Interpolation of the Maresh diaphyseal length data for use in quantitative analyses of growth

Abstract

The Maresh data are commonly used in bioarcheological growth studies as a representation of diaphyseal growth in a modern and healthy group of children. However, several problems with the way the data were reported have limited its use in quantitative analyses of growth. In this paper, we present updated and interpolated values for long bone length for age for use in calculating z‐scores, percentages of expected length, and other quantitative measures of growth. The Maresh mean and mean + 1 standard deviation values for the sexes separately and combined were first corrected for radiographic magnification. Several modeling approaches were then evaluated. This testing suggested that the best fit was provided by two third‐order polynomials fit to data ≤24 and ≥24 months, respectively. The resulting regressions were used to calculate age‐specific mean and standard deviation values in 1‐month intervals from birth until 12 years (0–144 months). Differences between the original and new values are minimal and do not exceed 1 mm. However, as the old Maresh values required rounding age down to the last attained threshold by as much as 5 months, there are differences between z‐scores calculated with original versus new values of up to 2 z‐score units, especially in children under 3 years of age where growth velocity is highest. Although these updated values do not solve the problems that made age estimation from the Maresh data unadvisable, they will be of use to researchers in conducting more precise growth studies in bioarcheological contexts.

Biological mortality bias in diaphyseal growth of contemporary children: Implications for paleoauxology

“…Z ‐scores are the preferred tool for comparing growth between groups of children because they quantify deviation from an expected measurement for age (e.g., height, length, weight), and thus enable comparison of growth across children of different ages. For more discussion on the advantages of z ‐scores in bioarchaeological growth studies, see Spake and Cardoso ( 2021 ) and for the use of z ‐scores in population health see WHO Expert Committee on Physical Status ( 1995 ). Z ‐scores for bone length for age were calculated from an interpolation of the Maresh ( 1943 , 1955 , 1970 ) reference data (Spake & Cardoso, 2021 ).…”

Bone structural data for the Denver longitudinal growth study

“…It is hoped that the availability of these individual data will allow many further ontogenetic and comparative studies. As noted earlier, long bone length data from the Denver Growth Study have already served as perhaps the most important modern comparative sample for a variety of bioarchaeological investigations (Humphrey, 2000; Spake & Cardoso, 2021). The addition of cross‐sectional diaphyseal and articular dimensions, as well as matched anthropometric dimensions, should increase opportunities to carry out studies of mechanically and physiologically relevant parameters in an ontogenetic context.…”

“…Anteroposterior (A-P) radiographs of the limbs were taken as part of the study beginning in 1935 and used in a series of investigations by Maresh of age-related changes in long bone lengths and diaphyseal external breadths, as well as fat and muscle breadths (Maresh, 1943(Maresh, , 1955(Maresh, , 1959(Maresh, , 1961(Maresh, , 1966(Maresh, , 1970. The long bone length data collected by Maresh have been used by a large number of other researchers to study basic patterns of human growth (Buschang, 1982a(Buschang, , 1982bRuff, 2003b;Smith & Buschang, 2004, 2005 and as a modern comparative sample in many bioarchaeological studies (e.g., Geber, 2014;Humphrey, 2000;Ives & Humphrey, 2017;Johnston, 1962;Lovejoy et al, 1990;Merchant & Ubelaker, 1977;Spake & Cardoso, 2021;Wall, 1991;Y'Edynak, 1976). An extensive battery of anthropometric measurements were also taken at each participant visit (for a full list, see Hansman, 1970; also see Himes, 2006).…”

A growth area

“…However, there are issues with the use of the Maresh dataset in palaeopathology. The Maresh dataset is focused on Euro-American children from middle to upper-middle class backgrounds, which may not be wholly representative of the growth patterns expected among past or contemporary populations [ 103 , 104 ]. For example, few if any longitudinal childhood growth studies exist for Asian populations [ 105 ].…”

“…The creation of both population-specific bone growth standards for non-Western contexts and quantification of the range of variation that may arise under challenging childhood circumstances would greatly help palaeopathological researchers identify the most appropriate contemporary comparative group and see if growth variability observed in the past overlaps with the variability observed under present stressful conditions. Differences between children from different geographic regions or socioeconomic contexts may or may not be evident upon further investigation [ 104 , 106 , 147 ]; however, this requires confirmation. As palaeopathologists contextualize results from non-survivors, ontogenetic information on modern children from different circumstances may clarify the spectrum along which bone growth may occur and improve our sense of how wide this spectrum can be.…”