Adaptive variability in the duration of critical windows of plasticity: Implications for the programming of obesity

Abstract

Developmental plasticity underlies widespread associations between early-life exposures and many components of adult phenotype, including the risk of chronic diseases. Humans take almost two decades to reach reproductive maturity, and yet the 'critical windows' of physiological sensitivity that confer developmental plasticity tend to close during fetal life or infancy. While several explanations for lengthy human maturation have been offered, the brevity of physiological plasticity has received less attention. I argue that offspring plasticity is only viable within the niche of maternal care, and that as this protection is withdrawn, the offspring is obliged to canalize many developmental traits in order to minimize environmental disruptions. The schedule of maternal care may therefore shape the duration of critical windows, and since the duration of this care is subject to parent-offspring conflict, the resolution of this conflict may shape the duration of critical windows. This perspective may help understand (i) why windows close at different times for different traits, and (ii) why the duration of critical windows may vary across human populations. The issue is explored in relation to population differences in the association between infant weight gain and later body composition. The occupation of more stable environments by western populations may have favoured earlier closure of the critical window during which growth in lean mass is sensitive to nutritional intake. This may paradoxically have elevated the risk of obesity following rapid infant weight gain in such populations.

Keywords: adaptation; critical window; developmental plasticity; growth; obesity; parent–offspring conflict.

Figure 1.

Schematic diagram illustrating two different ‘allocation games’ in which each of the mother and the offspring optimize their inclusive fitness. In the first game, the mother optimizes her allocation of parental investment (PI) across competing offspring (O₁–O₄). In the second game, which is sensitive to the first game, each offspring optimizes its allocation of that investment between competing phenotypic traits, such as muscle mass (M), vital organ mass (V) and energy stores in fat (F). These traits also represent allocations across life history functions (maintenance, growth, reproduction and immune function)

Figure 2.

Schematic diagram illustrating contrasting levels of stochasticity in four different ecological parameters. Some parameters such as altitude are relatively stable, others such as temperature may change relatively slowly, whereas others such as infection risk may alter suddenly

Figure 3.

Schematic diagram illustrating how the offspring can resolve the complexity of multiple ecological cues by sampling stable components of maternal metabolism, relying on maternal homeostatic systems to smooth over ecological disruptions. This means that when the offspring is weaned, it must either address all the ecological stochasticity, or develop its own buffering, which it can achieve through closing plasticity and canalizing traits

Figure 4.

Schematic diagram illustrating how the mother and the offspring have different optima for the duration of maternal care, according to the maternal fitness function (B), based on Haig’s model of optimal pregnancy duration [39]. The cost of continuing investment in the current offspring is given by C for the mother, but by rC for the offspring, which discounts some of these costs as it shares only 50% of its genes with its mother. Maternal fitness (M) is maximized by B–C, while offspring fitness (O) is maximized by B–rC. According to ecological conditions, the entire fitness function B may be shifted earlier or later, or the shape of the function may change, as shown in scenarios (a)–(d) and discussed in the article text