It seems to me that your question  effectively deals with two different topics:

i) "time at risk of exposure" (based upon the title) ;

ii) "time at risk of exposure effects" (based upon the text).

About the first one: in many cohort studies, the lenght of time of the exposure and the lenght of time at risk to be exposed are considered to be the same thing and assumed as a proxy of the overall dose (so using their values for estimates of  risk of disease / death). This doen't introduct a substantial bias in cases such as "risk for asbestosis in an asbestos textile factory", where any production and manteinance  worker is certainly exposed from the first day to the last day of work, but can entail a relevant misclassification between "homogeneous" exposure groups when the studied population includes ever-exposed subjects and others who are exposed just once a week or once a month. Clearly, the intensity of exposure is an other parameter to be anyway considered  along with the real duration  of the exposure.

About the risk that the exposure will produce adverse effects: a lag time has to be considered both at the beginnig of the follow-up (e.g. if an exposure started in 1985 is difficult that it could show carcinogenic effects before 1990) and in the last years of the follow-up (e.g: studying cancer mortality in a cohort followed to 2015 the more recent entrants - e.g. workers employed since 2010 - will contribute to the burden of man-years at risk of death  but not to the generation of cancer deaths causally related to the studied exposure).

I think that a good paper has to show to the reader, anytime this is possible, a description of clearly defined exposure profiles, well distinguishing both for duration and intensity to maximize the visibility of existing contrasts, and a separate analysis of the results referred to different exposure profiles and different lag times.

Hi, Ramiro,

May I give my tentative answer here and please tell me if I'm wrong.

Time at risk of exposure is an undetermined duration of time during which an individual is at risk of exposure whereas time of exposure is a measurable definite amount of time of follow-up summated till the exposure occurred.

one of the most comprehensive elucidations is chapter 5, Cohort Studies, in :

Checkoway Pearce and Crawford -Brown: Research Methods in Occupational Epidemiology OUP 1989.

one of the most comprehensive elucidations is chapter 5, Cohort Studies, in :

Checkoway Pearce and Crawford -Brown: Research Methods in Occupational Epidemiology OUP 1989.

Thanks for your contributions. 

I really appreciate it!

HAVE A MERRY CHRISTMAS

Best,

Ramiro

Few points:

1. The time for exposure, you give an example, study cohort consist of patient with DM at the time of hospital admission, it doesn't matter if someone of age 65 had DM for past 20 years or someone of age 55 had it for past 5 years or someone with age 35 had it for past 1 year. If this is not the research hypothesis you are looking at patients with DM at hospital admission. (So time during which exposure occurs- is 20, 5 and 1 year in this case, but if you take that anyone above 30 can have DM then risk years are 30, 25 and 5), the 2nd calculation is not needed, if this is the particular research hypothesis, most cases for how long they had exposure, is counted.

2. Another important things is outcome only should follow exposure, otherwise it is not outcome, please pay attention to this detail.

Thanks and Happy Holidays !

Rahul

I agree with Martín Cañón. The time during which exposure occurs and the time at risk of exposure effects depends on the specific exposure and disease studied. It is the difference between, for example, leukemia followed by atomic bomb explosion and leukemia followed by living near high power electrical transmission area. 

Hi Ramiro

In a cohort study methodology one can define theoretical time at risk of exposure, for example between 1/1/2015 to 12/31/2015. it is a time during which an individual is at risk of exposure exposure.

At the end of follow-up (for example 12/31/2015), time of exposure for each individual exposed is measured. This time depend on the follow-up for each individual included in the cohort. It may varied between 1 day to 365 days in this example. 

It seems to me that your question  effectively deals with two different topics:

i) "time at risk of exposure" (based upon the title) ;

ii) "time at risk of exposure effects" (based upon the text).

About the first one: in many cohort studies, the lenght of time of the exposure and the lenght of time at risk to be exposed are considered to be the same thing and assumed as a proxy of the overall dose (so using their values for estimates of  risk of disease / death). This doen't introduct a substantial bias in cases such as "risk for asbestosis in an asbestos textile factory", where any production and manteinance  worker is certainly exposed from the first day to the last day of work, but can entail a relevant misclassification between "homogeneous" exposure groups when the studied population includes ever-exposed subjects and others who are exposed just once a week or once a month. Clearly, the intensity of exposure is an other parameter to be anyway considered  along with the real duration  of the exposure.

About the risk that the exposure will produce adverse effects: a lag time has to be considered both at the beginnig of the follow-up (e.g. if an exposure started in 1985 is difficult that it could show carcinogenic effects before 1990) and in the last years of the follow-up (e.g: studying cancer mortality in a cohort followed to 2015 the more recent entrants - e.g. workers employed since 2010 - will contribute to the burden of man-years at risk of death  but not to the generation of cancer deaths causally related to the studied exposure).

I think that a good paper has to show to the reader, anytime this is possible, a description of clearly defined exposure profiles, well distinguishing both for duration and intensity to maximize the visibility of existing contrasts, and a separate analysis of the results referred to different exposure profiles and different lag times.