Clear definitions of, and rationale for defining variables as

explanatory
confounding (if required)
effect modifying variables (if required)

and rationale if the form they will take will be different from the form in which they were collected.

Explanatory variable: it is any variable that explains the response variable often called an independent variable or predictor variable. the independent variable in the study.
For example in this study and according the stated hypothesis are: the duration of sun light exposure, season and sun protection measures.

Confounding: is defined as an extraneous variable in a statistical model that correlates (positively or negatively) with both the dependent variable and the independent variable.

In this study confounding variables can be occupational sun exposure, dietary calcium intake, dietary vitamin D supplement, sunscreen use, wearing a hat outdoors and typical outdoor activities. all these variables acts as confounders and can have a negative or positive effect on the dependent variable (Vitamin D status) and the independent variable (duration of sun exposure, season and sun protection measures).

In assessing the first hypothesis related to the status of vitamin D and the effect of the duration of sunlight exposure. The stated confounders above (occupational sun exposure and typical outdoor activates) are not confounder and they take the form of the explanatory variable that will be tested in this hypothesis.

In assessing the second hypothesis related to the status of vitamin D and season.
In assessing the third hypothesis related to the status of vitamin D and sun protection measures. The stated confounders above (sunscreen use and wearing a hat outdoors) are the explanatory variables that will be tested for the stated hypothesis. However, they act as confounders if we are testing the first hypothesis for example.

Effect modifying variables: a variable that differentially (positively and negatively) modifies the observed effect of a risk factor on vitamin D status. Different groups have different risk estimates when effect modification is present. In this study skin colour (dark skin and fair complexion) can be the effect modifier variable. Easiest way to examine the nature of effect modifier is by stratifying the data by skin colour. Also by eyeball graphs we can know if skin colour has a modifying effect on vitamin D status.