Vision of Research on Human Linear Growth

Abstract

The human body grows in length from conception to the maximal adult height over two decades. The shortest male population averages ∼150 cm and the tallest ∼183 cm. Nonetheless the dimensions of head and trunk are highly comparable, with the vast difference in the leg length. Stunting is a personal condition in which an individual has a standing height-for-age (HAZ) of less than two standard deviations of the standard curve median. It is associated with increased mortality, morbidity, and functional deficits. The process of losing relative stature is known as linear growth retardation, first attributed to chronic protein deficiency, then to an assortment of micronutrient deficiencies, and most recently to inflammation from unhygienic environmental conditions. Public health intervention trials responding to each of these possibilities have failed to produce true reversal responses measured in the 10s of centimeters. As to biological insights, there is no convenient way to separate weight from length growth with sonographic monitoring, but a third of infants can be born stunted. Normative growth (standard curves) competes with epigenetic adaptation (programming) as the beacon for in utero growth. Major investments into field trials allow us to discard multiple micronutrients and water/sanitation/hygiene interventions as measures to reverse established stunting. The preponderance of evidence is against catch-up growth during puberty. Future publications will be in the conceptual domain, resolving metrics, while the full range of stimuli and exposures impeding growth will be elucidated. Advances in measurement techniques in anthropometry and immunology and endocrinology will be mobilized to the literature.

Introduction

Growth is a characteristic of all life from unicellular organisms to the most complex of plant and animal species. It operates in 2 interacting domains. The first relates to the organisms’ mass of liquid and solid tissues, generally assessed as weight on a balance scale. The second relates to linear or curvilinear dimensions that contribute the attributes of length, width, and thickness and determine the volumetric size of the organism. Clearly, there is an interaction between the mass and the volume domains, with linear—and often breadth—being generally irreversible in their extension, whereas gain and loss of total weight and distinct tissues is a fluid process. A thoughtful treatise on the geographic differences in deficiency of tissue weight with conservation of height (wasting) and loss of stature, with proportional weight for height (stunting), was written by Victora. ¹

In this review, we specifically focus on the linear dimensions (length or height/stature). Some species, such as dogs, cats, horses, humans, and most insects and birds, and so on, reach a definitive, genetically guided adult length, whereas fish, mollusks, and reptiles, as well as trees, bushes and vines, continue to elongate throughout their life span. Nutrient availability can be a determinant of growth velocity or final adult length. It has been recognized that the pure deficiency of any essential nutrient will result in reduction or cessation of growth of a young organism. In fact, the bioassay for essentiality of a specific nutrient was the effect on growth with specific depletion. ²

Moreover, Michael Golden ³ identified a phenomenon with the acute and total restriction of a family of nutrients that have no storage pools in the body, namely, N, Na, K, Mg, Cl, and Zn (type II nutrients), for which immediate [cessation] of growth is the survival response to avoid circulating or tissue dilution with fatal metabolic consequences. The field has been troubled, therefore, by the logical confusion between the truism postulate that nutrient deficiency can cause slowing of growth and the generalized assumption that observed slowing in growth is related to nutrient deficiency. Indeed, as outlined subsequently, non-nutritional factors play a profound role in linear growth, and understanding and managing the interplay of adverse factors are only partially understood.

Classical and Established Insights

Archeologists and anthropologists have demonstrated that human heights have varied over the millennia of Homo sapiens’ reign in this world. Fogel ⁴ documented a secular trend in height in Europe from the mid-19th century onward, putatively related to improvements in food supply (nutrition) and health. Indeed, the median heights of under-5 children vary widely from society to society ⁵ , as [do] the median adult heights for males and females. ⁶

Maximal Adult Height Potential

A focus across the tallest and shortest adults gives a useful baseline perspective on what the extent and elasticity of human growth really is and how much leeway of determinacy is affected by adversity and would theoretically be responsive to preventive or corrective measures in a clinical or public health context. The greatest average (median) across humanity is 182 cm in the Netherlands, ⁷ and it is considered that Dutch men have peaked out at the maximal human genetic potential for stature. The current practical definition of adult male stunting is 150 cm for men, which is the median height for elderly men in the Western Guatemala township of Nahuala (Unpublished findings, Guardado, 2019).

One perspective on this is that 150 cm is the default stature for humans, the height that can be reached under the very least propitious conditions for growth across humanity. That is, 80% of the maximal adult height is intrinsic, and the remaining 20% or 35 cm is malleable or modifiable within the context of the dietary and environmental exposures accumulated across the period of stature growth.

Classification of Human Short Stature

The foregoing considerations would logically evoke population distributions to be the locus of analysis and indicate that factors penetrating across whole societies would be an important focus. Unfortunately, a more clinical and individual binary diagnostic classification of “too short,” or not, emerged. This would come to dominate the paradigm of linear growth and its disorders in the nutritional community. With respect to linear growth, the earliest—and most enduring—insights relate to its classification and putative association with the protein and energy nutrition of an individual. ^8,9 The formal origins can be traced to 1972 to an article by John Waterlow in the British Medical Journal entitled “The classification of protein-calorie malnutrition.” ⁹ The preeminent concern of public health nutrition from the end of World War II into the 1970s had been clinical forms of protein-energy malnutrition (PEM): kwashiorkor and nutritional marasmus. ¹⁰ Waterlow asserted that growth deficits in weight and stature in children represented a lesser form of PEM and possibly could be a stepping stone on a pathway to the more severe, clinical forms. He coined the term “stunting” for the condition of notable, if not pathological, short stature in which the deficit fell below the 10th percentile of the growth reference charts of the era. He comments: “…the extent of height deficit in relation to age may be regarded as a measure of the duration of malnutrition…height for age gives a picture of past nutritional history.” ⁹ As a consequence, a synonym for stunting became “chronic malnutrition,” with the definitional implications of a relationship to origins in PEM.

The 10th percentile was an arbitrary cutoff. It was gradually replaced in an international context by 2 standard deviations around the mean of the reference population distribution. This corresponds to the 2.5th percentile, which is the criterion for stunting (as well as underweight and wasting) in the 2006 World Health Organization (WHO) Child Growth Standards. ¹¹

Waterlow ⁹ did introduce a complementary concept at the same time, which was “linear growth retardation” to be understood as the process of deterioration in height for age. It is virtually a reduced linear growth velocity relative to the first derivative of the median curve. Linear growth retardation is the process that eventually leads to overt stunting. If one begins with a superior stature, it may take years for the HAZ to decline to the −2 Z-score level. An individual slipping in his or her HAZ, however, has “normative” stature but does not likely have normal health.

Consequences and Correlates of Short Stature

The process of linear growth retardation could have adverse effects on life and health either by damage from the factors that produce the growth impairment or consequences of the size and proportions achieved into adulthood or both. ¹² The most consequential outcome of stunting in a public health context is its association with a shortened life span and increased mortality. ¹³

Several reviews have summarized additional consequences to health and function associated with the stunted condition. ¹⁴ There is growing evidence of the connections between slow growth in height early in life and impaired health and educational and economic performance later in life. Recent research findings, including follow-up of an intervention trial in Guatemala, indicate that stunting can have long-term effects on cognitive development, school achievement, economic productivity in adulthood, and maternal reproductive outcomes. As enunciated by Prendergast and Humphrey ¹⁵ :

We view this condition as a ‘stunting syndrome’ in which multiple pathological changes marked by linear growth retardation in early life are associated with increased morbidity and mortality, reduced physical, neuro-developmental and economic capacity, and an elevated risk of metabolic disease into adulthood. Stunting is a cyclical process because women who were themselves stunted in childhood tend to have stunted offspring, creating an intergenerational cycle of poverty and reduced human capital that is difficult to break.

Proposals of Macro- and Micronutrient Deficiencies for a Dietary Etiology of Short Stature

As documented in the publications by Waterlow and other adherents of the “protein gap” theory, short stature was a form of “chronic undernutrition” related to insufficient quantity and/or quality of dietary protein. ^9,16 As the underpinnings of a macronutrient origin weakened, propositions regarding specific micronutrient deficiencies emerged.

Macronutrients

Let it be stipulated that the severe dietary deficiency of energy sources, proteins, or both produce predictable and profound interruption of both weight and length growth in laboratory animals, livestock, and poultry; the degree, however, to which this models or reflects human realities in nonclinical settings is the crux of the matter. The nutrient deficiencies proposed are associated with linear growth impairment. These have included the quality and quantity of dietary protein and energy, specifically in the earliest explorations of linear growth in children. ^8,9,16 In fact, the first use of the term “stunting” came in the 1972 article by Waterlow entitled, “Classification and definition of protein-calorie malnutrition.” ⁹ It was inferentially linked to PEM by the title. Beaton and Ghassemi ¹⁷ shook the fundamentals of the protein-energy link to poor growth in the 1982 review of the results of dietary supplement trials, which to that date had showed generally no efficacy or effectiveness of the interventions. However, they noted that the extant programs had not been tightly targeted to the most susceptible individuals and left out the possibility that subgroups within the intervened groups received differential benefit. Contemporary analyses by Arsenault and Brown cast doubts regarding the validity of the currently high recommendations of protein during early growth; ¹⁸ they further assert that, when intakes are not low, most global complementary feeding patterns provide sufficient protein to meet true requirements. ¹⁹

Micronutrients

On the one hand, the putatively most common micronutrient deficiency, that of dietary iron, has not been linked with impairment of linear growth. ²⁰ In fact, the concept of “outgrowing” one’s iron supply as a mechanism for precipitating impaired erythropoiesis and clinical anemia is widely embraced. On the other hand, the atomically similar trace element, zinc, has been strongly implicated in linear growth. A syndrome of dwarfism and hypogonadism, variously associated with anemia, was described in adolescents in nomadic groups at desert oases in Iran and Egypt in the early 1960s, associated with rudimentary biomarkers of zinc status. ^21,22 This prompted a series of zinc supplementation intervention trials in short children or short-stature populations over the decades that followed. Around the turn of the millennium, the first meta-analyses of zinc supplementation trials appeared, ^23,24 suggesting minor but positive response amounting to no more than a few centimeters of catch-up. A more recent meta-analysis by Imdad and Bhutta ²⁵ showed that supplementation of zinc alone, when compared to in combination with other micronutrients, especially iron, produced a significant effect size; they report, however, that 10 mg/d of oral zinc resulted in only a 0.4-cm increase in linear growth compared to controls.

A mean differential growth of 0.025 cm per month below age 2 years and 0.05 cm per month up to 4 years was claimed for a large sample of vitamin A-supplemented and nonsupplemented children from 0 to 48 months of age. ²⁶ The sample size enabled achievement of statistical significance with these differences, but any biological importance to such miniscule growth increment is dubious.

Ramakrishnan et al addressed the issue of all micronutrients, singly and in combination, in supplementation trials aimed at enhancing growth of populations challenged by short stature. ²⁷ The meta-analysis of studies related to single-nutrient interventions with iron, vitamin A, and even zinc in samples of less than 5 years of age revealed no significant effects on linear growth. For the 20 interventions using multiple micronutrient supplements in children in this age range, a very modest, but significant, aggregate 0.09 Z-score effect size was reported.

Alternative Postulates for the Origin of Short Stature

Challenges to the paradigm of low HAZ signifying a forme fruste of PEM emerged only shortly after the formulation by Waterlow. ⁹ It was recognized in El Salvador ²⁸ and shortly thereafter in the Negev Desert in Israel ²⁹ that children classified as third-degree malnourished, based on severe height-for-age deficits, were misdiagnosed. In fact, although extremely short, they were in no way clinically malnourished and had appropriate weight for their diminished lengths.

Another paradox was the timing of the development of stunting, that is, in early childhood. ^30,31 Across the low- and middle-income settings of the world, the reduction in HAZ commences within the first month of life and decreases steeply through 6 months of life and on to a year. If one supposes that the infant diet consisted exclusively or predominantly of human milk, which is postulated to meet all nutrient requirements in the first semester of infant life, ³² this period of life should be protected for all infants. Acceleration of height loss might reasonably be expected to initiate with the advent of nonmilk complementary feeding but not directly from birth.

Thus, over the years since 1972, a series of alternative hypotheses for the process of linear growth retardation have emerged, prioritizing the poor utilization of essential nutrition, rather than its scarcity from the diet. Implicated in the alternative theses were noxious exposures from the unsanitary and impoverished environments in which poorly growing populations tended to reside. Poor nutrient absorption was the basis of a notion implicating “Tropical enteropathy” as a contributing cause for stunting. ³³ Complementary to the gut dysfunction hypothesis was the conjecture that a persistent immunostimulation from both pathogenic and nonpathogenic microbes activated the catabolic effects of inflammatory hormonal mediators (cytokines). ^34,35 A contemporary review and reflection by Millward ³⁶ mobilizes the supporting evidence in the intervening years. Thus, ambient stressors entered the discussion as the elements of importance in the production of linear growth retardation.

Emerging Insights

Leg Length and Linear Growth Retardation

The non-nutritional origin of short stature has been the counterproposition to the original PEM theory and its corollaries in the micronutrient malnutrition domain. A clarification of the specific anatomical features of this so-called “chronic undernutrition” has refined the discussion and understanding. Namely, anatomic contribution to short stature is asymmetrical. The elongation of the lower extremities is more severely affected in comparison to the components of the upper body. Penny ³⁷ illustrated that 90% of the trunk (head and torso) length was preserved in a stunted Peruvian girl compared to her normal-height, age-matched peer, but the legs were only 75% of the reference. Bogin and Varela-Silva placed a more quantitative frame around this, showing an 11-cm difference in height between US-reared and Guatemalan-reared 9-year-olds, with over 8 cm of the difference attributable to the length of the lower extremities. ³⁸

In what they term the “thrifty phenotype,” Pomeroy and colleagues, ³⁹ from the biological anthropology disciple, underscore the relative conservation of cranial volume (head circumference) and trunk size in children of indigenous ancestry raised at sea level in Peru (Lima) when compared to the major sacrifice in the lower extremities for those raised at altitude (Ayacucho, at 2761 meters). They ascribe a teleological significance to these disproportions as an adaptation to preserve the size and function of the most vital organs, including the brain and the pulmonary, circulatory, excretory, and digestive systems.

The elongation of long bones is dependent on 3 primary factors: (1) a nutritional factor, assuring a necessary availability of the nutrients required for assembling bony tissue, primarily amino acids, calcium, and phosphorus; (2) a cellular factor, consisting of dense proliferation and hypertrophy of chondrocytes in the growth plate to convert the collagenous matrix into cortical bone distal to the epiphysis; and (3) a hormonal process of signaling the coordination of osseous synthesis within the chondrocytic stratum. ⁴⁰

Choline: A Potential First-Limiting Nutrient in Bone Elongation

The renowned professor at Giessen University in Germany, Julius Liebig, developed the concept of the first-limiting nutrient, meaning that in a situation of multiple nutrient deficiency, recovery may not occur with nutrient intervention unless the vitamin or mineral producing the manifestation of interest is addressed. The composition of human milk was mobilized early as evidence against candidate nutrients such as amino acids, calcium, and phosphorus being insufficiently available for forming new bone in infants. Complementing the observation on leg-length determination of shortness in stunting, the obscure—but ubiquitous—nutrient choline merits consideration as a major nutritional factor.

In a mechanistic sense, evidence for a role of choline-containing phospholipids in cartilage and bone formation emerged 2 decades ago ⁴¹ and has been refined ever since. A phosphocholine/phosphoethanolamine phosphatase enzyme functions in the mineralizing process in cartilage cells of the growth plate to release phosphorus for the consolidation of hydroxyapatite for bone mineral. ⁴² Studies in a knock-out mouse model indicate, moreover, that another choline-related enzyme, choline kinase beta, is essential to the mineralization of developing cortical bone. ⁴³

An observational study among deprived children in Malawi brings the choline consideration into a human context. ⁴⁴ A metabolomic study on analytes from Malawian children between 12 and 59 months, with a 62% stunting prevalence, provides the insights. Choline-bearing metabolites, part of the serum glycerophosphate pool, including numerous subvarieties of lyso-, diacyl-, and acyl-alkyl phosphatidylcholines, were among the analytes most strongly associated with HAZ classification. The finding showed disordered patterns of these choline metabolites—both higher and lower concentrations—in the stunted children when compared to those with HAZ scores equal to or above −2. Meanwhile, sphingomyelins are primarily derived from ceramide and choline; their circulating concentrations were lower in the stunted individuals. The authors comment:

The findings of the study also support the notion that stunted children do not receive sufficient dietary choline, as reflected by low serum sphingomyelins and alterations in phosphatidylcholines. Future studies are needed to characterize serum choline in children with stunting.

Recently, attention to the regular consumption of hen’s eggs has emerged in the literature related to linear growth. ^45,46 The high choline content of eggs and frequent consumption as a dietary source might underlie any biological linkage of the practice to improved early linear growth.

In Utero Growth Standards

Impaired growth is not exclusively a phenomenon occurring in the infancy and toddler years and can begin during the period of in utero development of the fetus. The pattern of fetal growth and the duration of gestation determine the size (length, weight, head circumference) of a newborn at birth. Intrauterine growth restriction (IUGR) has been defined as the condition of poor gestational growth, with a birth weight below the 10th percentile of a reference birth distribution. To some, it is equivalent to small-for-gestational age (SGA). The aspect of SGA is important, as IUGR is usually distinguished from premature, which also produces low birth weight, even when the growth trajectory had been normal in utero. The origin of IUGR can be related to hereditary diseases or congenital abnormalities with an endogenous basis in the fetus, or it can be caused by infections or the lack of nutritional substrate delivery, including that of oxygen and energy as well as other macro- or micronutrients.

The use of ultrasound to estimate total fetal volume and body weight dates back to the 1980s. The technology has made progressive strides over the past 4 decades in precision and validity, ⁴⁷ but random errors persist in its application over time. ⁴⁸ Obstetric epidemiologists have led the way in a mission to develop a distribution of normative fetal growth, constructing a massive international collaborative consortium called the International Fetal and Newborn Growth Consortium for the 21 Century (INTERGROWTH-21st) conducted in obstetric centers for 5 years in 8 countries: Brazil, China, India, Italy, Kenya, Oman, United Kingdom, and United States. ⁴⁹ One of its first products was the INTERGROWTH-21 Fetal and Newborn growth standards. ⁵⁰

The INTERGROWTH adherents insist that the fetal growth curve is analogous to the WHO curve for children, ¹¹ insofar as it is regarded as a prescriptive standard, a pattern to be followed to represent optimal fetal growth. This disregards that it is achieved in situations of optimal maternal conditions of environment, diet, and health and that a fetus cannot be molded into the prescribed size with interventions other than the natural experience of the reference population.

As stated initially, the unique focus in IUGR is that of volume and weight, not length. It can logically be concluded that fetal size would generally correlate with linear dimensions but not in a tightly predictive manner. As with physical anthropology in live children, both long and slender and short and stocky phenotypes can share a common body weight in fetuses. This would loosen the linear and ponderal associations. Indeed, in an application of the INTERGROWTH standards in Zimbabwe, an initial association between fetal growth status and mortality disappeared when adjustments were made for length (stunting). ⁵¹ In utero length trajectories would be a useful subject of study, but weight and volume indices are unlikely to be adequate proxies for fetal linear growth, and the current crown-rump dimension excludes the critical variance in the lower extremities.

Multimicronutrient Supplementation

A conceptual and technical breakthrough for the provision of the amounts of energy and protein needed to rescue and rehabilitate children with severe acute malnutrition came in 1999 with the innovation of Ready-to-Use Therapeutic Food (RUTF). ⁵² Conventional formulas were powdered and milk based, requiring safe water for their reconstitution and with limited conservation time in the prepared form. The RUTF was a lipid-based spread in aluminum foil packaging; the anhydrous nature of fat suppresses the growth of bacteria and fungi, allowing room-temperature storage. ⁵²

Meanwhile, anemia was receiving attention in the beginning of the new millennium after the 1998 emission by Stoltzfus and Dreyfuss of guidelines for oral iron and folate supplementation. ⁵³ From multiple industry and academic quarters came the idea to combine multiple essential nutrients with the iron vehicle as a value-added strategy to support vitamin and mineral nutrition. ⁵⁴ One emerging idea was to use the environment-resistant and quasi-food format of the lipid-based RUTF in smaller amounts for the iron and assorted micronutrient delivery process to create tasty and field-stable spreads, eventually to be known as small-quantity, lipid-based nutrient supplements (sq LNS).

Adu-Afarwuah et al, ⁵⁵ then a graduate student at the University of California at Davis, led a randomized field trial to compare 3 modalities of multinutrient supplements (tablet, powder, and sq LNS) for their relative efficacies to correct anemia. Each intervention was equivalent in reducing anemia. A secondary finding was an improvement in weight for age in the treatment arm offering the lipid-based intervention. This observation enabled Kathyrn Dewey and collaborators to parlay this ponderal growth observation into a multinational follow-up project, focusing on anemia and assorted growth effects. This effort was called the International Lipid-based Nutrient Supplement (iLiNS) Project. ⁵⁶ In going from the limited observation in Ghana to an international scale in Asia (Bangladesh) and elsewhere in Africa (Malawi), it morphed in 2 important ways: (1) in most instances, mothers would receive some sort of nutritional support during pregnancy and 6 months of lactation; and (2) it focused not on ponderal growth but on linear dimensions. Four prospective iLiNS field trials of effectiveness of randomized and controlled nature are of interest. ^{57 -60}

In Malawi, ⁵⁷ a 6-arm supplementation initiative was conducted restricted to children, through the period from 6 to 18 months of life. A no-treatment control group was compared to daily rations of 10-g LNS, 20-g LNS alone, 20 g of LNS with milk, 40-g LNS alone, and 40-g LNS with milk. On the aggregate, the cohort grew by 13 cm in length over the year, with a −0.45 Z-score, and with no differences across treatments. None of the treatments retarded relative height loss.

Bangladesh was the site of an effectiveness trial in which mothers received a nutrient supplement through pregnancy and the first 6 postpartum months, while their offspring received one or another multiple micronutrient supplement for the 18 months thereafter. ⁵⁸ In one set of dyads, the mothers received LNS, whereas the others received standard iron and folic acid supplementation (IFA) for up to 14 months. From 6 months of age, LNS was given to infants of the first mother–infant dyad set (LNS–LNS) and to a second dyad set (IFA–LNS). Infants in the third group received multimicronutrient powder (IFA–MNP), and control group infants received no supplement (IFA-none). The only remarkable differences were found in comparisons between the LNS–LNS dyads and the IFA–MNP dyads, as the former had a 30% lower stunting prevalence than the latter at 18 months, declining to 21% lower at 24 months. At that end line, moreover, the former group demonstrated a 0.13 LAZ increase over the latter. These variables did not differ from one another nor from the aforementioned groups in the control (IFA-none) or the other lipid group (IFA–LNS) at either time point.

Randomized field trials in Malawi ⁵⁹ and Ghana ⁶⁰ shared a virtually identical, complex design involving 3 treatment arms with supplemented mother–infant dyads. Pregnant women were randomized to receive 1 of the 3 forms of daily nutritional supplements from early gestation to 6 months postpartum: IFA group, which switched to placebo at birth; multiple micronutrient tablet throughout (MMN group); and sq LNS. The infants in the first 2 groups received no supplementation, and those of mothers receiving sq LNS received an appropriately sized pediatric dose of the same supplement from 6 to 18 months. Final measurement of length was made in all infants at age 18 months. At this end line point, toddlers at the Malawian study site were over 2 cm shorter and 3 times more stunted than their peers at the Ghanaian site. The respective IFA, MMN, and sq-LNS groups in Malawi had lengths of 77.0, 76.9, and 76.8 cm, respectively, corresponding to 32.7%, 35.6%, and 37.9% stunting rates. No significant differences were found among treatments. In Ghana, the respective lengths by treatment group were 79.1, 79.1, and 79.7 cm, with stunting rates of 13.7%, 12.9%, and 8.9%. Statistical tests showed significant superiority of the 0.6 cm of additional height with the lipid-based supplement.

We might summarize the effects of multiple micronutrients delivered to mothers and children (or just children) in this iLiNS series as generally paltry, perhaps in one instance (Ghana) because there was very little short stature to address. Since none of the protocols cast the LAZ distribution to the center channel of the growth standard, ¹¹ only palliation would result from implementation of these measures.

Water, Sanitation, and Hygiene (WASH)

As the negligible contributions toward correction toward a fully normal height were emerging from the multinutrient supplement intervention trial, the other hypothesis of impaired growth was being tested, that of exposure to environmental stressors. ^35,36

Observational and cohort studies

Two studies, each using one or the other putative biomarkers of environmental enteric dysfunction, show convergent findings related to exposure to safer or less-safe household water. With respect to linear growth, the study in Timor-Leste ⁶¹ did not report on growth, and in Uganda, ⁶² the additional growth was not of a magnitude to alter the stunting rates in the community.

Randomized intervention trials

Three major WASH studies have been published within the last year from Kenya, ⁶³ Bangladesh, ⁶⁴ and Zimbabwe. ⁶⁵ The former 2, under the umbrella of the iLiNS, have a common protocol, the design and rationale of which has been published. ⁶⁶ The common protocol included 7 arms: passive controls (no treatment, measurement), active controls (no treatment, monitoring, measurement); nutrition (daily sq LNS only to child from 6 months onward); water (home chlorination); sanitation (construction of improved latrine); handwashing (soap and hand-washing station); combined WASH (water, sanitation, and hand-washing); and combined plus nutrition (all 4 interventions). Mothers were enrolled during pregnancy for participation until their offspring were measured at 24 months of age. The findings were parallel in both Kenya ⁶³ and Bangladesh. ⁶⁴ All of the interventions, with the exception of water, reduced diarrhea and both arms with 18 months of sq LNS showed a modest increase in LAZ. However, none of the individual components nor the conjunction of all 3, in the absence of nutritional support, resulted in any significant improvement in height.

The Sanitation Hygiene Infant Nutrition Efficacy (SHINE) Project in Zimbabwe represents the third major intervention trial. ⁶⁵ As described:

Clusters were randomly assigned (1:1:1:1) to standard of care (52 clusters), IYCF (20 g of a small-quantity lipid-based nutrient supplement per day from age 6 to 18 months plus complementary feeding counselling; 53 clusters), WASH (construction of a ventilated improved pit latrine, provision of two handwashing stations, liquid soap, chlorine, and play space plus hygiene counselling; 53 clusters), or IYCF plus WASH (53 clusters).

Analysis of outcome was restricted to the infants born to HIV-negative mothers. Both groups with the nutritional intervention superseded in linear growth the groups without, with 0.16 greater HAZ and a 25% lower stunting prevalence (27% vs 35%). No benefits of the WASH intervention on linear growth were detected.

A very recent systematic review and meta-analysis included 10 selected randomized controlled trials involving interventions with 1, 2, or all 3 of the wash components in 16 473 children (7776 in the interventions and 8687 in the control groups) ⁶⁷ ; it was published too early, however, to capture the contribution from any of the major studies in Bangladesh, ⁶³ Kenya, ⁶⁴ or Zimbabwe ⁶⁵ reviewed earlier. Their interpretation is that the evidence shows efficacy. However, the standard mean differences for effects on HAZ scores among the comparisons of interest were only of the order of 0.14 to 0.22 Z-scores, which is negligible for any corrective effects or full repair of growth impairment.

Catch-up Growth during Adolescence and Puberty

We have sufficiently mapped the timing of the faltering of growth, ^30,31 with the greatest magnitude of relative stature loss (decline in HAZ) occurring in the first 24 months, persisting at slower rate out to 36 months. Leroy et al ⁶⁸ have admonished, moreover, that the advance of impaired linear growth can extend out to 5 years. With the predominance of stunting occurring early, the public health response has been to focus prevention and rehabilitation on the so-called “window of opportunity” of the first thousand days (the interval from conception to the second birthday). The corollary of this pattern of impaired linear growth, moreover, was the dictum that efforts to improve height after this early formative period were fruitless.

Questioning the validity of the early-intervention-only strategy for stunting was a group of researchers from The Gambia, who reviewed growth data from a cohort series from that nation. They suggested that the pubertal growth spurt of adolescence represented another opportunity for recovering the stature of growth restricted children, well beyond the so-called critical window. ⁶⁹ The interpretation of the suggested plasticity and pliability was challenged in a letter to the editor, ⁷⁰ making the important distinction between “relative” growth, as expressed in the HAZ, and “absolute” growth, as measured in actual cm gained. The improvement in adolescence occurs with the former, but is not reflected in actual catch-up in centimeters. This view is reinforced by several published commentaries. ^71,72 The recent findings from a 14-year cohort experience among children from the first year of life in Timor-Leste, one of the nations with the highest stunting prevalence in the world, is illustrative and contributory to the theme. ⁷³ A group of biological anthropologists confirms the absence of any absolute linear catch-up in centimeters as the short-stature children pass through puberty.

The Alternative View to an Imperative for Maximizing Body Size

“Feed it and it will grow” and “the bigger the better” may be applicable to a shade tree in the front yard, but these axioms have been problematic to scholars who reject the supremacy of the normative HAZ in the human context for evaluation of child nutrition in public health. The late Prof David Barker, a British physician and epidemiologist, studied birth weights of individuals born 5 decades or more previously in Sheffield, England, and found an association between lower weights at birth and earlier mortality and morbidity from noncommunicable diseases in adulthood. ⁷⁴ Postulation was made of a conditioning to more efficiently retain and store energy either from a “thrifty gene” (derived through human evolution as a compensation for famine in the Paleolithic lifestyle) or in utero “programming” (an epigenetic mechanism, which instilled energy retention based on signaling of impending, postpartum food scarcity). With his seminal observation and its ramifications, Barker, with his “Barker Hypothesis,” became the progenitor of the Developmental Origins of Health and Disease (DOHaD) school of analysis. ^75,76

Although the seminal observations are based on weight, we can postulate some proportionality through overall body size to slower in utero elongation and shorter postpartum stature. The underlying point of the DOHaD contribution is to posit small size as an appropriate adaptation to life in a deprived environment, based on experimental studies in livestock in which the maternal diet and offspring diets were differentially manipulated with respect to energy restriction or energy abundance. ^77,78 This is known as the “match–mismatch” paradigm in which later health for the smaller offspring of an energy-restricted dam is better over the long run if energy intake remains low. By contrast, having a surfeit of food is detrimental. Conversely, the larger offspring of well-fed dams do better with lifelong food abundance, and they are more sensitive to the adverse consequence of energy restriction. In the human context, small size at birth and slower growth could be adaptations to a harsher environment with scarcer food resources. In a practical sense, however, chances for lifelong food scarcity for an individual are relatively low, as global trends in food security and calorie availability and probability of migration from the site of birth to an area of greater food-energy abundance will intercede. This contemporary reality would tend to negate any protective adaptive benefit of the thrifty gene expression or programming for efficient energy retention. ⁷⁹

Future Projections

A major purpose of this series is to project to the future of research and publication in the area of interest, namely, human linear growth on the occasion of the 40th anniversary of the Food and Nutrition Bulletin (FNB). What will the FNB or other international journals be receiving as submissions in the arena of growth over the next decade and beyond? In the absence of a crystal ball, one must read the tea leaves of the insights from established and emerging research findings as part of the exercise.

Conceptual Departures

On the conceptual front, the literature can be expected to modify prevailing considerations of linear growth retardation toward a more benign affection for the individual and to realize that reduced growth is occurring across any affected population, in line with the perspective of Leroy and Frongillo. ⁸⁰ A series of revised outlooks and emphases, as outlined in Box 1, are likely to emerge in the upcoming literature.

Box 1

Conceptual Considerations Going Forward.

– To de-emphasize inquiry into dietary and nutrient-intake origins of short stature, given the comprehensive evidence of minimal to negligible relative and absolute linear growth recovery from intervention trials with macronutrients and individual and combined micronutrients.

– To progressively replace the use of the cutoff criteria of −2.0 and −3.0 Z-scores for linear growth impairment (or even less stringent variants such as −1.5 and −1.0 Z-scores) with consideration of the distribution of the entire population of interest.

– To disregard short stature, per se, as being a mediator of morbidity and mortality; further inquiry into consequences, including differences in phenotypic characteristics in longer and shorter individuals, should be extended and pursued through adolescence into adulthood and later life.

– To expect hypotheses, designs, and expectations in linear growth-related intervention studies in which biological and public health–relevant increases in stature and stature potential of the order of 10s of centimeters of deficit—rather than a few centimeters—are heralded by the research community.

– To disaggregate the effects of an initial linear retardation in elongation growth to peak adult stature from those of senescent, postgrowth stature reduction in elderly individuals from short-stature settings.

Etiology of poor linear growth

If covering potential dietary deficiencies with supplementation and reducing exposure to microbes with behavioral and structural measures have failed to reverse linear growth retardation, the search for etiologies of short stature will continue. Maternal exposures reflected in offspring growth and modeling of impact of multiple candidate exposures of adversity would be accentuated in these future departures. Box 2 outlines the considerations.

Box 2

Considerations Regarding Adverse External Stimuli and Exposures.

– To continue exploration of remediable maternal factors resulting in up to 30% of stunting appearing at birth as a consequence of reduced elongation during in utero development.

– To explore novel noxious exposures, such as to aflatoxin, ⁸¹ which can be excreted in maternal milk and potentially affect the growth of breast-fed infants.

– To investigate in depth the adverse psychological and emotional factors among disadvantaged women on infant growth ⁸² as transmitted by maternal behavior or through hormonal signaling in breast milk.

– To consider potential synergistic combinations among adverse factors, in which a primary or companion factor(s) must be present for linear growth retardation to be induced.

Advances in anthropometric measurements

Anthropometric measurements of all aspects of growth, including linear growth, have been standard for a decade. The measurement error is well understood such that the interval between successive measures in a given individual or the size of a population sample to reveal a defined difference are routinely applied in the clinic or the field. Digital technology has the capacity to improve comfort, measurement sensitivity, or both. The former is inherently welcome from ethical and subject experience perspectives. However, since larger effects and effect sizes may be of more intrinsic interest and meaning, the real benefits at the sensitivity level for future literature remain to be demonstrated. These considerations are detailed in Box 3.

Box 3

Considerations on Sensitivity and Resolution of Anthropometric Measurements.

– To perfect a process of 3-dimensional ultrasonography that could provide an accurate linear dimension of a fetus in utero, despite its curled (fetal) position. This would allow for monitoring of growth length of the fetus. Debate would ensue, of course, from the International Fetal and Newborn Growth Consortium for the 21 Century (INTERGROWTH 21) perspective on the one hand and the Developmental Origins of Health and Disease (DOHaD) viewpoint on the other as to whether adherence to a “standard linear growth pattern for a fetus” is desirable.

– To develop imaging technology with minimal radiation or white or infrared laser beams to increase accuracy and measurement precision for infant length, especially at the moment of birth.

– To produce an alternative to photographic images from cameras to a shadow projection approach would obtain an accurate and reproducible length/height for children derived from the dimensions of the shadow of the body imaged on a back screen, in a supine or upright position. This could be less cumbersome. Incorporating image-opaque markers at anatomical landmarks into the shadow-image approach, especially to define trunk length as a fraction of the total (trunk–height ratio).

Technological and Investigative Advances

Advances in engineering technology and molecular biology and medicine will impact future literature (Box 4). These will be applied at the interface of immune and hormonal responses at the interface of immunology and endocrinology and their cutting-edge techniques. The entire spectrum of the Omics will advance with refinement of separation techniques and mathematical discrimination. In the field setting, urine, saliva, breast milk, and feces are easy to collect and process, making certain types of Omics-based inquiries increasingly suitable for and accessible to population surveys. Finding a mathematical manner to use continuous—rather than categorical—variable classification would make application consonant with the short-population approaches in the evolving epidemiology. ⁸⁰

Box 4

Considerations Regarding Advancing Technology.

– To understand the point(s) of interruption in the hormonal cascade regulating all aspects of proliferation and metabolism of growth plate chondrocytes at the endocrine, paracrine, and apocrine levels. This might result in insights for potential medicinal intervention to preserve appropriate signals in the face of adverse stimuli.

– To apply advancing chemical analysis and heat-map computer calculation technology will extend the various levels of Omics, such as metabolomics ⁴⁴ and proteomics. ⁸³

– To associate microbiota in the vaginal birth canal, skin of mother and infant, maternal milk, and obviously fecal microbiomes with early linear growth. Publications associating a specific pattern of intestinal microbial flora are growing ^84,85 and this will undoubtedly be one emerging tool to differentiate individuals with differing degrees of shortened stature.

Conclusions

John Waterlow identified the concept of “stunting,” and this has come to dominate the public health and, to a lesser extent the clinical, evaluation of human stature. He further endowed us with the construct of linear growth retardation, a process by which an individual decreases his or her position in the distribution of height toward the stunting criterion. ⁹ Growing to one’s full genetic potential in height is considered an ideal by most of the public health community. We have identified a series of circumstances that constrain individuals from reaching adult life with that achievement.

We cannot, as yet, predict with exactitude the stature with which one or another individual is endowed by his or her genes; we compensate by assessing height in “adequacy” terms in relation to the lower boundaries of collective reference or standard distributions for groups growing under ideal conditions of diet, health, and parenting care, free of the constraints that retard linear growth. Growing poorly is associated with a series of adverse consequences in survival, health, and social functioning. To some extent, these are mechanical consequences of short stature, but to a large extent, they derive from the stressors and insults that produce retardation of growth as such. They would be manifest and problematic across the entirety of societies with high rates of stunting and are not exclusive to those exceeding the arbitrary criterion of the 2.5 percentile on the growth standard.

As discussed here, the mandate in the past, present, and future is to identify constraints impinging on linear growth. Impaired child growth may better be seen through the lenses of an endocrinologic (trophic and stress hormones) and immunologic (proinflammatory mediators) problem than a dietary or nutritional problem. Based on the nonperformance of scientific trials, we are learning to be restrained regarding the promise of nutrient supplementation to reverse linear retardation such that this line of publication is likely to recede. We should be judicious in testing hypotheses that require pharmacological manipulation of the hormonal response, guided by prudent prior appraisal of benefits versus risks.

The vision, therefore, may not be that all populations in all places attain the maximum genetic potential for linear growth but rather that height would be in harmony with the highest well-being possible, given relevant environmental circumstances. Clearly, one of the factors among these circumstances will be drivers of ponderal growth in a world currently facing sufficiency or surfeit of dietary energy, with adverse consequences, but with dietary scarcity and imbalances as future threats from climate change. Scientific publications will likely not lack submissions related to growth, but the FNB’s role may become more about diet in the management of appropriate health in the face of short stature than it has been through the era of the belief in its dietary determination.