EDIT: I've been putting more thought into this, and it turns out that this cannot be right. I'm dumb.
Also, I agree with Todd about almost everything else, except Fitness. By request (he didn't actually ask me to rant, more like post a one-liner), I'd like to share what I think so people can either agree or disagree. I'll do my best to make it as referenced as possible so that others who aren't familiar with the concepts can take part in the discussion as well. Here are my thoughts:
First some summaries in more or less my own words:
The adaptive response from controlled stress. The five parts of this are stimulation, intensity, constraint, gradualism, and recovery.
When your body responds to stimulus, whether beneficial or not, in an effort to retain homeostasis, your body responds to inhibit or counteract the initial response (let's call this respond the opponent process). The opponent process is strengthened with increase of the intensity of the stimulus, increase of the duration of the stimulus, or decrease of the time between stimulus exposure.
I will define this as gaining strength or muscle. Usually one follows the other, and both are generally considered good.
Now first argument that I'd like to make is that getting fitter is a "communication" response (through hormones, proteins, and the like; I'm not arguing that any specific substance improves fitness, but rather a combination that tells your body that it's time to grow) rather than the popular view that you gain muscle through tearing and rebuilding muscle fibers. In one study (http://www.ncbi.nlm.nih.gov/pubmed/11782267
), training only the legs produced just as much arm strength (and significantly more half way through the study) as training the arm. Also training both resulted in the most strength gains. In another study (http://www.nejm.org/doi/full/10.1056/NEJM199607043350101
), men being supplemented with testosterone gained more muscle without exercising than those who have been exercising. Likewise, those that did both had the most gains. I believe this is sufficient evidence to show that the actual muscle effort doesn't have to be there as long as the "communication" response is. This also explains the effectiveness of compound exercises (ones that exercise multiple muscle groups) as opposed to isolation exercises (ones that focus on one muscle group at a time).
Since we know that exercise does something that stimulates fitness. We can make a little diagram like this:
Something that occurs with exercise > "communication" response > getting fitter
My next arguments will be based on what I hope to be intelligent guess work. They aren't necessary for the conclusions that I draw, but they help me and hopefully you attempt to understand why things would work instead of just what works. Unfortunately though, guess work is necessary with the research I've read since they measure markers (eg. muscle activation, certain proteins, or hormone ratios) that don't necessarily result in getting fitter, but they do provide stepping stones for theories on such matters.
When looking at what exactly muscles do, they simply contract. It would make sense then to assume that muscle activation level. By the size principle (of which I believe there are exceptions, but not that many), smaller motor units are stimulated before larger motor units. Muscle activation level basically measures how far along the scale of no motor units to all motor units are engaged. I hope that you agree with me that it makes sense to assume (take note that this is only an assumption) that muscle activation is that something, since it's what muscles do. Our diagram then becomes:
Muscle activation > "communication" response > getting fitter
Not too controversial, I hope. One thing that results from this assumption is that the actual amount of weight lighted (assuming we are weightlifting, which is commonly used in studies to get fitter) is irrelevant to muscle activation (http://www.scsepf.org/doc/291208/Paper1.pdf
). The paper contains an analysis of 21 different resistance training studies, where 20 show no significant difference in strength gains compared with only 1 that did (take note that "significant" generally means that their result is 95% not by chance; since the odds appear to be 1 in 21 for a difference in strength gains, it does appear that this "significant" difference occurred by chance). This paper claims that weight doesn't matter, as long as the effort is the same. Another paper (http://www.ncbi.nlm.nih.gov/pubmed/2369907
) that supports this shows size gains with weights that, by common belief (many people think that 60% of 1 Rep Max, the maximum weight they can lift for 1 repetition, is the threshold for hypertrophy), should not result in muscular hypertrophy. And another study (http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0012033
) showed that training at 30% of 1 Rep Max to muscular failure (when you can no longer lift) is equally or more beneficial than training at 90% of 1 Rep Max (this study doesn't actually measure getting fitter, simply certain markers that they believe are part of the "communication" response). So from this, we can say that effort results in muscle activation.
Effort > Muscle Activation > "Communication" Response > Getting Fitter
Now, I'd like to talk more about the link between effort and muscle activation. The first paper in the last paragraph (http://www.scsepf.org/doc/291208/Paper1.pdf
) doesn't include weights lower than their 20 Rep Max. To quote the author of the paper "Very high RMs (e.g., loads lighter than 20 RM) or an extensive time under load (e.g., longer than 2–3 minutes) may involve mechanisms of fatigue that are not conducive to stimulate optimal increases in muscular strength." The reasoning for is best said in Body by Science (page 49): “If you use a weight that is too light, the load will not be meaningful enough. You will recruit the slow-twitch fibers into service, but because they fatigue so slowly, by the time you have started to recruit the intermediate fibers, some of the same slow-twitch motor units will have started to recover. They will then recycle back into the contraction process, thus preventing you from ever engaging the higher-order muscle fibers.” This is consistent with the paper as well, where they found that it was always the last rep which resulted in the most activation. I find the fatiguing and recovery of motor units a very crucial point to my theory of getting fitter.
Enough of the why. Onto the details.
Now exercise is a stress on the body (hopefully a controlled one). It appears from anecdotal evidence that strength gain is an opponent process of exercise. The ways to strengthen the strength gain response would be increasing intensity, increasing volume, and decreasing intervals between exercise. With this knowledge of muscle activation, we could imagine the activation for the muscle or group of muscles as stairs tilted back. Every time we exert that muscle or group, we go up the stairs, when we don't, we fall down a certain amount. For training to be fruitful, obviously we'd have to rise more than we fell. Every chance of rest between reps would move it back ever so slightly, and in between sets, we'd drop by a substantial amount.
Because of this fatigue and recovery issue, I believe that performing any more than 1 set is a waste of time and could possibly hinder the process of getting fitter (little to no benefit, but over double the effort, time, and chance of injury since you do twice as much exercise). Adding sets doesn't provide intensity, and perhaps increases the interval between stimulus because of the necessary recovery. Studies even indicate that the added volume doesn't necessarily help. This study (http://physiotherapy.curtin.edu.au/resources/educational-resources/exphys/00/muscle_strength.cfm
) analyzes many different studies and shows that most show no significant benefit with multiple sets. Only one of the 14 mentioned showed a significant difference, which could be more easily attributed to chance than not. Other studies (http://www.ncbi.nlm.nih.gov/pubmed/9777681
) mention the same.
Another application of this knowledge on fatigue and recovery would be to produce more force in less time. Using the same stairs idea, you would need at least as much force to lift the resistance a certain distance (assuming we are weightlifting for example), thus in the image that I talked about for the stairs, you would go up by the same amount, but since the rest time is lower, you would go down less, and thus reach the necessary intensity in less time (the top of the stairs). Studies on slow vs fast lifting are, from what I've seen, either poorly designed (the one study I saw that stated lifting slow was better didn't even measure strength with the same type of exercise) or inconclusive. Based on anecdotal evidence, slow lifting appears better for joints, and fast lifting for athletic performance. Because of these points, I prefer lifting as fast as possible (effort >>> getting fitter).
One idea, that I have no evidence of but seems to be a conclusion of effort >>> getting fitter, is that once you reach the final intensity, you can apply a reverse pyramid set-up, to continue doing the set with either less resistance, or a smaller range of motion, possibly even leading to an isometric hold before absolute failure (meaning you absolutely cannot lift). I personally lean towards less resistance, because training at your strongest range of motion would probably only increase the imbalance. This seems advantageous because the fatigue would activate larger muscle fibers that normal with the less resistance, and if you can in fact lift more, then you can exert more effort and hopefully make it higher in the figurative stairs.
You might think that reducing resistance would require rest in between repetitions. Because of this, I'm actually a big fan of bodyweight training. Most people believe that bodyweight training is mostly for endurance after a certain amount of strength. And if you do the same exercise all the time and don't gain weight, then it is (look at the quote about using excessively light weights). An amazing book combining the principles of progressive overload (or the gradualism part of hormesis) is Convict Conditioning by Paul Wade. You increase resistance by doing progressively harder bodyweight exercises. Of course there is a limit to what you can do, but honestly, neither you nor I will probably reach that. Or you could buy a weighted vest when you can do 20 one-arm handstand pushups. With bodyweight training, it takes practically no time at all to shift to a version with less resistance (eg. from clap pushups to pushups to girl/knee pushups or from jump squats with knees to chest to jump squats to squats to assisted squats), plus you save money and time on the gym.
Here is the most controversial part of my fitness thoughts: occlusion training, or kaatsu as they call it in Japan. The summary: partially restrict blood flow to your peripheries, do almost any exercise, get fitter. Studies show that, using this technique, you can gain muscle by just walking (http://jap.physiology.org/content/100/5/1460.short
), gain more strength (http://www.springerlink.com/content/1014223x4r820702/
), get bigger muscles (http://www.ncbi.nlm.nih.gov/pubmed/16015131
), gain more strength and get bigger muscles (http://jap.physiology.org/content/88/6/2097.abstract
), and get larger muscles even if they aren't all restricted (aka chest muscles grow with occluded arms) (http://onlinelibrary.wiley.com/doi/10.1111/j.1475-097X.2010.00949.x/abstract
). It even works on highly trained athletes (they were elite rugby players supposedly) (http://www.ncbi.nlm.nih.gov/pubmed/11990743
). Oh and according to this survey (http://www.jstage.jst.go.jp/article/ijktr/2/1/2_5/_article
) of people who use kaatsu (in Japan), the risks are minimal and it's safer than weightlifting (assuming you gradually start occlusion training, the same way you wouldn't attempt to lift huge weights when starting weightlifting). I know that you may think that this is dangerous, your limbs will have to be amputated, and that your time is too valuable to read the safety survey that I linked. Compare this with the medical use of a tourniquet though, which is said to be safe from 10 minutes (anecdotal evidence) to 3 hours (http://www.ncbi.nlm.nih.gov/pubmed/7439813
). Also a properly applied tourniquet results in total blood flow restriction, which sounds more dangerous than partial. The amazing use of this is that 1) you can use light weight (meshing perfectly with bodyweight training), 2) the safety concern makes faster exercise better (fitting with the idea that a fast workout of one set is at the very least more time efficient), 3) less weight = less muscle damage (this can result in training more often, or simply enjoying your non training time more).
I do have theories on how occlusion training works, but I don't feel that making assumptions about that could lead to any useful hypothesis for training.
As for training frequency, I think the most important factor is convenience. You will get stronger training once or twice a week. Or you could train like Bulgarians twice a day. I personally like doing a bit of both, by splitting my workout by exercise and training 5-10 minutes a day. So it's like a total of 2 workouts per week, but once or twice a day.
-use occlusion training (optional)
-do bodyweight exercises
-switch to an easier version when you reach failure
-keep doing that till you can't
-don't do more than one set
-go do something else