past or future?

epidemiological study of some description is what you'll need to do.

Thank you for that important addition Philip.

YOUR CAN USED PERCEIVED CHANGES IN THE ENVIRONMENT LIKE  DISEASE OUTBREAK  OVER TIME USING SOCIAL RESEARCH APPROACHES SUCH AS LIKERT-TYPE SCALE TO RANK THE PERCEIVED CHANGES AND IMPACT ON  CHILD MORTALITY. 

Thank you Umazi. Perceived changes in the environment, yes I partly agree, maybe you can enlighten me more whether such a scale exists for climatic changes. Building on your suggestion, one could use weather-related calamities such as the current FLOODS IN MALAWI or other drastic changes such as DECLINE of agricultural(plants and livestock) productivity, dried-up lakes have in East Africa or declined in levels of L. victoria affecting Hydro electric power generation e.t.c, extreme drought in northern Kenya....Using these and others as indicators of climate change it maybe possible to relate to under-5 mortality. Sounds a vast study though.

I would pay attention to the three major ocean cycles that impact that area the PDO, AMO and IOBD. I may be writing next Monday on the impact of the location of the Aleutian Low on drought in that area. You can google me at Sig Silber Econointersect.com and my weekly weather and climate report gets posted there late Monday nights and you will be able to find out what I have to say and if you go to Page II of my report you will find more information that may be of value.

For the next 50 years or so the ocean cycles will probably have more of an impact than Global Warming but how you translate that into under 5 mortality is another question. One has to identify the causes of infant mortality.

I am 100% certain that you will not be able to attribute any part of infant mortality to climate change not that there is not a connection but because

A. 15 years is less than the cycle length of ocean cycles.

B. The accuracy of climate models is not sufficient to estimate the delta of climate during that 15 year period.

C. There are probably many factors that determine infant mortality that are more significant than temperature, precipitation, storms etc.

Thank you Silber. I am going through your website, "Sig Silber Econointersect.com" and there seem to be very useful information. I would ask you a few questions later when I am done. Many thanks

I think my email address is posted with Research Gate. I really need to add it to my bio at Econonintersect.com. I am planning to address the impact of the Aleutian Low this Monday if I get that far. The PDO is possible changing phase and amazingly the PDO impacts East Africa. All three oceans do. This is important because any time series of epidemiological data most likely will be correlated with the state of these three oceans. But I am not sure that one can separate it out statistically. A lot of degrees of freedom. But if you come up with something I want to include it in my book.

Quick note: It might help to formulate hypotheses to be tested for instance if due to temperature rise the frontier of certain diseases has moved up into highlands, e;g; malaria in the Ethiopian highlands.

another quick note: subjective recall questions going back in time are extremely unreliable.

Thank you Gideon for that valuable contribution, I largely agree with with you on your second quick note.