Birth Weight-Specific Mortality

Figure 3: Weight-specific neonatal mortality for the US, 1950 and 1998

Birth weight by itself would not have caught the attention of epidemiologists were it not for its association with infant mortality. The relation of mortality to birth weight has a highly distinctive pattern (Fig. 3). Mortality ranges more than 100-fold across the spectrum of birth weights. (The figure shows mortality on a log scale in order to accommodate this huge range.)

The reverse-J pattern of weight-specific mortality is found in all populations, and occurs with fetal mortality (stillbirths) and with neonatal or infant mortality (Wilcox 1983b). While high mortality among small babies is one of the chief justifications for studying LBW, note the continuous rise of mortality with lower weight. The mortality curve provides no particular justification for 2500 grams as the criterion for risk.

Stability of the curve

One fundamental aspect of birth weight-specific mortality is the constancy of its shape. US neonatal mortality fell 75% between 1950 and 1998 (from 20 to 5 per thousand) with no change in the basic shape of the curve (Fig. 3). This constancy in the shape of this curve over time may be surprising, since much of the improvement in US infant survival over the past fifty years is assumed to be due to better medical care for very small babies. The absolute decline in mortality has indeed been greatest among small infants. However, the relative decline in mortality has been fairly uniform across all birth weights (a constant distance on the log scale), with least change at the smallest weights.

The general contrast seen between these two mortality curves is typical. The crucial difference in birth-weight-specific mortality between any two groups is usually the height of the mortality curves, rather than their shape (more details in Wilcox 1983b).

There is one more feature of the mortality curve that becomes apparent only when the curve is considered in relation to the birth weight distribution. This feature is the foundation for the next section, The Wilcox-Russell hypothesis. (If you've already read "The Wilcox-Russell hypothesis", then go directly to The Analysis of Infant Mortality.)

Why study birth weight?

A short history of low birth weight

The low birth weight paradox

Frequency distribution of birth weight

Birth weight specific mortality

The Wilcox-Russell hypothesis

The analysis of infant mortality

Beyond low birth weight

Print options

Bibliography