Insulin loses potency when stored at high temperatures but refrigeration is not available in many low-income countries. So paediatric endocrinologist Dr Graham Ogle and colleagues decided to enter a range of cooling devices typically used across the world into their own Olympics.

The 14 devices from Sudan, Ethiopia, Tanzania, Mali, India, Pakistan and Haiti included clay pots of various designs, vegetable gourd, goat skin, and a bucket filled with sand. They also tested two commercially available “Frio” brand cooling bags and tested one prone, and one lying flat.

Internal and ambient temperature and ambient humidity were measured simultaneously every five minutes for 88 hours in a well-ventilated internal room in Kartoum, where the average temperature was 31C and humidity was 32%.

However, the Mali pots kept breaking in transit so they were studied in Bamako where it was 32.9C and 39.8% humidity, Ogle told delegates.

The research team found that all devices reduced temperature by an average of 2.7 – 8.3C depending on the device. When expressed as % maximal cooling the efficacy ranged from 20.5% to 71.3%.

“Some pots were a little bit effective, some pots were very effective but they were all better than doing nothing,”  Ogle said.

But the four outstanding performers, the gold medal winners if you like, were the standard cooling wallet, two of the devices from Sudan and one of the pots from Ethiopia.

These devices were very effective at reducing temperatures by 7 or 8 C on average, with efficacy in the 60-70 percents, he said.

However, the absolute winner was actually the goat skin because of its large evaporative area, he said.

The results showed that devices constructed from traditional materials were effective in reducing insulin storage temperatures.

“Even in the hottest climates storage temperatures at or close to room temperature should be achievable”.

Assuming human insulin is at least as stable as animal derived insulin this should permit safe storage outside a refrigerator for up to five months with negligible loss in potency (≤2%) based on data from Pingel and Volund (1972) and Storvick and Henry (1968), he said.

However new studies on the stability of human and analog insulins were needed to determine when these pots need to be used, he concluded.