ABSTRACT
Large parts of Africa are prone to malaria epidemics. Advance epidemic warning would give health services an opportunity to prepare. Because malaria transmission is largely limited by climate, climate-based epidemic warning systems are a real possibility. To develop and test such a system, good long-term malaria and climate data are needed. In KwaZulu-Natal (KZN), South Africa, 30 years of confirmed malaria case data provide a unique opportunity to examine short- and long-term trends. We analysed seasonal case totals and seasonal changes in cases (both log-transformed) against a range of climatic indicators obtained from three weather stations in the highest malaria incidence districts, using linear regression analysis. Seasonal changes in case numbers (delta log cases, dlc) were significantly associated with several climate variables. The two most significant ones were mean maximum daily temperatures from January to October of the preceding season (n=30, r2=0.364, P=0.0004) and total rainfall during the current summer months of November-March (n=30, r2=0.282, P=0.003). These two variables, when entered into the same regression model, together explained 49.7% of the total variation in dlc. We found no evidence of association between case totals and climate. In KZN, where malaria control operations are intense, climate appears to drive the interannual variation of malaria incidence, but not its overall level. The accompanying paper provides evidence that overall levels are associated with non-climatic factors such as drug resistance and possibly HIV prevalence.
Subject(s)
Climate , Disease Outbreaks/statistics & numerical data , Malaria/epidemiology , Humans , Incidence , Linear Models , Malaria/etiology , Malaria/transmission , Rain , Risk Factors , Seasons , South Africa/epidemiology , TemperatureABSTRACT
Seven borderline personality and ten eating disorder patients were compared using demographic, descriptive/behavioral, and ego function measures on admission to an inpatient unit. Demographic and descriptive data analysis revealed areas of overlap, consistent with other reports of co-morbidity between the two groups. Ratings of intrapsychic function were made using Bellak's Ego Function Assessment profiles, and showed that the two groups shared deficits in the areas of impulse/affect control and sense of identity, but had otherwise easily distinguishable profiles. The authors feel that studying ego functioning is a useful way of assessing co-morbidity in borderline personality and eating disorder patients.
