Physiology behind EKG effects of hypokalemia and hyperkalemia

In terms of myocardium cell potential:
hyperkalemia = depolarized resting potential, but Decreased excitability.
hypokalemia = hyperpolarized resting potential, but Increased excitability

In terms of EKG surface potential:
hyperkalemia = mixed unreliable effects on QT (shortened -> prolonged with worsening hyperK; see discussion in comments below), peaked T wave, wide QRS, ST depression
hypokalemia = prolonged QT, flat Twaves, U waves

The confusion comes from the fact that these action potentials are a lot more complicated than what there is time to cover in most Physiology courses.  Sorry for the long post below but it is necessary if you don't want to memorize the above and want to answer these "why" questions.

3 important concepts:


1) Some Na Channels will begin opening while the RMP is below "threshold potential".

Threshold voltage of a cell is not a fixed magical voltage level where all of the sodium channels suddenly turn on.  Sodium channels are voltage-gated and turn on probabilistically.  The more positive the membrane voltage, the higher the probability a sodium channel opens.  So even at very negative voltages, most sodium channels stay closed but occasionally a sodium channel may open briefly.  These brief openings conduct such a small current that they have very little effect on the RMP.  At these very negative RMP values, the only channels that are open significantly are "leak" K channels and thus the RMP (which is kind of a weighted average of the reversal potentials of all the conducting ions) remains close to the K reversal potential.  Now the (maybe to some) surprising part: if the membrane voltage depolarizes a little bit but is still significantly below (more negative to) normal "threshold potential" of a cell, some Na channels will start opening (although most stay closed). Again this is because Na channels are probabilistic and voltage dependent and start opening as the membrane starts to depolarize, long before you get to "threshold".  Since your membrane voltage has depolarized some, the probability of Na channels opening is no longer close to 0 and some of them will start opening (even though you are below "threshold").


2) Threshold voltage is kind of a moving target that depends on the size of Na and K currents relative to each other.

At sub-threshold depolarizations of membrane voltage, the opening of some Na channels does pull the membrane voltage higher (depolarizes) but as long as there is significant leak K current pulling the membrane voltage towards hyperpolarization, the Na current is sufficiently offset by the K current and the membrane voltage remains "sub-threshold".  The "threshold" phenomenon occurs when membrane voltage is elevated enough that a large enough number of Na channels open to overcome the K channels.  The membrane voltage is like a weighted average so as Na channels open, voltage is pulled towards the Na reversal potential (positive) from its resting state close to the K reversal potential (negative).  If enough Na channels open to overcome the hyperpolarizing influence of the K channels, the effect becomes a positive feedback loop: as Na channels open, the membrane voltage becomes more positive and as it becomes more positive, more Na channels open.  This Na channel positive feedback causes the rapid phase 0 depolarization of the action potential.  It would run away unchecked if not for the fast inactivation of Na channels and a second set of K channels the voltage gated K channels (as opposed to the leak K channels).  These K channels' voltage dependence is shifted more positive compared to the voltage gated Na channels and so they turn on later at more positive voltages, helping to drive membrane potential back to ward the negative reversal potential of K.


3) Sodium channels inactivate after opening.

In activation is a process that makes Na channels non-conducting after being open.  We talked about fast inactivation of Na channels (ball and chain mechanism).  There is also a slow inactivation (due to more overall conformational change of the Na channel protein).  Hyperpolarization relieves inactivation and resets the Na channels making them ready to fire.  At depolarized voltages, Na channels can open and some can go on to become inactivated.  This effect is key to the physiological effects of hyper/hypo-kalemia.


What happens in Hyperkalemia:

Reversal potential of K is shifted positive. Driving force for K to leave the cell via Leak K channels is decreased so there is less "leak" K current and the RMP is depolarized compared to normal.  The more positive RMP means more Na channels will randomly open and more of them will become inactivated.  This means fewer Na channels are available to respond by opening when full depolarization occurs.  This effectively ALSO RAISES THE THRESHOLD POTENTIAL.  Because there are fewer Na channels available to conduct current, you need a greater voltage to get enough Na current to overcome the effect of the K channels.  Thus the "threshold" voltage, the voltage where Na wins over K, is now higher.  Early in mild hyperkalemia, the effect on K currents and RMP is more important than the effect on Na channels and Threshold potential, so overall there is a small increase in excitability as the RMP is slightly closer to threshold potential.  Soon however, the hyperkalemia's effect on the Na channel inactivation dominates and you get an overall DECREASE in excitability due to decreased availability of Na channels for phase 0.  This results in a slow phase 0 depolarization and a slow or widened QRS complex. To reiterate, significant hyperkalemia will cause decreased excitability.


What happens in Hypokalemia:

Essentially reverse of above.  Low serum K = shift negative in reversal potential of K = increased K driving force = more leak K current and hyperpolarized RMP.  This causes increased availability of Na channels for firing due to relief of inactivation and actually shifts threshold potential towards the negative.  Thus there is NOT an increase in the "distance" of the RMP from threshold.  Increased Na channels = faster depolarization and lower threshold voltage = INCREASED excitability.


Effect on T-waves:

The flat/peaked T-waves are a little more complex: the voltage dependent K channel proteins (as opposed to the leak K channels which are simpler) like to have a potassium ion sitting in the pore near the extracellular end of the channel.  This helps keep them conducting and prevents them from collapsing shut.  In hypokalemia, there is decreased extracellular K, so more of the voltage gated K channels are collapsed shut.  In hyperkalemia, very few of these voltage gated K channels are collapsed and many more of them are ready to fire when the cell depolarizes sufficiently.  Thus in hyperkalemia, more K current after depolarization (ie after phase 0, so during phase 2 and 3) means, phase 2 is shortened and phase 3 is faster.  This results in shortened APD, shortened QT and peaked T wave.  Opposite in hypokalemia - collapsed K channels = less K current in phase 2 and 3 = long phase 2 and slow phase 3 = prolonged QT and flat T wave.


SA node: 

HR changes are not a typically discussed as part of the cardiac syndrome caused by hyper/hypo kalemia. Take the following with a grain of salt. This is what makes sense but really it is much more complex than our discussion. 
I did want to mention regarding atrial phase 4 depolarization that although a Sodium current is involved and K currents also become important in phase 4 (directly and in response to Acetylcholine) the channels involved are different in terms of both the Na and K channels than those mentioned earlier so do not confuse them. Phase 4 is a balance of Na and K currents where there is normally enough Na current to slowly overcome the K current and depolarize to threshold. These SA node phase 4 Na channels open at much lower voltages (are in fact hyperpolarization activated sodium channels...very different) and have a very different voltage dependence than the phase 0 Na channels. Here the important K channels are also different but similar to some of the potassium channels listed above, they also like to have a potassium ion near the extracellular pore to keep them from collapsing shut. Thus their activity is also directly affected by the presence of extracellular potassium acting directly on the channel protein conformation independent of the "driving force" effect. Therefore the conductance of potassium falls with hypokalemia even though the concentration gradient is increasing. The effect on conductance is greater than the effect on the driving force and the net effect is less potassium current and FASTER phase 4 depolarization leading to faster heartrate. The Na channels in the SA node DO NOT INACTIVATE with slow inactivation at subthreshold voltages and thus we don't have to worry about them the way we did before. Hyperkalemia would do the opposite, more K channels conducting current which fights the outward sodium current and therefore slows down phase 4 depolarization and therefore slows down heartrate. 

Interestingly, Acetylcholine (from vagus nerve endings) acts to activate an additional class of atrial potassium channels which are activated by ACh and these additional K channels can boost the potassium current to hyperpolarize the SA node and slow down phase 4 leading to the chronotropic (HR slowing) effect of the vagus nerve. 



Weakness:
In terms of weakness, some hand waving is necessary.  Decreased excitability will obviously present as weakness but increased excitablity can also cause weakness through "fatigue"-type effects.  Furthermore, the K changes will have myriad changes on other ion transporters directly and indirectly by changing other ionic gradients. Both hyper and hypo kalemia cause problems in conduction. You need correct amount of K for conduction, not excessive nor too little (that is why they both cause pathological ECG changes). The discussion above is for myocardium and EKG findings.  Behavior may be somewhat different for nerve and skeletal muscle.