Table 1.

Models investigating associations between environmental variation and child mortality

No.	Model	E		Spring temperature
		AIC	ΔAIC	AIC	ΔAIC
0	BASE	11269.96	0.00	11269.96	0.00
1	BASE + current	11171.60	−98.36	11266.15	−3.81
2	BASE + early-life	11266.26	−3.70	11261.23	−8.73
3	BASE + early-life + current	11173.13	−96.83	11260.05	−9.91
4	BASE + early-life:current	11174.38	−95.58	11259.12	−10.84
5	BASE + social:current	11172.64	−97.32	11269.33	−0.63
6	BASE + social:early-life	11268.85	−1.11	11260.90	−9.06
7	BASE + age:current	11172.50	−97.46	11265.90	−4.06
8	BASE + age:early-life	11260.88	−9.08	11258.65	−11.31
9	BASE + age:(current + early-life)	–	–	11254.75	−15.21
10	BASE + current + age:early-life	11172.93	−97.03	–	–
11	BASE + age:(current + early-life) + Current:early-life	–	–	11255.17	−14.79

The table presents a comparison of binomial generalized linear mixed-effects models testing the effects of early-life and current environmental conditions and their interactions on mortality in children (aged 1–15 years). The numbered models, where current and early-life effects were either E (parish-wide infant mortality) or spring temperature, were compared with each other using AIC values. The best-supported models for each of E and spring temperature have the lowest AIC value and are shown in bold; the ΔAIC values are shown relative to the base model. ‘+’ indicates additional terms in the model structure, while ‘X:X’ indicates an interaction term between X and X.