Serial Changes in Cranial Asymmetry During Helmet Therapy for Deformational Plagiocephaly

Abstract

Background:

To analyze the serial changes in cranial asymmetry (CA) during helmet therapy for deformational plagiocephaly (DP).

Methods:

The subjects were 159 patients with DP who visited National Center for Child Health and Development between October 2011 and March 2014 and completed helmet therapy. The authors retrospectively collected information from medical records and analyzed the rate of improvement of deformation.

Results:

The average age at the start of the helmet was 24.1 weeks old [standard deviation (SD): 5.0], and the average treatment period was 21.2 weeks (SD: 5.3). The average CA before the start of helmet therapy was 16.3 mm (SD: 4.2), and it improved to 7.7 mm (SD: 3.5) after therapy. The analysis in the present article was limited to 16 weeks after initiation of helmet therapy when 1/4 of patients completed the therapy. In general, improvement of CA occurred inverse proportionally to the enlargement of head circumference. In subgroup analysis, for those who started earlier than 24 weeks of age or those with the severity of Argenta 1 or 2, improvement occurred rapidly and reached nearly a plateau at 12 weeks of treatment. For those started later or those with more severe deformity, improvement occurred slower and it continued even after 12 weeks of treatment.

Conclusion:

The rate of improvement was rapid at the early stage of treatment for those started earlier or with milder severity. This rapid improvement was obtained when the cranial enlargement was rapid. By understanding the course of improvement, evaluation of treatment responsiveness and ideal timing to finish therapy is possible.

Deformational plagiocephaly (DP), also known as positional plagiocephaly or nonsynostotic plagiocephaly, is defined as a flattening on one side of the cranium due to persistent external pressure.¹ And deformational brachycephaly (DB) is defined as flattening of both sides of the posterior skull. Risk factors for DP include multiple gestation, male sex, premature delivery, assisted delivery, and the presence of congenital muscular torticollis.^2–4 Flattening of the cranium is thought to be caused by head preference, in which the head is constantly turned to one side, and the head growth is suppressed in the area due to constant pressure from the weight of the head. In severe cases, once the flattening has progressed to some extent, the head preferentially turns to rest on the flattened area and makes the deformation fixed or even more severe.

In 1992, the American Academy of Pediatrics proposed the “Back to Sleep Campaign” to encourage carers to lay their babies on their backs with the aim of preventing sudden infant death syndrome.⁵ With the spread of this recommendation, supine position became the mainstream instead of prone position, and the prevalence of DP increased from 0.3%⁶ to up to 46.6%.³

Treatment options include head-repositioning therapy, physical therapy, and helmet therapy.⁷ The Guideline from Congress of Neurological Surgeons recommends that patients with DP are initially introduced with physical therapy, and for moderate to severe deformity, helmet therapy is applied.⁸ There is considerable evidence of the efficacy of helmet therapy.^7,9

In Japan, the accurate prevalence of DP is unknown but is reported to be around 65%,¹⁰ which is much higher than the US and European countries because Japanese infants traditionally spend a long time in the supine position. In recent years, parents have become more concerned about head deformities, and the number of head shape consultations is increasing rapidly. However, helmet therapy is still not widely performed compared with other countries, and there are few reports to date.^11–13

In 2011, National Center for Child Health and Development, located in Tokyo, Japan, opened the “Clinic for Baby’s Head Shape”, and has been providing helmet therapy for DP. From past analysis, we reported that statistically significant improvement was observed in cranial asymmetry after helmet therapy. We also reported that cranial asymmetry after helmet therapy was statistically larger in the group who started helmet therapy at older than 6 months or whose cranial asymmetry before treatment was more severe.¹³

From the experience of helmet therapy for DP, we often experience a pattern that the rate of improvement is rapid at the early stage of treatment especially at first 1 to 2 months and gradually becomes slow, but this tendency has not been confirmed with scientific evidence.

The purpose of this study is to analyze the serial changes in cranial asymmetry during helmet therapy.

MATERIALS AND METHODS

In this study, we retrospectively collected information on helmet therapy for DP from medical records and analyzed the rate of improvement of deformation. This study was approved by the institutional review board at National Center for Child Health and Development (No. 2021-030).

The subjects were 159 patients who visited the Clinic for Baby’s Head Shape of National Center for Child Health and Development between October 2011 and March 2014 and completed helmet therapy using the protocol described below (Fig. 1). Briefly, 178 patients started helmet therapy, and 12 patients did not complete the therapy due to lost to follow-up (n=9), death due to heart disease (n=1), or parents chose to quit because of profound sweating (n=1) or reinstated to employment (n=1). Out of 166 patients who completed helmet therapy, 7 were excluded from analysis because of pure deformational brachycephaly without plagiocephalic deformity (n=3), data deficit (n=2), scan data acquired from computed tomography (CT) and not from scanner (n=1), or cost of helmet not covered by research fund (n=1).

FIGURE 1.

Patients.

All patients underwent clinical evaluation according to the Argenta classification¹⁴ and anthropometric measurements. Head circumference, cranial asymmetry (CA), and cephalic index (CI) were measured using a craniometer and a tape measure according to the description of Loveday et al¹⁵ CA is an index of asymmetry of the cranial vault, and defined as the absolute value of the difference in the length of right-anterior and left-anterior diagonals 30 degrees from the midline. A larger value in CA means more severe head asymmetry. We defined CA <5 mm as normal, 5 to 9 mm as mild, 10 to 14 mm as moderate, 15 to 19 mm as severe, and 20 mm or more as most severe.¹³ CI is an index of brachycephaly and dolichocephaly and defined as cranial width divided by cranial length multiplied by 100. On the basis of the measurement results of healthy Japanese infants, CI <79.2 is defined as dolichocephaly and >93.8 is defined as brachycephaly¹⁶ (Fig. 2A). X-ray of the skull was taken if craniosynostosis could not be ruled out with physical examination.

FIGURE 2.

(A) Evaluation of head deformity. Above left: measurement of CA, the index for cranial asymmetry. Above right: measurement of CI, the index of dolicho/brachycephaly. Middle: Argenta classification. Type 1: unilateral flattening of the occiput. Type 2: ipsilateral ear shifts anteriorly. Type 3: ipsilateral brow protrusion, Type 4: ipsilateral cheek protrusion, Type 5: temporal protrusion. (B) Michigan Cranial Reshaping Orthosis. CA indicates cranial asymmetry; CI, cephalic index.

Patients with DP who were younger than 4 months were initially introduced with physical therapy and were evaluated again at 4 months old. Patients who were equal to or older than 4 months and whose head deformity were equal to or more severe than Argenta type 2 were introduced with helmet therapy. Helmet therapy was initiated if the parents of the baby desired the treatment. We used a Michigan Cranial Reshaping Orthosis from Danmer Products (Fig. 2B). An infant’s head was scanned using an optical scanner (OMEGA Scanner, Ohio Willow Wood Company), and obtained data was modified with specialized software (OMEGA Tracer, Ohio Willow Wood Company) to create a helmet data. The helmet was manufactured based on this data and sent to Japan. Patients wore the helmets 23 hours a day after a 1 to 2 weeks of break-in period. The helmet was adjusted every 3 to 4 weeks. Argenta classification, head circumference, CA and CI were measured at every visit. Helmet therapy was terminated when the helmet became tight or when the family was satisfied with the head shape. At the end of the treatment, infant’s head was scanned again.

The following data were collected retrospectively from the medical record: age at which helmet therapy was started, Argenta classification and CI at the start and end of the treatment, and CA and head circumference at every follow-up time during the treatment period. Records of helmet wearing-time was collected from the parents at every visit.

For the patients in the present study, part of the cost of helmet therapy was provided by the institutional research fund (institutional fund; grant number 21-14) for the purpose of evaluating safety and efficacy of the treatment.

Statistical Methods

First, we compared the Argenta classification, CI, and CA before and after treatment using paired t test.

Next, we constructed the following mathematical model to examine serial changes in head circumference and CA during helmet therapy (model 1, Fig. 3). Similarly, we stratified and analyzed serial changes of CA by Argenta classification (model 2, Fig. 4) and by age at the start of helmet therapy (model 3, Fig. 5).

FIGURE 3.

Serial change in CA and head circumference. Line: average. Band: 95% CI. Note that the rapid improvement in CA is gained during the time when head circumference is rapidly expanding. CA indicates cranial asymmetry.

FIGURE 4.

Serial change in CA according to different Argenta types. Line: average. Error bar: 95% CI. Arrows: time point at which initially rapid rate of improvement becomes relatively slow. Note that this time point is early for the patients with milder deformity, and later for more severe cases. CA indicates cranial asymmetry.

FIGURE 5.

Serial change in CA according to different age at start of helmet therapy. Line: average. Band: 95% CI. Note that the bands do not overlap for the group of patients who started helmet therapy later than 23 weeks and those started earlier for most of the time course, suggesting significant difference in the improvement rate. CA indicates cranial asymmetry.

We dealt with the data up to 15 weeks of helmet therapy, because after this time point, patients with rapid improvement start quitting the therapy and the rate of missing data became >25%.

In model 1 (Fig. 3), we used a linear mixed-effects model to analyze the serial changes of CA and the head circumference. The fixed effects were age, Argenta classification before helmet therapy (pre-Argenta), time (as a continuous variable) of a second-order polynomial, and the interaction terms of the time of a second-order polynomial and pre-Argenta. We also included the random intercept and slope in subject level in the model. We calculated the least-squares mean estimates and the 95% confidence intervals of the CA and the head circumference to confirm the longitudinal change.

In model 2 (Fig. 4), we used a linear mixed-effects model to analyze the changes in the CA by pre-Argenta at each measurement week. The fixed effects were age, pre-Argenta, time (as a discrete variable), and the interaction term of time and pre-Argenta, assuming an unstructured covariance matrix as a repeated-measures covariance structure. We calculated the least-squares mean estimates and the 95% CIs of the CA at each time.

In model 3 (Fig. 5), we used a linear mixed-effects model to analyze the serial changes of CA by the age category (wk: 20<, 20–23, 24≥) at the start of the helmet therapy. The fixed effects were age category, time (as a continuous variable) of a second-order polynomial, and the interaction terms of the time of a second-order polynomial and age category. We also included the random intercept and slope in subject level in the model. We calculated the least-squares mean estimates and the 95% CIs of the CA to confirm the longitudinal change in each age category.

RESULTS

Three hundred eighty-seven infants visited the Clinic for Baby’s Head Shape between October 2011 and March 2014, and 366 infants were diagnosed with DP/ DB.

Of the 21 excluded infants, 6 infants were normal skull without DP (CA <5 mm), 10 infants had suspected craniosynostosis on x-ray or CT, and 5 infants presented head deformity due to other causes (such as brachycephaly due to Down syndrome). Among the 366 patients with DP/DB, 86 patients who were younger than 4 months were initially introduced physical therapy. They were evaluated again at 4-month old, and 18 patients were excluded due to improvement. Among 348 Patients who were older than 4 months old, including the 68 patients after physical therapy, 252 patients were more severe than mild deformity (CA>9). Among them, 178 patients were introduced with helmet therapy and 166 patients completed. 9 patients did not complete helmet therapy because of lost to follow-up (9 patients), death due to heart disease (1 patient), or the parents chose to quit the therapy because of profound sweating and crying at night (1 patient) or because parents were reinstated in employment (1 patient). Finally, we analyzed 159 patients excluding 7 patients (3 patients with pure DB without plagiocephalic deformity, 2 patients whose data were missing, 1 patient whose scan data was acquired from CT scan and not from an optical scanner, 1 patient whose cost was not covered by research fund) (Fig. 1).

Supplemental Table 1 (Supplemental Digital Content 1, http://links.lww.com/SCS/H275) shows the demographics of 159 patients. The average age at the start of the helmet was 24.1 weeks old [standard deviation (SD): 5.0] and the average treatment period was 21.2 weeks (SD: 5.3).

The Average CA before the start of helmet therapy was 16.3 mm (SD: 4.2), and it improved to 7.7 mm (SD: 3.5) after therapy. A statistically significant improvement was observed not only in CA (P<0.001, paired t test), but also in CI (P<0.001, paired t test). (Supplemental Table 2, Supplemental Digital Content 2, http://links.lww.com/SCS/H276)

Figure 3 shows the serial changes in CA and the head circumference during helmet therapy. The bands show the 95% CIs. The rate of improvement of CA was rapid at the early stage of treatment. In addition, the head circumference enlarged rapidly in the early stage during which rapid improvement of CA was obtained. In average, the improvement in CA gained in the first 6 weeks [5.3 mm (SD:3.5)] reached more than 60% of the improvement gained through the total duration of helmet therapy [8.6 mm (SD:3.9)].

For those started earlier than 24 weeks of age or those with severity of Argenta 1 or 2, improvement in CA occurs rapidly and reaches nearly plateau at 12 weeks of treatment. For those started later or those with more severe deformity (Argenta 3 or more), improvement in CA occurs slower and it continues even after 12 weeks of treatment (Figs. 4 and 5).

DISCUSSION

There are many studies reporting the efficacy of helmet therapy for DP. However, most of these studies report the pre-treatment and post-treatment results, and the rate of improvement during the period of helmet therapy has not been matter of debate. Robinson et al described that infants typically show significant improvement after the initial several weeks in the helmet therapy, and achieve most of the correction by 3 months, but the serial changes in CA was not presented.¹⁷

Our study showed that the rate of improvement of CA was rapid at the early stage of treatment and gradually slows down. This tendency was more evident in the group of patients who started helmet therapy at a younger age and whose deformity was milder.

The improvement in CA and cranial growth were approximately inversely proportional.

The child’s head grows rapidly up to ~81% of the size of adults by 1 year old.¹⁸ The principle of the Michigan Cranial Remolding Helmet is to promote the growth of the skull in the flattened area by completely decompressing the area without applying excessive pressure to the overgrown part. The results of this study, the high improvement rate of CA during the period of rapid skull growth, supported this principle of helmet therapy. This result also complies with our experience and shows that helmet therapy is most effective during period of rapid skull growth.

The change of improvement rate occurred at an earlier time in patients with milder deformities compared with more severe patients. We propose the following hypotheses to explain this observation. In Argenta type 1, which is a deformity limited to the cranium, the deformity is rapidly resolved with expansion of the head by decompression of the flattened area with the helmet therapy. In contrast, in type 2 and above, which is accompanied by deformation of the cranial base (i.e., ear position and brow area), cranial enlargement is not sufficient for the improvement of deformity in the cranial vault, and the growth of cranial base has to take place, which occurs slowly compared with the growth of the cranium.¹⁹

In addition, we showed a difference in the rate of improvement depending on the age at the start of helmet therapy. The patients who were younger than 6 months at the start of helmet therapy achieved statistically greater improvement at the early stage compared with the group older than 6 months. This is probably due to the fact that the head growth is especially rapid up to 6 months of age. There is no complete consensus on when to start the helmets, but many authors reported that younger than 6 months is the optimal age.^12,20–25 Our results support these opinions by adding evidence of serial improvement during helmet therapy. We conclude that the most rapid improvement can be obtained by starting the helmet at younger than 6 months when the cranial enlargement rate is rapid.

By clarifying the progress of improvement in head deformity, we suggest the following. First, this finding can be used for the evaluation of treatment responsiveness in helmet therapy. The improvement of CA during the helmet therapy of DP is rapid at the early period of therapy regardless of the severity and age at the start of helmet therapy, although the rate of improvement is slower for those with severe deformity (Argenta type 3 and above) or whose helmet therapy were started at older than 6 months of age. The improvement rate becomes slower as the head enlargement becomes gradual. In the case of poor response at the early period of treatment, re-evaluation of treatment compliance and adherence (e.g., is helmet worn 23 h a day?), helmet adjustment (e.g., is helmet fitted properly?), or disease conditions (e.g., craniosynostosis, brain maldevelopment) are warranted.

Secondly, by understanding this improvement pattern, it is possible to set the end time of treatment. By avoiding the prolonged period of helmet therapy during the late phase when the therapeutic effect is low, it is possible to reduce the treatment burden of the patient and guardians.

The number of cases is too small to generate a reliable prediction model of CA improvement at present, but accumulating experience with more patients enables constructing such model in future. For now, Figure 3, which is the average of all cases with different severity and starting timing can be used as a general guide for the transition of CA. In actual practice, If the serial changes of specific patient are in parallel with the curve of the CA in Figure 3, it can be suggested that the treatment is progressing smoothly.

In our study, several factors were thought to affect the rate of improvement. First of all, because this was a retrospective study, the severity of plagiocephaly and age at the start were unevenly distributed. This was dealt with by mixed-effect model. In addition, the study population was limited to a single institution in Japan, which may affect the generalizability of the results to other populations and settings. Secondly, because patients kept wearing the helmets until head deformities resolved in this study, there can be a selection bias that the patients with more severe deformity accumulated in the group with longer-term of treatment. This was dealt with by limiting the analysis up to 15 weeks at which time point participating rate became <75% by early “graduation” of rapidly improving individuals. This cutoff timing seemed appropriate because serial changes in CA was relatively stable by this timing.

However, short-term results presented in this study (up to 16 wk) and long-term follow-up data will allow for a more comprehensive assessment of the lasting effects of helmet therapy, which is an area for future work.

Thirdly, regarding the compliance of helmet application, the bias caused by was thought to be small because the compliance was checked with record filled out by the guardians every 3-4 weeks under the instruction of wearing 23 hours a day.

Fourthly, closer examination of other factors that influence improvement rates, such as socioeconomic status and the presence of concurrent treatment, may further reduce bias.

Finally, the measurement was performed by 3 doctors, and the degree of agreement among the examiners was not confirmed. But there was a general trend that the value of CA decreased as treatment progressed, and it is speculated that there was not a large discrepancy in inter-rater consistency.

In this research, part of the cost for the helmet therapy was covered by research fund. This may have had some influence on family’s decision on selecting helmet or non-helmet therapy for their babies. However, the indications for helmet therapy has been unchanged throughout the study period, and treatment course would not have been influenced by the price of the helmet therapy.

CONCLUSION

In helmet therapy for DP, improvement of CA occurs inverse proportionally to the enlargement of head circumference. For those started earlier than 24 weeks of age or those with the severity of Argenta 1 or 2, improvement occurs rapidly and reaches nearly plateau at 12 weeks of treatment. For those started later or those with more severe deformity, improvement occurs slower and it continues even after 12 weeks of treatment. By understanding the progress of improvement, evaluation of treatment responsiveness and ideal timing to stop helmet therapy is possible.

Supplementary Material

scs-36-0531-s001.docx ^{(15.3KB, docx)}

scs-36-0531-s002.docx ^{(18.4KB, docx)}