Conclusion

We show a possible relationship between the annual incidence of MOH-diagnosed NS and the annual number of conflict incidents and deaths in preceding years. This supports an association between NS and wartime activities.3 ,4 ,7 If the association is true, there is a latent period of approximately 5–6 years between the peak incidences of conflict/deaths and NS cases, a disease that affects children tightly clustered around the ages of 5–15 years of age.1 Civil conflict was active in Acholi Sub-Region when the first reports of NS appeared in 1998, with thousands of residents fleeing villages in Gulu District in July 1996 after a wave of LRA violence. Cases of NS declined between 2008 and 2011 in line with the cessation of the war and signing of a peace agreement between the Ugandan government and LRA in February 2008.

There are numerous reasons why conflict theoretically could be associated with NS. Exposure to warfare chemicals is readily posited but dismissed as highly improbable given the known neurotoxic properties of such substances, none of which causes repetitive head nodding from atonic seizures, let alone a progressive seizure disorder. Sudanese communities affected heavily by NS also experienced war and displacement but reported no symptoms consistent with neurotoxic exposures when questioned in 2002.9 Moreover, Tanzanian children with signs consistent with NS acquired the brain disease in the absence of war or civil conflict.19 Severe psychological trauma resulting from the sight of injury and death, and the personal fear associated therewith, have also been advanced as causal of ‘Psychological NS’, but populations in other war zones have not succumbed to a comparable illness. Furthermore, there are no known reports of NS among the thousands of children who were abducted by the LRA and forced to conduct atrocities. Documentation is available of incidents described as ‘human rights abuses by the LRA’ and ‘human rights violations by Ugandan Government forces’.2 ,6

There is also an apparent relationship between the peaks of NS cases in Kitgum District and earlier peak influxes of households into IDP camps. The 1997 peak influx is followed 7 years later by an elevated number of new NS cases in 2004 (2003–2005), and the 2003 large influx of households anticipates a larger peak in new NS cases 5 years later in 2008. Conditions in the IDP camps were exceptionally poor, with overcrowding, violence, food insecurity and high potential for disease transmission. In 2005, a government survey of Kitgum estimated an IDP population of 310 111 persons, 21% of whom were under 5 years of age. At the time of the survey, over 66% of children were reported to have been ill sometime in the previous 2 weeks. Crude mortality rates were ∼2 deaths per 10 000 per day and double that rate for children under the age of 5 years. Top self-reported causes of death in IDP camps were malaria/fever (34.7%), AIDS (15.1%) and violence (10.5%). An estimated 1216 persons were killed and an additional 304 (mostly children) abducted during the first half of 2005. Water was obtained from protected sources but water intake was low and waiting times high. Infant feeding practices were poor and, for children under the age of 5 years, the traditional disease concept of two lango or gimiru, a combination of oral thrush, malnutrition and diarrhoea, was the second most commonly reported cause of death.20 The World Food Programme provided food because ability to grow crops was limited due to security concerns.15 Food quality was often extremely poor and, under normal conditions, would have been considered inedible. In some cases, security was so weak that food deliveries did not occur, with resulting hunger and malnutrition. Insecurity in IDP camps led to a migration of children (‘night commuters’) to seek shelter in Kitgum hospitals, schools, municipal buildings, verandahs, parking lots and other open spaces.21 Unfortunately, there are no data on the incidence of NS among night commuters.

The major limitation of this study is the accuracy of reports of the number of deaths, displaced persons and NS cases. Cited data are drawn from reports prepared by the Uganda MOH, international bodies and non-profit and other organisations. Reason for caution is illustrated by the report of a massacre on 7–12 January 1997 when “up to 412 civilians were killed by armed attackers in northwest Kitgum subcounties of Lokung and Palabek and in nearby areas”.6 This contrasts with the report of 128 deaths in Kitgum throughout 1997 recorded in the ACLED database.16 ACLED death data are derived conservatively from a variety of sources, including research publications and reports from humanitarian agencies and local media. If a report mentions hundreds of deaths, ACLED records the number as ‘one hundred’, which might explain the discrepancy.

Taken in concert, there appears to be a reasonable correlation between IDP camp intake peaks and delayed NS peaks, which in turn show correspondence with prior conflict incidents and deaths. Since the disease is clearly associated with prevailing environmental factors, but apparently not with polluted drinking water, suspicion falls on infectious and/or nutritional factors, including food type, quality, spoilage or chemical contamination. This meshes with 2002 findings from then-southern Sudan where NS was associated with onchocerciasis and prevalent in sessile communities dependent for food on crops grown in small gardens but absent in cattle herders, the latter having access to meat, blood and milk.9 When affected South Sudanese were questioned by CDC investigators as to the source of their food, an unexplained association emerged between NS and use of garden food rather than food purchased from the market or provided by the World Food Programme (US Centers for Disease Control, unpublished data). Additionally, a link between food available in IDP camps and NS would explain the apparent absence of this brain disorder in abducted children: while they served as fighters, porters, sex slaves and baby-sitters, they likely had better access to food, including freshly killed human body parts. Thus, the nature of any association between NS, nutrition and materials used for food, in addition to infection with nematode microfilariae (particularly Onchocerca volvulus) and war-related neuropsychological factors, would appear worthy of focused investigation.