In 2010, a landslide in Bududa, Eastern Uganda, killed about 350 people and nearly 1000 affected households were resettled in Kiryandongo,
Western Uganda. A cross-sectional survey assessed household food insecurity and diet diversity among 1078 affected and controls. In Bududa,
the affected had a lower adjusted mean score of food insecurity than controls – 9·2 (SE 0·4) v. 12·3 (SE 0·4) (P< 0·01) – but higher diet
diversity score (DDS) – 7·1 (SE 0·1) v. 5·9 (SE 0·1) (P<0·01). On controlling for disaster and covariates, recipients of relief food had higher food
insecurity – 12·0 (SE 0·6) v. 10·4 (SE 0·3) (P=0·02) – whereas farmers had higher DDS – 6·6 (SE 0·2) v. 5·6 (SE 0·3) (P <0·01). Household size
increased the likelihood of food insecurity (OR 1·15; 95 % CI 1·00, 1·32; P <0·05) but reduced DDS (OR 0·93; 95 % CI 0·87, <1·00; P =0·04).
Low DDS was more likely in disaster affected (OR 4·22; 95 % CI 2·65, 6·72; P<0·01) and farmers (OR 2·52; 95 % CI 1·37, 4·64; P<0·01). In
Kiryandongo, affected households had higher food insecurity – 12·3 (SE 0·8) v. 2·6 (SE 0·8) (P <0·01) – but lower DDS – 5·8 (SE 0·3) v. 7·0 (SE 0·3)
(P =0·02). The latter reduced with increased age (OR 0·99; 95 % CI 0·97, 1·00; P <0·05), lowest education (OR 0·54; 95 % CI 0·31, 0·93;
P=0·03), farmers (OR 0·59; 95 % CI 0·35, 0·98; P =0·04) and asset ownership (OR 0·56; 95 % CI 0·39, 0·81; P <0·01). Addressing social
protection could mitigate food insecurity.