canadian Curves Consortium†
a
Keenan Research Centre, Li Ka Shing Knowledge Institute, St. Michael’s Hospital,
b
Institute for Clinical Evaluative Sciences,
c
University of Toronto,
d
Departments of Medicine and Obstetrics and Gynaecology, St. Michael’s Hospital, Toronto, ON, Canada

Abstract
Background: To evaluate if there are departures from the WHO Child Growth Standards (WHO-CGS) in postnatal
growth of healthy ‘Canadian’ children in Ontario up to age 2 years, including by infant feeding and ethnicity.
Methods: We included data on 9964 healthy, singleton children born in Ontario, Canada. Smoothed weight, length
and body mass index (BMI) percentile curves were generated using quantile regression for the Canadian cohort
from birth to age 2 years. Differences in percentile values were calculated comparing Canadian children vs. the
WHO-CGS.
Results: Canadian children under age 2 years were longer than the WHO-CGS at the 10th (0.8 cm), 50th (1.3 cm)
and 90th (1.9 cm) percentiles. Canadian children incrementally surpassed the WHO-CGS in weight after age
6 months, and in BMI after 9 months. By age 2 years, the 50th percentile weight of Canadian males was 823 g
(95% confidence interval (CI) 680, 965) higher than the WHO-CGS 50th percentile. Weight differences were seen
regardless of feeding practice, and were greatest among children of mothers born in Canada and Europe/Western
nations, and least for those of East Asian/Pacific or South Asian heritage. Among Canadian breastfed males, 18%
(95% CI 16, 19) of newborns and 26% (95% CI 20, 33) toddlers aged 2 years were classified by WHO-CGS as
weighing >90th percentile – much higher than the expected rate of 10%. Similarities were seen for differences in
BMI.
Conclusions: Healthy Canadian infants/toddlers are longer and heavier than the WHO-CGS norms. Explanations
for these discrepancies require further elucidation.

Keywords: infant and child growth, growth chart, growth standard, WHO, growth standards, length, weight, percentiles.

An important aspect of routine care for young chil-
dren is serial measurement of weight and length.1
Monitoring early growth helps practitioners assess
the health of children, including feeding adequacy.2,3
In 2006, the World Health Organization (WHO)
released universal child growth standards (CGS),
intended to describe the optimal growth of children.4
The WHO Multicentre Growth Reference Study
(MGRS) took place between 1997 and 2003, and
included children from Brazil, Ghana, India, Norway,
Oman and the US, who were deemed to be free of
Correspondence:
Joel G. Ray, Departments of Medicine and Obstetrics and
Gynaecology, St. Michael’s Hospital, University of Toronto,
Toronto, Ontario, Canada.
E-mail: rayj@smh.ca
†
Members of the Canadian Curves Consortium are in
Appendix 1.
health or environmental constraints on growth. They
used a longitudinal design between birth and
24 months, and a cross-sectional design between 18
and 71 months. Of the 1743 infants enrolled at birth,
882 (50%) complied with the criteria for the longitudi-
nal group, including exclusive or predominant breast
feeding until at least 4 months.4
Canadian authorities endorsed using the WHO-
CGS to monitor infant growth from 0 to 2 years,5
replacing the Centres for Disease Control and Preven-
tion 2002 growth charts that had predominantly
included formula-fed infants, who are prone to exces-
sive growth.6 Although the WHO-CGS are thought to
describe optimal growth, regardless of ethnicity, a
number of international studies have found marked
departures from the WHO-CGS, not entirely
explained by feeding practice.7–11 Despite widespread

© 2017 John Wiley & Sons Ltd

Paediatric and Perinatal Epidemiology, 2017, , –
2 A. L. Park et al.
adoption of the WHO-CGS in Canada, it is unclear if
the WHO-CGS reflects the growth of healthy Cana-
dian infants and toddlers. Comparisons between the
growth of Canadian children and those of the
WHO-CGS have been restricted to single-centre
hospital-based cohorts,12,13 which may limit their
generalisability. One exception was a prospective
study of only 73 infants, which sought to match the
inclusion criteria of the WHO-CGS.14 In addition, no
study has specifically evaluated performance of the
WHO-CGS in Canadian children born to parents from
some of Canada’s major immigrant source countries,
such as China, the Philippines and Pakistan, none of
whom were included in the MGRS. This is potentially
relevant, as we previously showed that World region-
specific birthweight curves more accurately predict
adverse neonatal and obstetric outcomes than a single
birthweight curve based on infants of Canadian-born
women.15
Herein, we compared the WHO-CGS to the postna-
tal growth of 9964 healthy Ontarian children up to age
2 years, including variations by infant feeding prac-
tices and maternal world region of origin. As Ontario
comprises over a third of the births and half of immi-
grants in Canada, and universal health coverage and
healthcare practices (i.e. primary care, obstetrics and
paediatrics) are largely the same across the provinces,
we henceforth refer to our Ontario cohort as
‘Canadian’.16
Methods
Study design and participants
We completed a retrospective population-based
cohort study using administrative health care data
for the province of Ontario, Canada. Analyses took
place at the Institute for Clinical Evaluative Sciences
(ICES), using datasets linked by unique encoded
identifiers. We identified the linked inpatient records
of delivering Ontarian mothers and their newborns
from the ICES MOMBABY dataset,15,17 which uses
data from the Canadian Institute for Health Informa-
tion’s Discharge Abstract Database (DAD). Included
were singleton live births in an Ontario hospital at
371/7 to 416/7 weeks’ gestation, between April 1,
2002 and March 31, 2013, who had at least one well-
baby/child visit before 750 days of age, with a mea-
sured weight or length in the Electronic Medical
Record Administrative data Linked Database
(EMRALD).18 In Canada, it is recommended that
children have a well-baby/child visit within the 1st
week of life, and at months 1, 2, 4, 6, 12–13, 18 (in
Ontario) and at 2–3 years, with weight and length
recorded at each visit.5,19 We excluded those with a
diagnosed congenital or chromosomal anomaly,
identified from hospital records at birth and up to
750 days of age, using the diagnostic codes Q00-Q99
in the International Statistical Classification of Dis-
eases, 10th Revision. We also excluded implausible
weight, length or body mass index (BMI) values
above +5 SD, or below 5 SD, of the sample median
for that age group. To enable comparison with the
WHO-CGS, we used their inclusion criteria, as
follows: single term births without a significant
morbidity, no known health or environmental
constraints to growth, including maternal smoking,
who were exclusively or predominantly breastfed
for at least 4 months, with introduction of comple-
mentary foods by age 6 months, along with partial
breast feeding to at least age 12 months.4
Eligible newborn records were linked by their
encrypted health care number to the EMRALD,18
which, at the time of this study, comprised electronic
medical records (EMRs) from 321 family physicians in
42 geographically distinct primary care practices
across Ontario. For each eligible infant in EMRALD,
we extracted data for their postnatal visits including
date of the visit, age, current weight and length, and
feeding practice. For each infant, we also examined
each entry in their EMR chart up to age 750 days for
any mention of present or past breast or formula feed-
ing, using an algorithm that we developed to search
structured and free text fields.20 Canadian family
physicians commonly use the structured Rourke Baby
Record system (http://www.rourkebabyrecord.ca)
for routine child health surveillance – including
growth and feeding – from birth to 5 years of age. For
EMRs without a Rourke Baby Record, chart abstrac-
tors manually extracted information about the type
and duration of infant feeding from free text fields.
Each infant’s predominant feeding type in the first
6 months of life was classified as: (i) exclusively
breastfed, (ii) mixed breastfed and formula-fed, (iii)
exclusively formula-fed, or (iv) unknown.
Maternal world region of origin was used as a practi-
cal proxy for ethnicity, categorised as follows: (i)
Canada, (ii) Europe and Western nations, (iii) Middle
East and North Africa, (iv) Sub-Saharan Africa and the
Caribbean, (v) Latin America, (vi) East Asia and the
© 2017 John Wiley & Sons Ltd
Paediatric and Perinatal Epidemiology, 2017, , –

Pacific, (vii) South Asia, and (viii) Other. Maternal
records were linked to the Ontario portion of the fed-
eral Immigration, Refugees and Citizenship Canada
(IRCC) Permanent Resident Database (PRD), also
housed at ICES.15,17 This database contains information
about original country of citizenship, for immigrants to
Canada from 1985 onward. Women not linked to the
PRD were classified as non-immigrant women. Over
90% of the latter comprise Canadian-born women, and
henceforth, are referred to as ‘Canadian-born’, but their
ethnic background is not documented within our
administrative datasets.
Residential postal code at the time of birth was used
to determine the neighbourhood income quintile and
rural residence, derived from Statistics Canada census
data.
Statistical analyses
We used nonparametric quantile regression methods
to generate smoothed weight-for-age (WFA), length-
for-age (LFA), and BMI-for-age (BMI-FA) percentile
curves for all children in our final cohort. Quantile
regression produces virtually similar results to the
Lamba-Mu-Sigma (LMS) method when the distribu-
tion of the response variable is roughly normal.21
Curves were fit employing a cubic spline and the use
of a smoothing algorithm, with knot locations deter-
mined by backward stepwise regression. Separate
curves were generated for males and females, and
also by type of infant feeding and by maternal world
region of origin. BMI – the ratio of weight (in kg) to
recumbent length (in m2) – was calculated for a subset
of infants who had both a weight and a length
recorded at the same visit. BMI is highly correlated
with weight-for-length from 0 to 2 years of age.22
Percentiles were based on growth measurements at
birth (weight only), and subsequent primary care vis-
its centered around key time points of 7 days (range
1–11), 28 days (range 17–34), 2 months (range 50–
69 days), 4 months (range 100–129 days), 6 months
(range 160–189 days), 9 months (range 250–279 days),
12 months (range 350–379 days), 15 months (range
350–379 days), 18 months (range 500–559 days) and
24 months (range 680–749 days). If a child had more
than one weight (or length) measurement within the
aforementioned age groups, then we selected that
closest to the target age. Age groups were chosen to
correspond with those defined in the WHO-CGS
methodology.23
© 2017 John Wiley & Sons Ltd
Paediatric and Perinatal Epidemiology, 2017, , –
Growth of Canadian children vs. WHO-CGS 3
We calculated differences and 95% confidence inter-
vals (CI) in the 10th, 50th and 90th percentile WFA and
LFA; and in the 3rd, 50th and 97th percentile BMI-FA,
between Canadian children and those of the WHO-
CGS. Estimated 95% CIs were calculated using the sex-
and age-specific sample-based errors from our cohort
and the coefficients of variation from the WHO-CGS.23
Differences were calculated by sex and infant feeding
type, as well as maternal world region of origin.
Next, we estimated the proportion of exclusively
breastfed Canadian children that would be considered
an unhealthy size according to the WHO-CGS thresh-
olds. For weight and length, we determined the pro-
portion (95% CI) in each age group who were <10th
percentile or >90th percentile according to the WHO-
CGS. For BMI, we calculated the proportion in each
age group who would be classified as wasted (<3rd
percentile) or overweight (>97th percentile) using the
WHO-CGS. All weights, lengths and BMI values were
converted to WHO-CGS percentiles. CIs for binomial
proportions are exact Clopper-Pearson 95% confi-
dence limits.
Statistical analyses were performed using SAS for
Linux, Version 9.4 (SAS Institute, Cary, NC) and R for
Linux, Version 3.1.2 (R Foundation for Statistical
Computing, Vienna, Austria). The study was
approved by the institutional review board at Sunny-
brook Health Sciences Centre, Toronto, Ontario.
Results
Formation of the final cohort of 9964 healthy, single-
ton term-born infants in EMRALD is shown in Data
S1. There were 48 556 WFA measurements, 32 681
LFA measurements, and 31 037 BMI-FA measure-
ments. The average child had 5 weight and 4 length
measurements (data not shown). Compared to all hos-
pital births in Ontario, those that linked to the
EMRALD database were somewhat more likely to
have a Canadian-born mother and reside in a rural
area, but were otherwise comparable (Data S2). We
were unable to determine infant feeding practices for
2188 (22%) infants.
Weight-for-age
At birth, the 50th percentile weight of Canadian males
(3530 g vs. 3346 g) and females (3380 g vs. 3232 g)
was higher than that of the WHO-CGS (Data S3). By
6 months of age, there was consistent departure in
4 A. L. Park et al.

weight differences at the 10th, 50th and 90th per- estimate, except for breastfed and mixed-fed females
centile (Figure 1). For example, by age 2 years, the dif- at 24 months of age (Data S5).
ference at the 50th percentile was 823 g (95% CI 680,
965) among males, and 776 g (95% CI 635, 917) among
BMI-for-age
females (Data S3). These differences were even seen
for exclusively breastfed infants (Figure 1). The 50th percentile BMI was significantly lower in
The 50th percentile weight difference between the Canadian children than the WHO-CGS up to
Canadian cohort and the WHO-CGS varied signifi- 6 months for males and 9 months for females; after
cantly at birth by maternal world region of origin, age 12 months, the 50th percentile BMI was signifi-
mostly among infants of mothers born in Canada, Eur- cantly higher in the Canadian cohort vs. the WHO-
ope and Western nations (Data S4a and b). Thereafter, CGS, regardless of infant feeding (Data S6). The
differences were not appreciable up to age 6 months, threshold for infant wasting was generally higher in
but were once again evident after 9 months of age. the WHO-CGS than among Canadian males up to
12 months and females up to age 9 months (Figure 3).
Thereafter, there was little difference in 3rd percentile
Length-for-age
BMI among males, and females, until age 18 months,
From birth to age 2 years, compared to their WHO- when positive differences emerged (Figure 3). The
CGS counterparts, Canadian children were on aver- pattern of <3rd percentile BMI differences varied
age 0.8 cm longer at the 10th percentile, 1.3 cm longer slightly for breastfed males and breastfed females.
at the 50th percentile, and 1.9 cm longer at the 90th The threshold for an infant classified as overweight
percentile (Figure 2). The 50th percentile length was was generally higher in the WHO-CGS than that
higher among Canadian children than the WHO-CGS among Canadian infants up to 4 months of age

2.00 2.00
All Breas ed Formula-fed Mixed-fed
1.75 1.75
Difference in percen le weight, (kg)

1.50 1.50

1.25 1.25

1.00 1.00

0.75 0.75

0.50 0.50

0.25 0.25

0.00 0.00

– 0.25 – 0.25

– 0.50 – 0.50
0 1week 1 2 4 6 9 12 15 18 24 0 1week 1 2 4 6 9 12 15 18 24 0 1week 1 2 4 6 9 12 15 18 24 0 1week 1 2 4 6 9 12 15 18 24

Age (months) Age (months) Age (months) Age (months)

Number:
Canada 5119 2945 2241 2529 2175 1721 1319 1938 1869 1887 1252 2101 1450 1056 1222 1061 825 618 895 832 784 528 632 347 313 340 313 233 180 262 239 262 139 1281 1030 745 825 683 560 403 584 545 492 299
WHO 890 391 426 423 408 411 392 412 418 442 594 890 391 426 423 408 411 392 412 418 442 594 890 391 426 423 408 411 392 412 418 442 594 890 391 426 423 408 411 392 412 418 442 594
2.00 2.00
All Breas ed Formula-fed Mixed-fed
1.75 1.75
Difference in percen le weight, (kg)

1.50 1.50

1.25 1.25

1.00 1.00

0.75 0.75

0.50 0.50

0.25 0.25

0.00 0.00

– 0.25 – 0.25

– 0.50 – 0.50
0 1week 1 2 4 6 9 12 15 18 24 0 1week 1 2 4 6 9 12 15 18 24 0 1week 1 2 4 6 9 12 15 18 24 0 1week 1 2 4 6 9 12 15 18 24

Age (months) Age (months) Age (months) Age (months)

Number:
Canada 4845 2853 2136 2374 2035 1594 1234 1783 1748 1820 1139 1953 1345 983 1112 949 755 577 784 762 763 401 603 363 285 324 275 219 181 259 254 269 163 1206 1003 745 775 653 501 359 540 494 443 297
WHO 838 396 450 445 444 441 439 451 447 469 596 838 396 450 445 444 441 439 451 447 469 596 838 396 450 445 444 441 439 451 447 469 596 838 396 450 445 444 441 439 451 447 469 596

Figure 1. Difference in weight-for-age between Canadian minus WHO-CGS males (top panel) and females (bottom panel), at the 10th
(red circles), 50th (black squares) and 90th (blue triangles) percentiles. The dashed line at 0 indicates no difference.

© 2017 John Wiley & Sons Ltd

Paediatric and Perinatal Epidemiology, 2017, , –
 Growth of Canadian children vs. WHO-CGS 5

3.0 3.0
All Breas ed Formula-fed Mixed-fed
2.5 2.5
Difference in percen le length, (cm)

2.0 2.0

1.5 1.5

1.0 1.0

0.5 0.5

0.0 0.0

–0.5 – 0.5

–1.0 – 1.0
1 week 1 2 4 6 9 12 15 18 24 1 week 1 2 4 6 9 12 15 18 24 1 week 1 2 4 6 9 12 15 18 24 1 week 1 2 4 6 9 12 15 18 24
Age (months) Age (months) Age (months) Age (months)
Number:
Canada 2249 1769 2374 1975 1636 1073 1875 1441 1472 892 1098 806 1159 971 791 524 877 658 625 375 280 237 309 274 220 135 263 184 194 88 766 612 767 615 521 322 544 409 382 214
WHO 391 427 424 416 419 398 417 421 444 593 391 427 424 416 419 398 417 421 444 593 391 427 424 416 419 398 417 421 444 593 391 427 424 416 419 398 417 421 444 593
3.0 3.0
All Breas ed Formula-fed Mixed-fed
Difference in percen le length, (cm)

2.5 2.5

2.0 2.0

1.5 1.5

1.0 1.0

0.5 0.5

0.0 0.0

– 0.5 – 0.5

– 1.0 – 1.0
1week 1 2 4 6 9 12 15 18 24 1week 1 2 4 6 9 12 15 18 24 1week 1 2 4 6 9 12 15 18 24 1week 1 2 4 6 9 12 15 18 24
Age (months) Age (months) Age (months) Age (months)
Number:
Canada 2143 1700 2275 1874 1544 1048 1760 1348 1414 819 1012 780 1083 877 728 506 773 606 599 282 279 233 294 243 205 148 247 193 210 111 721 577 736 604 486 296 525 368 341 209
WHO 396 450 445 447 444 442 453 448 472 598 396 450 445 447 444 442 453 448 472 598 396 450 445 447 444 442 453 448 472 598 396 450 445 447 444 442 453 448 472 598

Figure 2. Difference in length-for-age between Canadian minus WHO-CGS males (top panel) and females (bottom panel), at the 10th
(red circles), 50th (black squares) and 90th (blue triangles) percentiles. The dashed line at 0 indicates no difference.

(Figure 3). Thereafter, the opposite was observed, a much higher than expected proportion of exclu-
such that by 24 months, the 97th percentile BMI was sively breastfed Canadian children were >90th per-
2.0 kg/m2 higher for Canadian males and 1.1 kg/m2 centile LFA on the WHO-CGS – as much as 30% at
for females, than their WHO-CGS counterparts (Fig- age 6 months (Figure 5). Correspondingly, the pro-
ure 3). Such departures were more pronounced for portion classified <10th percentile LFA on the
formula-fed than breastfed children (Figure 3). WHO-CGS was consistently lower than 10% during
the same age period (Figure 5).

Proportion of breastfed children classified as <10th

or >90th percentile weight or length Proportion of breastfed children classified as wasted
(<3rd percentile BMI), or as overweight (>97th
Among exclusively breastfed Canadian males, about
percentile BMI)
18% (95% CI 16, 19) of newborns, and 26% (95% CI
20, 33) of toddlers aged 2 years, were classified by Before 12 months of age, exclusively breastfed Cana-
WHO-CGS as weighing >90th percentile – much dian males were more likely to be classified by WHO-
higher than the expected rate of 10% (Figure 4). CGS as wasted, with a prevalence of 11% (95% CI 9,
The probability of a Canadian male being <10th 13) at birth and 7% (95% CI 5, 9) at 9 months; whereas,
percentile by WHO-CGS was generally less than from age 12 months onward, they were increasingly
expected. Canadian females followed a similar pat- likely to be classified as overweight, with a prevalence
tern as the males, in terms of being <10th per- of 13% (95% CI 7, 22) by age 24 months (Data S7). A
centile, but less so for weighing >90th percentile by similar pattern was observed for females, though less
age 2 years (Figure 4). From birth to age 18 months, pronounced (Data S7).

© 2017 John Wiley & Sons Ltd

Paediatric and Perinatal Epidemiology, 2017, , –
6 A. L. Park et al.

3.0 3.0
All Breas ed Formula-fed Mixed-fed
Difference in percen le BMI (kg/m2)

2.5 2.5

2.0 2.0

1.5 1.5

1.0 1.0

0.5 0.5

0.0 0.0

– 0.5 – 0.5

– 1.0 – 1.0

– 1.5 – 1.5
1week 1 2 4 6 9 12 15 18 24 1week 1 2 4 6 9 12 15 18 24 1week 1 2 4 6 9 12 15 18 24 1week 1 2 4 6 9 12 15 18 24
Age (months) Age (months) Age (months) Age (months)
Number:
Canada 2162 1739 2309 1933 1464 1044 1655 1378 1415 824 1047 792 1121 950 704 512 774 629 599 359 283 242 315 276 196 135 231 184 193 84 735 595 739 597 473 306 499 386 362 200
WHO 387 426 423 408 411 392 411 416 442 587 387 426 423 408 411 392 411 416 442 587 387 426 423 408 411 392 411 416 442 587 387 426 423 408 411 392 411 416 442 587
3.0 3.0
All Breas ed Formula-fed Mixed-fed
2.5 2.5
Difference in percen le BMI (kg/m2)

2.0 2.0

1.5 1.5

1.0 1.0

0.5 0.5

0.0 0.0

– 0.5 – 0.5

– 1.0 – 1.0

– 1.5 – 1.5
1week 1 2 4 6 9 12 15 18 24 1week 1 2 4 6 9 12 15 18 24 1week 1 2 4 6 9 12 15 18 24 1week 1 2 4 6 9 12 15 18 24
Age (months) Age (months) Age (months) Age (months)
Number:
Canada 2049 1673 2209 1829 1398 1017 1550 1279 1363 747 954 766 1040 855 673 487 686 573 578 270 274 235 295 245 190 151 231 192 204 102 700 566 718 587 431 283 463 345 325 195
WHO 392 449 445 444 441 439 451 442 469 591 392 449 445 444 441 439 451 442 469 591 392 449 445 444 441 439 451 442 469 591 392 449 445 444 441 439 451 442 469 591

Figure 3. Difference in BMI-for-age between Canadian minus WHO-CGS males (top panel) and females (bottom panel), at the 3rd
(orange circles), 50th (black squares) and 97th (green squares) percentiles. The dashed line at 0 indicates no difference.

In a post hoc analysis, we compared the 10th, 50th were most pronounced among children of mothers
and 90th percentile weights and lengths for the Cana- born in Canada, or Europe and Western nations, Mid-
dian cohort, as generated by the quantile regression dle East and North Africa, and Sub-Saharan Africa
method used herein vs. that by the LMS method used and the Caribbean. Infants of mothers from East Asia
by the WHO-MGRS Group.23 No important variations and the Pacific, South Asia, and Latin America
were observed (Data S8a and b). The same was seen appeared to grow more similar to WHO-CGS. By age
comparing the methods for the 3rd, 50th and 97th per- 2 years, there was a 10% excess in the proportion of
centiles (Data S8c). male infants classified as overweight.

Comment Strengths and limitations

This study comprised a sample of nearly 10 000
Main findings
Canadian children who contributed two or more
Healthy Canadian children aged 0–24 months dis- weight measures between birth and age 750 days,
played greater length than the WHO-CGS currently and used over 48 000 weight and 30 000 length
adopted across Canada. Differences in weight, how- measures. Our inclusion criteria were similar those
ever, varied by age, such that Canadian infants used in the WHO study4 and our use of quantile
weighed more than their WHO-CGS counterparts regression generated comparable results to the LMS
both at birth, and from age 6 months onward. Corre- method used by the WHO-MGRS Group.23 As
spondingly, Canadian children displayed compara- EMRALD has a higher proportion of children with
tively lower BMI before 9 months and higher BMI Canadian-born mothers, our sample may have dif-
thereafter. Similar patterns were observed with all fered slightly from all Ontarians, even though we
types of infant feeding. That said, WFA differences performed a stratified analysis by maternal world

© 2017 John Wiley & Sons Ltd

Paediatric and Perinatal Epidemiology, 2017, , –
 Growth of Canadian children vs. WHO-CGS 7

30

25

Propor on (%, 95% CI)

20

15

10

0
0 1 week 1 2 4 6 9 12 14 16 18 24
Age (months)

30

25
Propor on (%, 95% CI)

20

15

Figure 4. Proportion of exclusively

breastfed Canadian male (top) and 10
female (bottom) children who would be
classified as <10th percentile weight-
for-age (red circles), or >90th percentile 5

weight-for-age (blue squares),

according to the WHO-CGS. The 0
dashed line indicates the expected 0 1 week 1 2 4 6 9 12 14 16 18 24
proportion of 10%. Age (months)

region of origin (Data S4a and b). Otherwise, the
newborns in EMRALD were similar to all those in
Ontario. The ethnic ancestry of Canadian-born
women was not available; nonetheless, the majority
were born 25–35 years earlier to parents of British
and European ancestry.24 EMRALD had EMR data
from 321 physicians across Ontario, and anthropo-
metric data routinely collected at well-baby/child
visits. Child weight and length were not obtained
using the strict measurement criteria applied in the
WHO MGRS. However, the potentially greater vari-
ability in growth measurements in EMRALD, par-
ticularly for length, which is more prone to
measurement error in young children, would not
account the consistently greater lengths in Canadian
children compared to the WHO-CGS.
In WHO MGRS, 1743 infants were enrolled at birth,
but only 882 (50%) complied with the criteria for lon-
gitudinal follow-up to age 2 years, including exclu-
sive or predominant breast feeding for at least
4 months of life.4,23 We attempted to follow the
WHO-CGS criteria, and presented separate analyses
by feeding practice. Even then, we could not deter-
mine feeding for 22% of infants. We also lacked data
on maternal smoking, but the rate in Ontario is under
10%.25
Implications
We found important differences in percentiles of
weight, length and BMI of young Canadian children
compared to the WHO-CGS, regardless of infant

© 2017 John Wiley & Sons Ltd

Paediatric and Perinatal Epidemiology, 2017, , –
8 A. L. Park et al.

40

35

30
Propor on (%, 95% CI)

25

20

15

10

0
1week 1 2 4 6 9 12 14 16 18 24
Age (months)

40

35

30
Propor on (%, 95% CI)

25

20

15 Figure 5. Proportion of exclusively

breastfed Canadian male (top) and
10 female (bottom) children that would be
classified as <10th percentile length-for-
5 age (red circles) or >90th percentile
length-for-age (blue squares) according
0 to the WHO-CGS. The dashed line
1week 1 2 4 6 9 12 14 16 18 24
indicates the expected proportion of
Age (months) 10%.

feeding practice. Since the WHO-CGS is currently
used across Canada, these differences could have a
substantial impact on how we interpret the ‘normal’
growth of Canadian children.
Our healthy Canadian cohort was markedly longer
than the WHO-CGS before 18 months, regardless of
feeding practice. In a prior Canadian study of a
selected cohort of 73 infants who were breastfed
according to the WHO guidelines, median newborn
length was markedly greater at birth, but approxi-
mated the WHO thereafter.14 This pattern was also
seen in UK infants.9 In Norway and Belgium, greater
child length in the first 2 years of life was reported,
including among exclusively breastfed Norwegian
children.7,8
Differences in maternal stature are hypothesised to
play a role in birth length.9,26 Among the countries
contributing to the WHO MGRS, the tallest average
maternal height was among Norwegians, and the
shortest was among Indian mothers, as were the cor-
responding birth lengths of their infants.23 While only
about 20% of size at birth is determined by additive
genetic variation, genetics come to explain nearly 80%
of variation in weight, and 40% of variation in height,
from 4 months onward.27,28 Thus, it is plausible that
postnatal growth, especially after 6 months of age,
may be partly expressed by parental ethnicity, which
may explain why certain groups of healthy children
deviate from the standards expressed in the WHO-
CGS.
In contrast to length, the direction and magnitude
of difference in weight and BMI percentiles between
Canadian children and the WHO-CGS depended on
age. Compared to their WHO-CGS counterparts,

© 2017 John Wiley & Sons Ltd

Paediatric and Perinatal Epidemiology, 2017, , –

Canadian children had a higher 50th percentile
weight at birth, and again, after 6 months, while the
50th percentile BMI was lower before 12 months
and higher thereafter. Correspondingly, a higher
than expected proportion of Canadian breastfed chil-
dren were wasted in the first 9 months and over-
weight the second year of life, according to the
WHO-CGS. Age-related differences in weight per-
centiles have been observed in a number of interna-
tional studies comparing national growth references
to the WHO-CGS.7–11 Rolland-Cachera et al.10 com-
pared national weight references from the US, UK,
Netherlands, Belgium and France to the WHO-CGS,
and revealed a strikingly similar pattern of lower
median weights after birth and up to 6 months of
life, with typically higher weights thereafter.
Although the trough between birth and 6 months
observed herein, and in other Canadian studies,12–14
was less pronounced, the consistency of this pattern
of divergence from the WHO-CGS across different
populations, and independent of feeding practice,
raises questions about using the WHO-CGS to moni-
tor the growth of Canadian children, particularly
before 6 months of age.
Some have suggested that the unexpectedly higher
weights of WHO-CGS children aged 0–6 months may
be from selective dropout of smaller infants.29 Around
50% of infants enrolled in the WHO MGRS were non-
compliant with breast feeding, and their growth mea-
sures were excluded. When an infant is smaller than
expected, exclusive breast feeding is more likely to be
replaced with mixed for formula feeding.29,30 In
Malawi, among exclusively breastfed infants under
age 6 months, more infants fell below the threshold
<3rd percentile WFA on the WHO-CGS than on
another growth chart.31 These children were at higher
risk of unnecessary referral to a specialist, with inter-
ruption of exclusive breast feeding, even when there
was no evidence of faltering growth.31 Hence, while
largely adopted for use in many settings, there is some
uncertainty around the use of the WHO-CGS as a
guide to healthy child growth in the first 2 years of
life. Certainly, a similar assessment among Canadian
infants and toddlers is required, including by parental
ethnicity.
We calculated BMI measures for Canadian infants
and toddlers in the first 2 years of life, offering
another useful indicator of early childhood growth.22
The BMI cut-points for classifying a Canadian breast-
fed infant as wasted, or a toddler as overweight, were
© 2017 John Wiley & Sons Ltd
Paediatric and Perinatal Epidemiology, 2017, , –
Growth of Canadian children vs. WHO-CGS 9
discrepant when using the WHO-CGS. Whether such
departures are due to obesogenic feeding practices or
normal healthy growth, needs to be determined.32
Conclusion
Our findings suggest that many healthy Canadian
infants and toddlers weigh more and are longer than
the norms laid out in the WHO-CGS. Specifically, the
degree to which these differences matter for short-
term and long-term health trajectories requires
longitudinal examination before reconsideration of
current use of the WHO-CGS in Canada can be
recommended.
Acknowledgements
The authors thank Dr. Daniel Roth of The Hospital for
Sick Children for his constructive comments over the
course of this project. This work was supported by a
Healthcare Renewal Policy Analysis grant from the
Canadian Institutes of Health Research (CIHR). JGR
holds a Canadian Institutes for Health Research Chair
in Reproductive and Child Health Services and Policy
Research, co-funded by the SickKids Foundation and
CIHR. MU holds a CIHR New Investigator Award.
The funders had no role in study design, data collec-
tion and analysis, decision to publish, or preparation
of the manuscript. This study was supported by the
Institute for Clinical Evaluative Sciences (ICES), which
is funded by an annual grant from the Ontario Min-
istry of Health and Long-Term Care (MOHLTC). The
opinions, results and conclusions reported in this
paper are those of the authors and are independent
from the funding sources. No endorsement by ICES
or the Ontario MOHLTC is intended or should be
inferred. Parts of this material are based on data and
information compiled and provided by the Canadian
Institute for Health Information (CIHI). However, the
analyses, conclusions, opinions and statements
expressed herein are those of the author, and not nec-
essarily those of CIHI. Conflict of Interest: The authors
report no conflict of interest.
References
1 Spencer NJ. Failure to think about failure to thrive. Archives
of Disease in Childhood 2007; 92:95–96.
2 Kramer MS, Barros FC, Demissie K, Liu S, Kiely J, Joseph KS.
Does reducing infant mortality depend on preventing low
10 A. L. Park et al.
birthweight? An analysis of temporal trends in the
Americas. Paediatric and Perinatal Epidemiology 2005;
19:445–451.
3 Kramer MS, Guo T, Platt RW, Sevkovskaya Z, Dzikovich I,
Collet JP, et al. Infant growth and health outcomes associated
with 3 compared with 6 mo of exclusive breastfeeding.
American Journal of Clinical Nutrition 2003; 78:291–295.
4 WHO Multicentre Growth Reference Study Group. WHO
Child Growth Standards based on length/height, weight and
age. Acta Paediatrica 2006; 450:76–85.
5 Collaborative A. Statement from Dietitians of Canada,
Canadian Paediatric Society, College of Family Physicians of
Canada, and Community Health Nurses of Canada.
Promoting optimal monitoring of child growth in Canada:
using the new World Health Organization growth charts.
Paediatrics and Child Health 2010; 15:77–79.
6 Dieticians of Canada. Determining the best use of WHO Growth
Standards and Growth References within the Canadian context.
Technical report and recommendations. Toronto, CA: Dieticians
of Canada, 2008. Accessed June 27, 2016. ProQuest ebrary.
7 J
ulıusson PB, Roelants M, Hoppenbrouwers K, Hauspie R,
Bjerknes R. Growth of Belgian and Norwegian children
compared to the WHO growth standards: prevalence below -
2 and above +2 SD and the effect of breastfeeding. Archives of
Disease in Childhood 2011; 96:916–921.
8 Roelants M, Hauspie R, Hoppenbrouwers K. Breastfeeding,
growth and growth standards: performance of the WHO
growth standards for monitoring growth of Belgian children.
Annals of Human Biology 2010; 37:2–9.
9 Wright C, Lakshman R, Emmett P, Ong KK. Implications of
adopting the WHO 2006 Child Growth Standard in the UK:
two prospective cohort studies. Archives of Disease in
Childhood 2008; 93:566–569.
10 Rolland-Cachera MF, Peneau S. Assessment of growth:
variations according to references and growth parameters
used. American Journal of Clinical Nutrition 2011;
94(6 Suppl):1794S–1798S.
11 Scherdel P, Botton J, Rolland-Cachera MF, Leger J, Pele F,
Ancel PY, et al. Should the WHO growth charts be used in
France? PLoS One 2015; 10:e0120806.
12 Yasin A, Filler G. Evaluating Canadian children: WHO,
NHANES or what? Journal of Paediatrics and Child Health
2013; 49:282–290.
13 Nash A, Secker D, Corey M, Dunn M, O’Connor D. Field
testing of the 2006 World Health Organization growth charts
from birth to 2 years: assessment of hospital undernutrition
and overnutrition rates and the usefulness of BMI. Journal of
Parenteral and Enteral Nutrition 2008; 32:145–153.
14 van Dijk CE, Innis SM. Growth-curve standards and the
assessment of early excess weight gain in infancy. Pediatrics
2009; 123:102–108.
15 Urquia ML, Berger H, Ray JG. Canadian Curves Consortium.
Risk of adverse outcomes among infants of immigrant
women according to birth-weight curves tailored to maternal
world region of origin. Canadian Medical Association Journal
2015; 187:e32–e40.
16 2011 National Household Survey Highlights: Fact Sheet 1.
Ontario: Ministry of Finance, Office of Economic Policy
Labour and Demographic Analysis Branch, 2013. http://
www.fin.gov.on.ca/en/economy/demographics/census/
nhshi11-1.html
17 Ray JG, Sgro M, Mamdani MM, Glazier RH, Bocking A,
Hilliard R, et al. Birth weight curves tailored to maternal
world region. Journal of Obstetrics and Gynaecology Canada
2012; 34:159–171.
18 Tu K, Mitiku TF, Ivers NM, Guo H, Lu H, Jaakkimainen L,
et al. Evaluation of Electronic Medical Record
Administrative data Linked Database (EMRALD). American
Journal of Managed Care 2014; 20:e15–e21.
19 Canadian Task Force on Preventive Health Care.
Recommendations for growth monitoring, and prevention
and management of overweight and obesity in children and
youth in primary care. Canadian Medical Association Journal
2015; 187:411–421.
20 Bartsch E, Park AL, Young J, Ray JG, Tu K. Infant feeding
practices within a large electronic medical record database.
Submitted for publication.
21 Wei Y, Pere A, Koenker R, He X. Quantile regression
methods for reference growth charts. Statistics in Medicine
2006; 25:1369–1382.
22 Furlong KR, Anderson LN, Kang H, Lebovic G, Parkin PC,
Maguire JL, et al. BMI-for-age and weight-for-length in
children 0 to 2 years. Pediatrics 2016; 138:e20153809.
23 WHO Multicentre Growth Reference Study Group. 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. Geneva: World
Health Organization, 2006.
24 Johnston B, Conly J. The changing face of Canadian
immigration: implications for infectious diseases. Canadian
Journal of Infectious Disease and Medical Microbiology 2008;
19:270–272.
25 Brown HK, Wilk P. Changes in smoking during pregnancy
in Ontario, 1995 to 2010: results from the Canadian
community health survey. Journal of Obstetrics and
Gynaecology Canada 2014; 36:878–884.
26 van Buuren S, Van Wouwe JP. WHO child growth
standards in action. Archives of Disease in Childhood 2008;
93:549–555.
27 Dubois L, Ohm Kyvik K, Girard M, Tatone-Tokuda F,
Perusse D, Hjelmborg J, et al. Genetic and environmental
contributions to weight, height, and BMI from birth to
19 years of age: an international study of over 12,000 twin
pairs. PLoS One 2012; 7:e30153.
28 Demerath EW, Choh AC, Czerwinski SA, Lee M, Sun SS,
Chumlea WC, et al. Genetic and environmental influences on
infant weight and weight change: the Fels Longitudinal
Study. American Journal of Human Biology 2007; 19:692–702.
29 van Buuren S. Effects of selective dropout on infant growth
standards. In: Lukas A, Makrides M, Siegler EE, Eds.
Importance of Growth for Health and Development. Nestle Nutr
Inst Workshop Ser Pediatr Program. Vevey/S. Karger AG,
Basel: Nestec Ltd. Vol 65, 2010. pp. 167–179.
30 Kramer MS, Moodie EE, Dahhou M, Platt RW. Breastfeeding
and infant size: evidence of reverse causality. American
Journal of Epidemiology 2011; 173:978–983.
31 Ahmad UN, Yiwombe M, Chisepo P, Cole TJ, Heikens GT,
Kerac M. Interpretation of World Health Organization
© 2017 John Wiley & Sons Ltd
Paediatric and Perinatal Epidemiology, 2017, , –

growth charts for assessing infant malnutrition: a
randomised controlled trial. Journal of Paediatrics and Child
Health 2014; 50:32–39.
32 Salsberry PJ, Reagan PB. Dynamics of early childhood
overweight. Pediatrics 2005; 116:1329–1338.
Appendix 1
Members of the Canadian Curves Consortium
Kathlyn Babaran-Henfrey, Emily Bartsch, Howard
Berger, Douglas Campbell, Maria Chiu, Lisa Colizza,
Leanne R. De Souza, Astrid Guttmann, Manavi
Handa, Christopher Longo, Jonathon Maguire, James
Meloche, Nir Melamed, Patricia Mousmanis, Alison
Park, Joel Ray, Henry Roukema, Jennifer Roy, Michael
Sgro, Graeme N. Smith, Karen Tu, Marcelo Urquia,
Pat Vanderkooy, William Watson, Nancy Watts,
Jacqueline Young.
Supporting Information
Additional Supporting Information may be found in
the online version of this article at the publisher’s
web-site:
Data S1. Flow chart of formation of the cross-sectional
cohorts with measurements of weight (Cohort 1),
length (Cohort 2), or both weight and length concur-
rently (Cohort 3), between birth and age 750 days.
Data S2. Characteristics of the 9964 singleton term
livebirths comprising the current study cohort from
the EMRALD database, compared to all singleton
term hospital livebirths in Ontario between April 1,
2002 and March 31, 2013. Standardised differences
were used to compare differences in means or
proportions.
Data S3. Difference (95% CI) in the 50th percentile
weight-for-age between Canadian minus WHO-CGS
males (top panel) and females (bottom panel). The left
axis shows the 50th percentile weight (kg) in Canada
© 2017 John Wiley & Sons Ltd
Paediatric and Perinatal Epidemiology, 2017, , –
Growth of Canadian children vs. WHO-CGS 11
(green circles) and WHO (blue triangles). The right
axis shows the difference (95% CI) in 50th percentile
weight (g) (black squares). The dashed line at 0 indi-
cates no difference.
Data S4. Difference (95% CI) in 50th percentile
weight-for-age between Canadian minus WHO-CGS
(a) males (b) females, by maternal world region of ori-
gin. The dashed line at 0 indicates no difference.
Data S5. Difference (95% CI) in the 50th percentile
length-for-age between Canadian minus WHO-CGS
males (top panel) and females (bottom panel). The left
axis shows the 50th percentile length (cm) in Canada
(green circles) and WHO (blue triangles). The right
axis shows the difference (95% CI) in 50th percentile
length (cm) (black squares). The dashed line at 0 indi-
cates no difference.
Data S6. Difference (95% CI) in the 50th percentile
BMI-for-age between Canadian minus WHO-CGS
males (top panel) and females (bottom panel). The left
axis shows the 50th percentile BMI (kg/m2) in Canada
(green circles) and WHO (blue triangles). The right
axis shows the difference (95% CI) in 50th percentile
BMI (black squares). The dashed line at 0 indicates no
difference.
Data S7. Proportion of exclusively breastfed Canadian
male (top) and female (bottom) children that would
be classified as wasted <3rd percentile BMI-for-age
(orange circles) or overweight >97th percentile BMI-
for-age (green squares) according to the WHO-CGS.
The dashed line indicates the expected proportion of
3%.
Data S8. (a) Difference in 10th, 50th and 90th per-
centile weights of Canadian children, comparing
Quantile Regression (QR) vs. Lamba-Mu-Sigma
(LMS) methods. (b) Difference in 10th, 50th and 90th
percentile lengths of Canadian children, comparing
Quantile Regression (QR) vs. Lamba-Mu-Sigma
(LMS) methods. (c) Difference in 3rd, 50th and 97th
percentile body mass index (BMI) of Canadian chil-
dren, comparing Quantile Regression (QR) vs.
Lamba-Mu-Sigma (LMS) methods.