Professional Documents
Culture Documents
related to this diagnosis between 1997 and 2010. The BMI and the relationship between the duration of follow-
medical records and radiographs of the identied up with the change in BMI over time. Multiple regression
patients were reviewed retrospectively. Demographic analysis was performed to predict the eect of the childs
information including the patients sex, ethnicity, age at initial surgery, number of extremity surgeries per-
health insurance type (Medicaid vs. commercial), age of formed to address the limb deformity, unilateral versus
disease onset (early onset vs. late onset), laterality (uni- bilateral involvement and the length of follow-up on the
lateral vs. bilateral), type(s) of surgery (osteotomy, hem- change in BMI values over time. Statistical signicance
iepiphyseodesis, contralateral epiphyseodesis, hardware was dened as P < 0.05.
removal), and chronologic age at rst surgery were ex-
tracted from the medical records. The patients height and
weight at each clinical encounter were retrieved from the
outpatient and/or inpatient records. If this information RESULTS
was not available, the childs primary care physician was During the study period, 68 patients with the diag-
contacted to get the pertinent demographic data. The nosis of Blount disease underwent surgical treatment. Of
BMI was calculated by dividing the patients weight in these 17 patients were excluded because of the following
kilograms by the square of the patients height in meters reasons: above 18 years of age at initial surgery (4 pa-
(kg/m2).3 tients) and unavailable preoperative and/or postoperative
Radiographic analysis included measurement of the height and weight measurements (13 patients). Thus, 51
mechanical axis deviation (MAD) and LLD based on patients with 70 aected extremities were included in the
the standing full-length anteroposterior radiographs of nal analysis. The demographic details including the
the lower extremity that was performed before the rst childs age at disease onset, laterality, sex, race, health
surgery and at latest follow-up. MAD was calculated in insurance status, and type of initial surgery is noted
millimeters as the perpendicular distance from the center in Table 1. Eighty-six percent (44/51) of patients under-
of the distal femoral condyles to the mechanical axis line went >1 surgical procedure, including hardware removal.
connecting the center of the femoral head to the center of All 47 patients who had gradual correction with external
the talar dome of the aected extremity.7 In patients with xation as the initial or subsequent surgery also under-
bilateral disease, the larger of the 2 MADs was used for went nutritional counseling while undergoing inpatient
statistical analysis. LLD was calculated in millimeters as rehabilitation for 2 to 6 weeks postoperatively.
the distance between the transverse lines drawn along the Details of the patients preoperative and most recent
most proximal aspect of the 2 femoral heads on the BMI and weight in kilograms are noted in Table 2. At an
standing radiograph. If the radiograph was performed average follow-up of 48 months [95% condence interval
with the patient standing on a lift under the short side to (CI) 39-56 mo], the mean BMI for the group had in-
level the pelvis, the height of this lift was added to the creased to 38 (95% CI, 35-41; P = 0.0006). During this
calculation. time period, the MAD improved from 80.5 mm medial
Patients were excluded if they were older than 18 (95% CI, 72-89 mm medial) to 16.1 mm medial (95% CI,
years at the time of their initial surgery or if their height 6-26.2 mm medial; P < 0.0001) and the limb length
and weight or full-length standing radiographs of their discrepancy improved from 19.6 mm (95% CI, 15.4-
lower extremity were not available preoperatively and 23.8 mm) to 10.9 mm (95% CI, 7.5-14.4 mm; P < 0.0002).
postoperatively. Compared with their preoperative BMI, 76% (39/51) of
Data were analyzed using the statistical package the patients had an increase in BMI at the last follow-
SAS, version 9.2 (SAS Institute Inc., Cary, NC). The up. Preoperatively, 13 patients were morbidly obese
mean values of each patients BMI, MAD, and LLD (BMIZ40), whereas at the latest follow-up, 20 patients
preoperatively and at the most recent follow-up were had a BMIZ40 (P = 0.002). The patient with the largest
compared using paired t test. Fisher exact test was utilized decrease in BMI, from 44 preoperatively to 28 (D16) at 43
to compare the proportion of children who were morbidly months postoperatively was a 16-year 10-month-old fe-
obese (BMIZ40) preoperatively and at the most recent male with early-onset Blount disease who had undergone
follow-up. The eect of variables such as the patients age multiple prior surgeries, including an attempt at medial
at disease onset (early onset vs. late onset), laterality, sex, tibial plateau elevation with internal xation at another
ethnicity, health insurance status, and MAD and LLD at institution (Fig. 1). She presented with left-sided knee
the most recent follow-up on the change in BMI between pain, genu varum, and limb shortening. Radiographs re-
the preoperative and most recent follow-up was com- vealed a residual MAD of 71 mm medial and 52 mm of
pared using unpaired t test. Pearson correlation co- LLD. She underwent removal of hardware, in prepara-
ecient was used to assess the relationship of variables tion for a repeat osteotomy and gradual correction, but
such as the childs age at initial surgery, number of ex- refused further orthopaedic procedures. She is the only
tremity surgeries performed to address each sided limb subject in our series who underwent subsequent bariatric
deformity and the length of follow-up on the change in surgery (gastric bypass procedure), and despite having the
BMI values preoperatively and at the most recent follow- largest amount of residual limb deformity and LLD at
up. Simple linear regression analysis was performed to follow-up, lost the most weight (preoperative 130 kg,
determine the eect of patients initial BMI on their nal postoperative 76.4 kg; D53.6 kg).
TABLE 1. Demographic Information of Patients With Blount Disease, Including Early-onset and Late-onset Subgroups
Early Onset, n (%) Late Onset, n (%) Total
No. Patients 23 (45) 28 (55) 51
Side aected
Right 4 (17) 7 (25) 11
Left 9 (39) 12 (43) 21
Bilateral 10 (43) 9 (32) 19
Sex
Male 9 (39) 23 (82) 32
Female 14 (61) 5 (18) 19
Race
Black 12 (52) 25 (89) 37
Others 11(48) 3 (11) 14
Insurance
Commercial 8 (35) 13 (46) 21
Medicaid 15 (65) 15 (54) 30
Type of initial surgery
Osteotomy 18 (78) 21 (75) 39
Osteotomy+hemiepiphyseodesis 0 (0) 4 (14) 4
Osteotomy+epiphyseodesis 1 (4) 1 (3.5) 2
Hemiepiphyseodesis 3 (13) 1 (3.5) 4
Epiphyseodesis 0 (0) 1 (3.5) 1
Hardware removal 1 (4) 0 (0) 1
Mean age (y) at initial surgery (95% CI) 7.1 (5.3-8.9) 12.8 (12.0-13.5) 10.2 (9.0-11.4)
Mean age (y) at initial BMI (95% CI) 6.8 (4.9-8.7) 12.6 (11.9-13.4) 9.9 (8.7-11.2)
BMI indicates body mass index; CI, condence interval.
Using bivariate analysis, there was no eect of the postoperative BMI (Fig. 1). In addition, as the length of
patients chronologic age at disease onset (P = 0.72), follow-up increased, there was a tendency for the patients
laterality (P = 0.09), sex (P = 0.11), ethnicity (P = 0.10), BMI to rise (P = 0.002; Fig. 2). On the basis of multi-
health insurance status (P = 0.93), change in MAD variate analysis, the length of follow-up was the only
(P = 0.09), or change in limb length discrepancy variable that was associated with a signicant increase in
(P = 0.43) on the change in BMI over time (Table 3). BMI in the postoperative period (P = 0.03).
Although there was no signicant association of the pa- In order to minimize bias, the data were reanalyzed
tients age at the initial surgery (r = 0.23, P = 0.11), after excluding the 1 patient (noted above) who only
number of surgeries (r = 0.14, P = 0.33), or unilateral underwent hardware removal and refused further re-
versus bilateral involvement (r = 0.28, P = 0.05), there alignment surgery. Although there were minor changes in
was a positive correlation between the length of follow-up the P-values based on multivariate analysis, only the
and the increase in BMI between the initial and latest length of follow-up remained signicantly associated with
follow-up (r = 0.42, P = 0.002). Furthermore, using an increase in the patients BMI in the postoperative pe-
simple linear regression, there was a signicant relation- riod (P = 0.0008).
ship (P < 0.0001) between the patients preoperative and
TABLE 3. The Effect of Certain Variables on the Change in BMI Between the Initial Evaluation and Most Recent Follow-up, Based
on Bivariate Analysis Using Unpaired t test
Mean (95% CI)
Variables Initial BMI Latest Follow-up BMI Change in BMI P Power
Diagnosis
Early onset (n = 23) 29.8 (25.9-33.8) 33.7 (29.3-38.0) 3.9 (0.3-7.4) 0.72 0.1
Late onset (n = 28) 38.8 (36.0-41.5) 42.0 (38.1-45.8) 3.2 (1.1-5.2)
Deformity side
Unilateral (n = 32) 35.1 (31.6-38.6) 37.4 (33.1-41.6) 2.3 (0.1-4.4) 0.09 0.7
Bilateral (n = 19) 34.2 (30.2-38.2) 39.7 (35.6-43.8) 5.5 (1.8-9.2)
Sex
Male (n = 32) 36.3 (33.2-39.5) 38.5 (34.4-42.6) 2.2 (0.2-4.1) 0.11 0.8
Female (n = 19) 32.1 (27.6-36.6) 37.7 (32.9-42.6) 5.6 (1.7-9.7)
Ethnicity
Black (n = 37) 35.1 (32.6-37.6) 37.6 (34.6-40.7) 2.5 (0.4-4.7) 0.10 0.7
Others (n = 14) 33.8 (26.3-41.2) 39.8 (31.4-48.1) 6.0 (1.9-10.0)
Insurance
Commercial (n = 21) 34.7 (31.4-37.9) 38.3 (34.8-41.7) 3.6 (0.8-6.3) 0.93 0.1
Medicaid (n = 30) 34.8 (30.9-38.7) 38.2 (33.5-42.9) 3.4 (0.7-6.1)
MAD at latest follow-up*
Normal (n = 8) 31.6 (21.1-42.2) 31.9 (22.9-40.1) 0.3 (-3.3-3.9) 0.09 0.9
Abnormal (n = 39) 35.8 (32.9-38.4) 40.3 (36.9-43.6) 4.5 (2.3-6.9)
LLD at latest follow-up (mm)
r10 (n = 24) 33.9 (30.2-37.6) 37.1 (33.1-41.1) 3.2 (0.4-6.0) 0.43 0.1
> 10 (n = 19) 37.2 (32.2-42.1) 41.9 (36.1-47.8) 4.7 (1.8-7.7)
*MAD normal range, 3 to 17 mm medial; MAD abnormal, MAD <3 or >17 mm medial.
BMI indicates body mass index; CI, condence interval; LLD, leg length discrepancy; MAD, mechanical axis deviation.
lose weight. Interestingly, the patient who lost the most rehabilitation, given the retrospective study design, we
weight had the largest amount of residual varus deformity were unable to fully ascertain the eect of such an inter-
and LLD, and despite having multiple surgeries before vention. Yet, on the basis of the increase in BMI in the
presentation, merely underwent hardware removal at our majority of patients over time, it is unlikely that such a
center. She was also the only patient in our study group short period of counseling was eective in having any
who had undergone a bariatric procedure. Although sustainable improvement in the childrens lifestyle or their
evaluating the eectiveness of bariatric surgery was not obesity. Fifth, we were unable to include 13 patients due
the focus of our study, such procedures could lead to to lack of availability of their preoperative or recent
sustained weight loss in certain obese adolescents.14 height and weight measurements. This may be a potential
Several investigators have reported on the inability source of bias. Even if we assumed that all of those 13
of the majority of adult obese patients with various children did not increase their BMI over time, still 61%
musculoskeletal ailments to lose weight despite under- (39/64) of the patients would have had a greater BMI
going elective orthopaedic procedures such as arthro- postoperatively compared with their preoperative meas-
plasty of the hip1519 and knee2022 joints as well as urements. Lastly, our study was underpowered to con-
successful ankle reconstruction23 and lumbar spinal de- dently establish the lack of association of certain
compression for neurogenic claudication.24 Our results variables such as the childs age at disease onset, health
demonstrate a similar pattern of weight gain among insurance status, and LLD with the change in BMI after
children with Blount disease. We found that the majority limb realignment. Given the possibility of a type 2 error, a
of children with Blount disease not only did not lose larger cohort of patients in each subgroup would be
weight after lower limb alignment surgery, but also that necessary to conrm our ndings.
>75% of these patients had an increase in their BMI at In conclusion, despite an improvement in lower
an average postoperative follow-up of 2 years. Factors limb alignment after surgical intervention and a trial of
other than the magnitude of limb deformity were likely up to 6 weeks of nutritional counseling during inpatient
related to their obesity, and most probably these issues rehabilitation in the majority of children with Blount
remained uncorrected despite successfully addressing disease, their BMI increased over time. These ndings
their genu varum and limb shortening. Thus, anticipating should be incorporated in the preoperative discussion
weight loss after satisfactory lower limb realignment in with the overweight childs care takers before undergoing
these patients seems unrealistic. surgical intervention, so that they may have realistic ex-
There are a few limitations in our study. First, being pectations after surgery. Furthermore, other strategies for
a retrospective case series, we were unable to follow these weight loss should be investigated in obese children with
childrens BMI prospectively at specic time intervals. Blount disease.
However, we were able to compare their BMI at 2 distinct
time points; preoperatively and at most recent follow-up.
Although our mean follow-up was only 48 months, we REFERENCES
were able to establish a trend of increasing BMI with 1. Sabharwal S. Blount disease. J Bone Joint Surg Am. 2009;91:
time. Whether this trend will continue over the patients 17581776.
adult lifetime would require a longer follow-up study. 2. Ogden C, Carroll M. Prevalence of obesity among children and
Second, we were unable to assess the inuence of factors adolescents: United States, trends 1963-1965 through 2007-2008,
2011. Available at: http://www.cdc.gov/nchs/data/hestat/obesity_
such as the childrens motivation to lose weight and their child_07_08/obesity_child_07_08.pdf. Accessed July 8, 2013.
ability to perform physical exercise related to knee pain, 3. WHO. Obesity and overweight. Available at: http://www.who.int/
limb deformity, or socio-economic constraints on the mediacentre/factsheets/fs311/en/. Accessed July 8, 2013.
change in BMI noted postoperatively. The role of these 4. Li Z, Bowerman S, Heber D. Health ramifications of the obesity
epidemic. Surg Clin North Am. 2005;85:681701.
potentially important factors could be assessed with a
5. Lobstein T, Baur L, Uauy R. TaskForce IIO. Obesity in children
prospective study design in the future. Third, we did not and young people: a crisis in public health. Obes Rev. 2004;5(suppl 1):
have a control group of patients with Blount disease who 4104.
did not have lower limb surgery. It is plausible that if we 6. Vander Wal JS, Mitchell ER. Psychological complications of
had a nonoperative cohort available, that such patients pediatric obesity. Pediatr Clin North Am. 2011;58:13931401.
7. Paley D. Principle of Deformity Correction. Berlin, Heidelberg:
may have demonstrated as much or greater gain in BMI Springer-Verlag; 2002;10.
over time. However, given the eectiveness of improving 8. Wills M. Orthopedic complications of childhood obesity. Pediatr
limb malalignment using various surgical techniques, Phys Ther. 2004;16:230235.
having such a control group would not be feasible. Fur- 9. Chan G, Chen CT. Musculoskeletal effects of obesity. Curr Opin
thermore, our purpose was merely to see whether there Pediatr. 2009;21:6570.
10. Sabharwal S, Root MZ. Impact of obesity on orthopaedics. J Bone
was any positive association between lower limb realign- Joint Surg Am. 2012;94:10451052.
ment and postoperative weight loss. We found that on the 11. Sabharwal S, Zhao C, McClemens E. Correlation of body mass
contrary, more than 75% of these children continued to index and radiographic deformities in children with Blount disease.
increase their BMI, despite improvement in their genu J Bone Joint Surg Am. 2007;89:12751283.
12. Sabharwal S, Wenokor C, Mehta A, et al. Intra-articular
varum. Fourth, although most of these children did un- morphology of the knee joint in children with Blount disease: a
dergo a 2 to 6 week period of nutritional counseling in the case-control study using MRI. J Bone Joint Surg Am. 2012;94:
early postoperative period while undergoing inpatient 883890.
13. Pirpiris M, Jackson KR, Farng E, et al. Body mass index and Blount 20. Zeni JA Jr, Snyder-Mackler L. Most patients gain weight in the 2
disease. J Pediatr Orthop. 2006;26:659663. years after total knee arthroplasty: comparison to a healthy
14. Barnett SJ. Contemporary surgical management of the obese control group. Osteoarthritis Cartilage. 2010;18:510514.
adolescent. Curr Opin Pediatr. 2011;23:351355. 21. Dowsey MM, Liew D, Stoney JD, et al. The impact of pre-operative
15. Jain SA, Roach RT, Travlos J. Changes in body mass obesity on weight change and outcome in total knee replacement: a
index following primary elective total hip arthroplasty. Correla- prospective study of 529 consecutive patients. J Bone Joint Surg Br.
tion with outcome at 2 years. Acta Orthop Belg. 2003;69: 2010;92:513520. [Erratum appears in J Bone Joint Surg Br. 2010
421425. Jun;92(6):902].
16. Woodruff MJ, Stone MH. Comparison of weight changes after total 22. Lachiewicz AM, Lachiewicz PF. Weight and activity change in
hip or knee arthroplasty. J Arthroplasty. 2001;16:2224. overweight and obese patients after primary total knee arthroplasty.
17. Aderinto J, Brenkel IJ, Chan P. Weight change following total hip J Arthroplasty. 2008;23:3340.
replacement: a comparison of obese and non-obese patients. 23. Penner MJ, Pakzad H, Younger A, et al. Mean BMI of overweight
Surgeon. 2005;3:269272. and obese patients does not decrease after successful ankle
18. Woolf VJ, Charnley GJ, Goddard NJ. Weight changes after total reconstruction. J Bone Joint Surg Am. 2012;94:17.
hip arthroplasty. J Arthroplasty. 1994;9:389391. 24. Garcia RM, Messerschmitt PJ, Furey CG, et al. Weight loss in
19. Middleton FR, Boardman DR. Total hip arthroplasty does not aid overweight and obese patients following successful lumbar decom-
weight loss. Ann R Coll Surg Engl. 2007;89:288291. pression. J Bone Joint Surg Am. 2008;90:742747.