March 30, 2016

expert reaction to study modelling the seasonal risk of Lyme disease in Scotland

Lyme disease is commonly spread through the bites of infected ticks and a group of researchers have published their work in the Journal of the Royal Society Interface which models risk of Lyme disease infection in Scotland based on temperature and geography.

Prof. Sarah Randolph, tick expert & retired Professor of Parasite Ecology, University of Oxford, said:

“In terms of any message from this paper of relevance to public health issues of Lyme disease risk, the following significant caveats should be taken into account.

“Specific predictions as model outputs are only as good as the model inputs (assumptions). The authors state clearly that this model is highly sensitive to the parameters for which the empirical data are least reliable (see section 3.1.3 model sensitivity and discussion). In this case, as in many other cases, models are therefore better used as tools to explore missing information and the interactions of the many elements of this highly complex system, rather than to supply reliable predictions of changing risk.

“The temperature data and forecasts used to drive the model are extremely crude – “mean weekly temperature surfaces” (see section 2 materials and methods). What temperature stream was used from which to derive these weekly means – hourly records, daily means, daily max, daily min, etc.? This matters a great deal for any mechanistic biological model of tick seasonal dynamics, because ticks are effectively ‘inert’ at low temperatures so that means that include low temperatures are biologically incorrect. Ticks are also ‘inert’ (in diapause) over the winter, so annual average temperatures are also incorrect inputs.  These points have been published many times.

“Predictions based on warming of +1, +2 and +3^oC far exceed anything of relevance in the near future.

“I have not had access to the supplementary material in which the models are described in detail (the devil is always in the detail). One major assumption that needs to be corrected, however, is that immature tick stages do not feed on, and are therefore not transported by, large host animals. Tick stages are distributed at different levels in the vegetation from where they seek hosts (‘quest’), larvae close to the ground, nymphs higher up and adults higher still. Adults and to a large extent nymphs do not contact and therefore feed on small rodents and small birds that live below their questing level – i.e. these tick stages are above the heads of small hosts. Larvae feed on anything whose feet touch the ground, including deer and sheep that both feed larvae in their hundreds (thousands, even) compared with tens of larvae on rodents. These points have been published many times.”

‘Modelling the seasonality of Lyme disease risk and the potential impacts of a warming climate within the heterogeneous landscapes of Scotland’ by Li et al. published in Journal of the Royal Society Interface on Wednesday 30th March. 

Declared interests

Prof. Sarah Randolph: No conflicts of interest