One of the hallmark features of weight loss research is the fact that the overwhelming majority of dieters will regain the weight, often with interest. [1-3]
The most fundamental first principle of weight loss is an energy imbalance, a calorie deficit. 
Calorie input vs. calorie output.
There is a lot of nutritional bullshit in the Googlesphere, and much of it comes from paradigms attempting to shift the goal posts and have you believe that calories don’t matter once your eating the “right” foods.
In this reductionist model, you’re potentially succumbing to a basic thinking error, a logical brain fart termed the Singe Cause Fallacy. Nutrition has plenty to offer up: carbs, fat, animal produce, gluten, dairy, sugar. Pick one. Blame it.
And you know where you’ll really get stumped? When you cut one of them out, and see a few pounds come off the scale. Then they’ve got you. You’re becoming a dogmatist. You’ve attained dietary enlightenment, because you know that sugar made you fat.
This dupery is evident all over the nutrition realm. Low-carb diets would have you believe that all you need to do is cut carbs that causes and fat will fall off. In reality, if you lose weight on a low-carb diet, its more likely because you increased protein, felt fuller after meals and ended up consuming less overall calories. 
Low-fat, vegan and vegetarian diets work because the high fibre intake, and focus on fruits and vegetables which are low in calorie density, mean you’re eating a ton of volume of food, but with little overall calorie content. 
If you were eating the Standard Western Diet (cereals for breakfast, bagel for snack, sandwich for lunch, pasta for dinner), and you decide to go on a gluten-free diet, then you’ll lose weight because you just removed an entire food group (wheat/gluten containing grains), and created a calorie deficit. That happens to be gluten-free.
If you lose weight on a juice diet, then you’ve lost weight because you spent three days consuming nothing more than 500 calories of liquidised vegetables.
You see the pattern here?
Surely there has to be one common denominator that is producing the same net effect?
In one word: adherence.
In a recent review in the Journal of the American Medical Association, entitled ‘A Call for an End to the Diet Debates’, the authors reviewed 4 different meta-analysis of between 13-24 trials each; a total of between 52-96 clinical trials . To quote directly from the paper:
“The dietary debate in the scientific community and reported in the media about the optimal macronutrient-focused weight loss diets sheds little light on the treatment of obesity and may mislead the public regarding proper weight management… The only consistent finding amongst the trials is that adherence – the degree to which participants continued in the program or met program goals for diet and physical activity – was most strongly associated with weight loss and improvement in disease related outcomes.” [Emphasis added].
You can lose fat on any diet, once you can adhere to over time.
Remember that adherence has to be superimposed on the only thing that matters, the First Principle of an energy deficit. So if adherence is going to be a problem, you’re in trouble. This is why listening to the evangelical crowd on what one food or food group you need to avoid is setting you up for the fall.
There is one thing in the research that is glaringly obvious: rigidity doesn’t work.
This has been repeatedly confirmed in research on calorie restriction and weight control behaviours, which have consistently shown that such behaviours predict weight gain in the future. [1,2]
A 2013 literature review interestingly differentiated between dieting (as an act of deliberate weight loss) and restrained eating (as an act of preventing weight gain) in non-obese individuals, and analysed the results of 40 studies, 20 on each approach . 75% of studies on dieting showed dieting to be a successful predictor of weight gain, in comparison to just 5% of studies on restrained eating.
So what is the difference between dieting and restrained eating?
Restrained eaters don’t identify with being on a diet, and were not purposefully restricting calories with the goal of losing weight. Dieters, on the other hand, were characterised by deliberate restriction of both food choices and calorie intake in an attempt to lose weight.
This is worth paying attention to, a call to pay better attention to behaviours in how we go about losing fat and eating better. The differentiation between the mentality of rigid and flexible dieting is important in this regard.
Rigid dieting fails, and creates unhealthy relationships with food, particularly in promoting overeating once the rigidity cannot be sustained. Flexibility, on the other hand, means not restricting any particular macronutrient or food group unless you have a specific reason to do so (either allergy or intolerance). It means no ‘good’ vs. ‘bad’ food mind-set. It means no crash dieting.
Flexibility means you can stick to the plan.
Ultimately what the JAMA review confirmed is that there is one principle that is paramount in weight loss: being able to stick to the plan.
To hammer the point home, here is another quote from the JAMA review:
“The assumption that one diet is optimal for all persons is counterproductive because this assumption ignores the variation in adherence influenced by food preferences, cultural or regional traditions, food availability, and food intolerances. These are independent of direct physiological effects of macronutrient composition on weight loss. The most important question is how to improve behavioural adherence.” [Emphasis added].
The best way to ‘improve behavioural adherence’ is to ensure that the approach taken is one that you can sustain. Ask yourself one simple question…will you be eating this way 1 week, 1 month or 1 year from now?
Be a flexible dieter.
Within that, you can eat to your personal preference.
Feel lethargic on a low-carb diet or suffer cravings? Don’t eat low-carb.
Love peanut butter? Eat more fats.
Worried about animal welfare? Be a vegetarian.
Suit your personal preference.