Verified by 
The question of what is a mental disorder and what is just an expected part of the “human condition” is an age-old, fraught and still unresolved question. Severe mental disorders are easy to recognize, but there is no clear line between normal and abnormal, and no definitive diagnostic test. So there is bound to be disagreement about the large swath of conditions in the grey zone.
Here’s a simple suggestion for making sense of the dilemma: Don’t try too hard to categorize mental disorders. Even though the standard diagnostic manual of mental disorders, called DSM-5, does indeed categorize them in distinct "boxes"—often usefully—most experts agree that the classification system is best considered just a guideline, to be applied flexibly. Think more fluidly about most mental disorders being on a continuum with normality—one end of a spectrum for a given human trait / tendency. Moreover, many mental disorders also overlap with each other, such that an individual might easily be “diagnosed” with several “disorders.” Visualize the spectrum graphically: plotted as a bell curve, like for IQ scores or height or other normally distributed traits in the population1.
Understanding the spectrum
Let’s consider anxiety as an example: We all experience anxiety, and it’s often quite necessary that we do, as a certain amount of anxiety motivates us, and alerts us to be cautious. But some people are prone to experience anxiety more intensely, more frequently and for longer duration than others, causing those individuals greater distress and impairment in their functioning. Those people are at the right hand side of the bell-curve for the anxiety trait and might be considered to have a moderate “disorder.” They may be very sensitive, excessively prone to nervousness, stress and worry, emotionally reactive, uptight. Their sensitivity can be a positive trait too—they may be exquisitely emotionally attuned.
Those with the most severe, disabling anxiety are at the extreme end of the population spectrum for this trait—the right hand tail of the bell curve on the graph, and can be considered to have a severe disorder. The level of severity with which we set our criteria defining any degree of anxiety disorder will determine the percentage of the population who will be “diagnosed” as suffering from such a disorder, versus merely being considered normal people who are somewhat anxious2. But remember, it’s not really a black or white question of “do you have a disorder or do you not?” It’s a shades-of-grey question of “to what degree is this trait / tendency causing you distress or impairing your functioning?”
At the other end of the curve (the left hand tail of our anxiety bell curve, if you will) are people who actually experience unusually little anxiety. This isn’t necessarily a good thing at all. Think of these people as relatively fearless and emotionally under-reactive. These people may take excessive risks and may be emotionally insensitive. At the extreme of this end of the spectrum, some of these people might even be predisposed to be psychopaths.
Of course, each of our full personalities, strengths and vulnerabilities is defined by a complex interplay of multidimensional traits, and no two of us are alike.
Why do extreme traits exist?
So, you might wonder, why do extreme traits exist at all? Because otherwise we would have gone extinct as a species a long time ago. Diversity of traits is essential for a species to survive and evolve. The environment keeps changing, and a trait that is a weakness in one environment at one time and place (e.g. too cautious, or too exploratory) might well turn out to be a strength in another environment at another time and place3.
Statistically, some individuals in every generation will inherit more extreme versions of traits at either end of the spectrum. While extremes of a trait might seldom be advantageous, they are an inevitable statistical result of genetic diversity, and that diversity is certainly adaptive for the species as a whole. You can’t have a bell curve distribution of traits without the two extremes or “tails” of the curve4.
Why the apparent modern epidemic of mental illness?
One reason may be that we have artificially skewed our environment in modernity such that cultural and technological evolution have outstripped our much more slowly-evolving biological capacity to handle these kinds of stresses (time management, organization, attention, etc.). We are also getting better at understanding and recognizing mental health problems. And younger generations are more likely to label personal difficulties as mental health issues.
It's not yet clear whether there is a true increase in mental disorders. This is probably because much hinges on definitions—where you draw those vertical lines on the graph. Those lines are determined in no small part by the demands on human functioning in today's world and by society's expectations of what is normal and what is not (i.e. what is impaired functioning), given those demands.
This should not be mistaken for a position that mental disorders do not exist or are purely a social construct. Mental disorders, like all aspects of brain and behaviour, are biologically based and can be extremely debilitating. Some have very clear biological correlates. The issue here is the lack of a sharp, fixed line dividing impaired functioning (and significant distress) from "normal" functioning, and how the position of those lines will be different in different environments.
References
1. It's possible that certain traits are not exactly normally distributed as a bell curve and may have a more asymmetrical distribution.
2. Even clearly abnormal conditions that are not just accentuations of normal human traits, such as head injuries, developmental delay, dementia, or schizophrenia, have a range of severity, with the mildest versions being practically indistinguishable from normality.
3. ADHD is another good example of a trait or set of traits that probably often had adaptive advantages in paleolithic environments (people with these traits would have been more adventurous and exploratory). But in today's highly structured environment it tends more often to be disadvantageous—these people are unfocused, easily bored, restless and stimulus-seeking. To understand ADHD using the model suggested here, on the X-axis of the bell-curve replace anxiety with attention span / focus (and closely associated executive functions like cognitive-control or self-control, i.e. self-discipline). Consider also the liability of being at the opposite end of the spectrum.
4. Interestingly, even psychosis may exist on a continuum: See “Controversies in Psychiatric Diagnosis: What Is a Mental Disorder? And When Are Irrational Beliefs Delusional?” Dr. Ralph Lewis, Skeptic Magazine, December 2013.
