96

■Bio-mathematics, Statistics and Nano-Technologies: Mosquito Control Strategies

models were fitted to data from many sites, others were fitted to single-site data. In as much

as fitting to single-site data does not necessarily limit relevance of the models themselves,

it is difficult to exclude the possibility that they are only applicable to the specific context of

the studied data-set, and predictions may not be extrapolated to different areas/scenarios.

Another interesting thing to do considering the pitfalls of both deterministic and agent-

based modeling approaches is to tactically combine the two, so as to capture the desired

population size, heterogeneity and study period [196].

5.3

DISCUSSION

Naturally acquired immunity against malaria plays an important role in the reduction

of the health burden posed by malaria, as it protects numerous people who are routinely ex-

posed to the infection against severe disease and death. In this review, the principal features

of NAI to malaria have been reevaluated in terms of how different modellers perceive them

and how the ideas have evolved. However, some of the underlying mechanisms behind this

protection are not yet clear. Owing to the need to understand NAI on an individual level,

modelling strategies have gradually evolved from deterministic compartmental models to

stochastic ABM which build on more advanced knowledge base than the purely determin-

istic models that preceded them for decades. The ABMs came in nicely but quite late in

the evolution of malaria models, as needed to cope with issues like interactive variability

in parasite and host dynamics that were first identified with deterministic model fits to data

sets for single individuals. Thus, more recent idea of modelling suggests that immunity in-

dicators should be functions of parasite density and diagnostic detection sensitivity rather

than binary factors.

Contrary to the suggestion that protective immunity is acquired after one or two infec-

tions, more recent studies demonstrated that immunity to severe malaria is more gradually

acquired with exposure. While it is believed that sterile immunity against infection is never

completely achieved, chronic, subacute carriage is the rule among adults. Such chronic car-

riers are less vulnerable to the parasite infection and have partially reduced infectivity to

mosquitos. Given these dynamics, it is easy to understand how malaria transmission is so

robust and difficult to eliminate. This review considers the opinion that adults in endemic

areas are completely immune and do not contribute considerably to the infection reservoir

as unreliable. Thus, recent models such as those of the OpenMalaria models, are adopting

the concept of modeling infectivity in relation to the asexual parasite densities of hosts, and

also linking the infectivity of human to that of mosquito infection. There is still a need for

improved models explicitly describing the impacts of age and exposure in the infectivity

of humans to mosquitos in endemic areas.

Furthermore, the mechanism of control of malaria parasites upon reinfection or su-

perinfection in semi-immune individuals, has been addressed. The convincing account is

that superinfections or infections that share antigens recognized by the already existing

immune responses can be controlled such that it would not be severe. There is very little

evidence that NAI has much impact on the duration of infections since models that have

evaluated this concept found no general trend with age. It is accepted that naturally im-