First, it is necessary to know how is ECG with P-wave, QRS-complex, and T-wave generated. ECG is recording of a sum of action potential propagation through a given region of the heart. The very important part is the conductive system of the heart that usually propagates an electric signal very fast.
PQ interval defines how long it takes to electric potential to propagate from SA-node to ventricles. QRS part is very fast, between 0.08 and 0.1 sec. QRS reflects the propagation of the potential through His-Purkinje part of the conductive system plus contraction of ventricles. ST-segment is given by time of depolarization of ventricles.
From said information, it is obvious that any disease, infection, ischemia, infection, ion disbalance, etc. can change the duration of different segments of ECG recording. Actually, those changes can and often do serve as a navigation tool towards a diagnosis of heart disease.
The length of various segments of ECG recording can be modulated by physiological changes of the body or the heart itself, by the cardiovascular center, adrenal hormone, thyroid hormone, etc.
It is very difficult or impossible to tell only from an ECG recording what is changed in a given patient. You must know the laboratory results as well to discern various pathologist from physiological changes.
Basically, in general, it is very difficult to make any conclusion from ECG only. Nevertheless, there is ongoing research based on complex systems measures that are trying to discern different modes of heart responses and diseases.
You can, for example, use ECG to discern different positions of patients tilted from vertical, standing, position to lying position with elevated legs.
The area of this research is quite broad and deep. Nonspecialists get easily lost there as might be unclear for them what is the important result and what not.
There are must be a lot of so-called contributors to the variation. Probably most significant one is heart rate. Ideally heart rate can be corrected so that the relation with other factors can be clarified. However, how to correct these intervals by heart rate is a well debate.
In addition, any disease that comprises the myocardium such as ischaemic heart disease or non-ischaemic cardiomyopathy will have an effect on conduction velocity and the subsequent intervals on an ECG (depending on the site of disease).
Not a single mechanism of our body is fixed. The variations can be physiological or pathological. Since the ECG is the recording of electrical activity of heart, the factors responsible would be either related to the origin and conduction of electrical activities in heart, innervation of heart, changes in ionic composition of blood, temperature or hormonal effects.
Just thought even physiologically there are a couple of factors that impact those intervals. However, if we could find the most significant impactor(s), using ways to correct them when quantifying heart intervals, it might help to bring a clearer map when applying those intervals in pathologies.