HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2311020963v1 231/1/129    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Megremis, S. D. Articles by Tsilimigaki, A. M. Search for Related Content PubMed PubMed Citation Articles by Megremis, S. D. Articles by Tsilimigaki, A. M.

Spleen Length in Childhood with US: Normal Values Based on Age, Sex, and Somatometric Parameters¹

Stylianos D. Megremis, MD, PhD, Ioannis G. Vlachonikolis, MA, DPhil and Amalia M. Tsilimigaki, MD

¹ From the First Department of Radiology (S.D.M.) and Second Department of Pediatrics (A.M.T.), Venizelio General Hospital, Heraklion, Crete, Greece; and Department of Social Medicine, School of Health Sciences, University of Crete, Voutes, Greece (I.G.V.). Received August 1, 2002; revision requested October 1; final revision received July 23, 2003; accepted September 16. Address correspondence to S.D.M., 23 Arsinois St, 71303 Heraklion, Crete, Greece (e-mail: efstel@med.uoc.gr).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To investigate with ultrasonography (US) normal spleen length in healthy children.

MATERIALS AND METHODS: The study comprised 512 healthy children (274 girls) with ages ranging from 1 day (full-term neonate) to 17 years who were examined between 1996 and 2001. The main sample comprised 454 children (249 girls) with body measurements (weight and height) between the 5th and 95th percentiles of the relevant growth curves. The remaining 58 children (25 girls) with body measurements outside the normal ranges formed a separate sample used for cross-validation. None had a problem that could affect spleen size. The relationships between the US-measured spleen length and age and body parameters were studied with nonlinear regression and multiple (backward stepwise) regression techniques. Normal ranges and related statistics were estimated and tabulated according to age group and sex. Spleen length growth curves and upper limits defined by the 90% upper confidence limit (UCL) are presented in graphs according to height, weight, and body surface area (BSA).

RESULTS: Spleen length was highly correlated with age, height, weight, and BSA; there was no statistically significant difference between the sexes. The exact pattern of these relationships was nonlinear (polynomial type of third order for age, height, and weight and exponential type for BSA). Multiple regression analysis indicated that age, height, and either weight or BSA had significant positive associations with spleen length. The spleen lengths among the sample of 58 children whose height and weight were outside the normal ranges of growth parameters did not influence the proposed upper limits (almost all were within the 90% UCLs with respect to height and weight for the main sample).

CONCLUSION: Normal spleen lengths and ranges in childhood were obtained with US in a large sample of individuals.

© RSNA, 2004

Index terms: Spleen, size, 775.92 • Spleen, US, 775.1298 • Ultrasound (US), in infants and children, 775.1298

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
A large number of pathologic entities can affect splenic size, and the clinical examination is far from accurate to detect small increases in size (1). The spleen may be palpable in 15%–17% of healthy neonates (2,3) and 10% of healthy children; however, in most individuals, it must be two to three times its normal size before it is palpable (3).

In children, knowledge of normal splenic size in relation to age and other biometric parameters (height and weight) of physical growth is of paramount importance for the determination of mild splenomegaly.

Ultrasonography (US) is an established safe, quick, and reliable method for the calculation of splenic dimensions, and among all the latter that have been used in the past, spleen length at the hilum is considered the most reproducible linear measurement (4).

In order to establish our own standards and suggest upper limits and to provide additional data to the literature on this subject, the purpose of our study was to investigate normal spleen length in healthy children with US.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Subjects, Measured Parameters, and US
Between February 1996 and December 2001, we prospectively examined 512 Greek children (274 [53.5%] girls and 238 [46.5%] boys) on the island of Crete. The age range was 1 day (full-term neonate) to 17 years; the mean age was 6.07 years (SD, 4.76), and the median age was 5.67 years. These children came to our hospital as outpatients for either a follow-up examination or a routine check-up. The follow-up examination concerned previously treated conditions, such as mild upper respiratory tract or urinary tract infections, while the check-up concerned routine cases with atypical recurrent abdominal pain or an innocent heart murmur. In addition, all the relevant medical records were checked by a pediatrician (A.M.T.) to ensure that selected children had no pathologic abnormalities (inflammatory, metabolic, congestive, traumatic, collagenous or hematologic diseases and malignancies) that could affect splenic size.

The study was approved by the scientific board of our hospital, and parents’ informed consent was obtained for all selected children. Underweight or overweight individuals or those with short or tall stature and small-for-date neonates, whose height (length) and weight were outside the corresponding ranges defined by the 5th and 95th percentiles of the relevant growth curves used in our hospital, were excluded from the main analysis. Instead, the data of these children served as the cross-validation sample of the main results; the cross-validation sample was used to examine whether the results of the main sample would be similar to or different from those of children whose height and weight were within the normal ranges of growth parameters.

In this way, the main sample comprised 454 pediatric individuals (249 [54.8%] girls and 205 [45.2%] boys); the mean age was 6.04 years (SD, 4.76; age range, 0–17 years). The second (cross-validation) sample comprised 58 pediatric individuals (25 [43.1%] girls and 33 [56.9%] boys); the mean age was 6.29 years (SD, 4.75; age range, 0–15.5 years). There was no significant difference between the proportions of girls (² = 2.85, df = 1, P = .091) or between the mean ages (t = 0.38, df = 510, P = .701) of the two samples.

Height (length in neonates and babies), weight, and body surface area (BSA) were also recorded by a pediatrician (A.M.T.). The BSA was calculated with the US device and the following equation: BSA = (Height)^0.725 · (Weight)^0.425 · 0.007184.

Measurements of spleen length with US were obtained by a radiologist (S.D.M.). The subject was lying in a supine or a slightly right lateral decubitus position. The measurement of spleen length was the optically maximal distance (ideally at the hilum) on the longitudinal coronal view (between the most superomedial and the most inferolateral points). Measurements were made during quiet breathing; in the older children, measurements were made while they were holding their breath. In most (87.5% [448 of 512]) cases, we obtained three sequential measurements and calculated the mean; thus, we ensured minimum intraoperator variation and greater accuracy and reliability of measurements. In the remaining 12.5% (64 of 512), that is restless infants and young children, we were satisfied with two measurements. Neither preparation nor sedation was needed.

We used high-resolution real-time US scanners (ATL Ultramark 8 or ATL HDI 5000; ATL Ultrasound, Bothell, Wash). We used a 3.5-MHz sector transducer and a 2–5-MHz convex probe for the older children. Alternatively, we used a 5.0-, 7.5-, or 10.0-MHz sector transducer and a 4–7-MHz pediatric probe for infants and younger children. For neonates and small infants, we also used a 5–12-MHz linear array transducer.

Statistical Analysis
Differences of continuous variables between two independent groups were assessed with the t test and the nonparametric Mann-Whitney test; more than two groups were assessed with analysis of variance (5). The application of the nonparametric test was to ensure that departures from normal distributions or small sample sizes (in subgroup comparisons) were properly handled.

In cases of multiple t tests, the Bonferroni and modified Bonferroni procedures (6) were used to control the increase in type I error (eg, rejection of the null hypothesis when in fact it was true; this is known to increase when many tests are performed). Associations between spleen length and each of the four variables—age, height, weight, or BSA—were assessed with the Pearson correlation coefficient; to identify the exact pattern of the relationship, nonlinear regression analysis was performed. Multiple regression, applied in a backward stepwise fashion, was used to test the independent influence of all investigated factors on spleen length (5). All P values are reported together with their significance level (< .05). All statistical analyses were performed by a statistician (I.G.V.).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Differences and Regression Analysis
Differences between the 11 age groups were highly significant (analysis of variance, F = 177.17; df = 10, 443; P < .001): older children had longer mean spleen length. In contrast, there were no significant differences between the sexes (t = 0.69, df = 452, P = .493).

When we studied the sex differences within age groups (Table 1), we found four age groups where the difference was marked or significant (t tests): 3–6 months (P = .05), 1–2 years (P = .054), 6–8 years (P = .012), and 10–12 years (P = .043). After we adjusted the level of significance to control the overall type I error of 11 multiple tests (with either the Bonferroni or modified Bonferroni procedure, = .0045 for 11 tests), none of the differences in the four age groups was significant. Similarly, when we tested the sex differences for spleen length within the four age groups with the nonparametric Mann-Whitney test, again no significant results were found.

First, the regression analysis with age in months as the independent continuous variable confirmed a high multiple correlation (R²) between spleen length and this variable (R² = 0.780). The relationship was the polynomial type (third order) and is shown in Figure 1. The regression coefficients indicated a significant overall growth pattern of spleen length in childhood with increasing age (within the range of 0–204 months). The significant nonlinear coefficients (quadratic and cubic) confirmed that the overall growth pattern has a nonuniform increasing rate.

View larger version (29K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Scatterplot shows spleen length plotted against age. Regression curve (black line) and approximate 90% upper confidence limit (UCL [dashed line]) are also presented. x = individual values.

View larger version (28K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Scatterplot shows spleen length plotted against height. Regression curve and approximate 90% UCL are also presented. Keys are the same as for Figure 1.

View larger version (30K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Scatterplot shows spleen length plotted against weight. Regression curve and approximate 90% UCL are also presented. Keys are the same as for Figure 1.

View larger version (29K): [in this window] [in a new window] [Download PPT slide]   Figure 4. Scatterplot shows spleen length plotted against BSA. Regression curve and approximate 90% UCL are also presented. Keys are the same as for Figure 1.

Multivariate Analysis
Table 2 shows the results of the multiple regression analysis (backward stepwise), with age, height, and BSA as explanatory variables. This was necessary, as the latter variables were intercorrelated. It can be seen that all except weight have significant independent positive associations with spleen length (R² = 0.815). The nonselection of weight in the final model can be explained by the fact that weight and BSA were highly correlated (r = 0.980). In fact, exclusion of BSA from the analysis yielded a final model, which included weight, with a highly significant coefficient (ß = 0.029, standard error = 0.008, P < .001) and similar results as before for age and height (R² = 0.813).

View larger version (27K): [in this window] [in a new window] [Download PPT slide]   Figure 5. Similar scatterplot as in Figure 2 incorporates spleen lengths of individuals with tall () or short () stature. Keys are the same as for Figure 1.

View larger version (27K): [in this window] [in a new window] [Download PPT slide]   Figure 6. Similar scatterplot as in Figure 3 incorporates spleen lengths of overweight () or underweight () individuals. Keys are the same as for Figure 1.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

We chose to evaluate only spleen length with US as a reliable and reproducible measurement, according to our experience and that of other investigators (4). Furthermore, it has been shown that spleen length correlates well with measurements at autopsy (17). As in other studies (4,9), we presented all relevant statistics tabulated in 11 age groups. We also displayed results for both sexes with consideration of two facts: Boys had consistently longer, but nonsignificant, mean and median spleen length compared with girls of all ages (except for the median in the 2–4-year age group). Although nonsignificant, the differences according to sex were marked in some age groups.

Although researchers in the aforementioned studies rounded up spleen lengths, we avoided approximating them in Table 1 with regard to descriptive statistics. In these articles, proposed upper and/or lower values mostly concern subjects whose ages are in the boundaries of each age group.

Spleen length was highly correlated with age and all body parameters (height, weight, and BSA). This was displayed also in the graphs that show both the regression curve and the 90% UCLs. Such graphs are useful for cross-checking assessment of spleen lengths according to age, with corresponding assessments of the same spleen length according to each body parameter. We did not draw curves according to sex in order to avoid overcomplication of the graphs.

In 1983, Dittrich et al (8) reported the first spleen size normogram with sonographically determined volume estimation in 194 healthy children, but this method is considered very cumbersome, time-consuming, and not reproducible (4). In 1991, Rosenberg et al (4) used data from 230 healthy children of all ages and suggested upper-limit guidelines for spleen length in 11 definite age groups without mention of somatometric factors. When we compared the median spleen lengths according to age group, our findings roughly seemed to agree with theirs.

In 1998, Loftus and Metreweli (7) sonographically measured both spleen length and kidney length in 256 healthy Chinese children, and they compared spleen lengths with the results of the study of Rosenberg et al (4). They found that spleen length in Chinese children to the age of 15 years was similar to that of Western children. They also suggested that splenomegaly should be suspected in children if the spleen is more than 1.25 times longer than the adjacent kidney.

In the same year, Konus et al (9) examined 299 healthy children by using sonography and provided results for both spleen length and transverse spleen dimension (in coronal section). Unfortunately, the age groups used in their study were not identical to those in ours, but we can make an approximate comparison by taking into account the differences in both age and height ranges. This reveals that their results were also not very different from ours.

In 2000, the series of Capaccioli et al (10) was an in vivo anatomic study of 237 healthy children. The researchers sonographically measured the three dimensions of the spleen (length, width, and thickness) and tabulated the means for every year (0–12 years). They reported that spleens were longer in male individuals, especially from the 5–6-year age group to older age groups, but did not specify if the sex differences were significant. Finally, the presentation of their results does not allow a realistic comparison, although it seemed roughly that our mean spleen lengths were longer than were theirs in most of the age groups.

None of the previous studies includes individuals outlying normal growth curves; researchers in these studies were trying to avoid potential deviations (distortion) in normal spleen lengths. We found that by taking into account the relevant somatometric parameters most of the children with short or tall stature or with low or high weight have spleen lengths within the normal range. Finally, small-for-date neonates—with a mean weight (at the time of measurement) of 1,850 g and a mean height of 44.8 cm—as expected, had the shortest spleens (mean, 3.1 cm) in the normal scatterplot.

We believe that by collecting the largest series so far, to our knowledge, of US-determined pediatric spleen length measurements, we have estimated more accurately its normal values in childhood. Our aim was to provide a more objective assessment of mild splenomegaly during routine abdominal US. Our results could be used as a practical and comprehensive guide to indicate the normal spleen length range for every child, according to his/her age and his/her body habitus. With this in mind, we chose the spleen length 90% UCLs so as to distinguish and thus better assess individuals with markedly long spleens outside the normal range but whose body parameters are within the normal range.

	ACKNOWLEDGMENTS

 
We thank Athanasios Alegakis, PhD, for his invaluable help in the completion of this study and Evaggelia Sfakianaki, MD, for all her support and fruitful comments.

	FOOTNOTES

 
Deceased

Abbreviations: BSA = body surface area, UCL = upper confidence limit

Author contributions: Guarantors of integrity of entire study, S.D.M., I.G.V.; study concepts and design, S.D.M., I.G.V.; literature research, S.D.M.; clinical studies, S.D.M., A.M.T.; data acquisition, S.D.M.; data analysis/interpretation, S.D.M., I.G.V., A.M.T.; statistical analysis, I.G.V.; manuscript preparation, definition of intellectual content, and final version approval, S.D.M., I.G.V.; manuscript editing and revision/review, S.D.M., I.G.V., A.M.T.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Zhang B, Lewis SM. A study of the reliability of clinical palpation of the spleen. Clin Lab Haematol 1989; 11:7-10.[Medline]
2. Mimouni F, Merlob P, Ashkenazi S, Litmanovitz I, Reisner SH. Palpable spleens in newborn term infants. Clin Pediatr (Phila) 1985; 24:197-198.
3. French J, Camitta BM. Splenomegaly In: Nelson textbook of pediatrics. 15th ed. Philadelphia, Pa: Saunders, 1996; 1439.
4. Rosenberg HK, Markowitz RI, Kolberg H, Park C, Hubbard A, Bellah RD. Normal splenic size in infants and children: sonographic measurements. AJR Am J Roentgenol 1991; 157:119-121.[Abstract/Free Full Text]
5. Armitage P, Berry G. Statistical methods in medical research 3rd ed. Oxford, England: Blackwell, 1994.
6. Simes RJ. An improved Bonferroni procedure for multiple tests of significance. Biometrika 1986; 73:751-754.[Abstract/Free Full Text]
7. Loftus WK, Metreweli C. Ultrasound assessment of mild splenomegaly: spleen/kidney ratio. Pediatr Radiol 1998; 28:98-100.[CrossRef][Medline]
8. Dittrich M, Milde S, Dinkel E, Baumann W, Weitzel D. Sonographic biometry of liver and spleen size in childhood. Pediatr Radiol 1983; 13:206-211.[CrossRef][Medline]
9. Konus OL, Ozdemir A, Akkaya A, Erbas G, Celik H, Isik S. Normal liver, spleen and kidney dimensions in neonates, infants and children: evaluation with sonography. AJR Am J Roentgenol 1998; 171:1693-1698.[Abstract/Free Full Text]
10. Capaccioli L, Stecco A, Vanzi E, Brizzi E. Ultrasonographic study on the growth and dimensions of healthy children and adults organs. Ital J Anat Embryol 2000; 105:1-50.[Medline]
11. Rasmussen SW, Christensen BE, Holm HH, Kardel T, Stigsby B, Larsen M. Spleen volume determination by ultrasonic scanning. Scand J Haematol 1973; 10:298-304.[Medline]
12. Zhang B, Lewis SM. Use of radionuclide scanning to estimate size of spleen in vivo. J Clin Pathol 1987; 40:508-511.[Abstract/Free Full Text]
13. Markisz JA, Treves ST, Davis RT. Normal hepatic and splenic size in children: scintigraphic determination. Pediatr Radiol 1987; 17:273-276.[CrossRef][Medline]
14. Ishibashi H, Higuchi N, Shimamura R, Hirata Y, Kudo J, Niho Y. Sonographic assessment and grading of spleen size. J Clin Ultrasound 1991; 19:21-25.[Medline]
15. Schlesinger AE, Edgar KA, Boxer LA. Volume of the spleen in children as measured in CT scans: normal standards as a function of body weight. AJR Am J Roentgenol 1993; 160:1107-1109.[Abstract/Free Full Text]
16. Prassopoulos P, Cavouras D. CT assessment of normal splenic size in children. Acta Radiol 1994; 35:152-154.[Medline]
17. Loftus WK, Chow LT, Metreweli C. Sonographic measurement of splenic length: correlation with measurement at autopsy. J Clin Ultrasound 1999; 27:71-74.[CrossRef][Medline]

This article has been cited by other articles:

				M. Friedrich-Rust, C. Koch, A. Rentzsch, C. Sarrazin, P. Schwarz, E. Herrmann, A. Lindinger, U. Sarrazin, T. Poynard, H.-J. Schafers, et al. Noninvasive assessment of liver fibrosis in patients with Fontan circulation using transient elastography and biochemical fibrosis markers J. Thorac. Cardiovasc. Surg., March 1, 2008; 135(3): 560 - 567. [Abstract] [Full Text] [PDF]

				S. Megremis, A. Alegakis, and M. Koropouli Ultrasonographic Spleen Dimensions in Preterm Infants During the First 3 Months of Life J. Ultrasound Med., March 1, 2007; 26(3): 329 - 335. [Abstract] [Full Text] [PDF]

				S. D. Chernausek, P. F. Backeljauw, J. Frane, J. Kuntze, L. E. Underwood, and for the GH Insensitivity Syndrome Collaborative Gr Long-Term Treatment with Recombinant Insulin-Like Growth Factor (IGF)-I in Children with Severe IGF-I Deficiency due to Growth Hormone Insensitivity J. Clin. Endocrinol. Metab., March 1, 2007; 92(3): 902 - 910. [Abstract] [Full Text] [PDF]

				R G Hosey, C G Mattacola, V Kriss, T Armsey, J D Quarles, J Jagger, D Vigil, and C M Lebrun Ultrasound assessment of spleen size in collegiate athletes * Commentary * Commentary. Br. J. Sports Med., March 1, 2006; 40(3): 251 - 254. [Abstract] [Full Text] [PDF]

				A. A. Safak, E. Simsek, and T. Bahcebasi Sonographic Assessment of the Normal Limits and Percentile Curves of Liver, Spleen, and Kidney Dimensions in Healthy School-Aged Children J. Ultrasound Med., October 1, 2005; 24(10): 1359 - 1364. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2311020963v1 231/1/129    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Megremis, S. D. Articles by Tsilimigaki, A. M. Search for Related Content PubMed PubMed Citation Articles by Megremis, S. D. Articles by Tsilimigaki, A. M.