You haven’t defined health. 
Health is one of those tricky concepts. Everybody knows what it is, but nobody can define it. To understand why this is a problem consider the second item on your list “regular exercise”. It might seem obvious that regular exercise is a cheap way to increase health, but in the youngest age groups particularly exactly the opposite is true. Once we separate out those individuals who can’t easily exercise for pre-existing health reasons (like being in a wheelchair) young people (18-35 IIRC) are significantly more likely to suffer health problems if they exercise regularly. (Don’t ask for a reference for that. I read it in a reputable publication, but I no longer have access to it.)
That might sound screwy, but the reason is quite simple: sports are dangerous. People who indulge in exercise suffer everything from torn muscles and blisters from running, spinal injuries from lifting weights, assault injuries in match fights through to severe head and spinal trauma from falls and crashes. And of course any number of other sports injuries should be obvious. Ask anyone who exercises regularly and they will almost certainly have suffered some sort sporting injury in the last 6 months.
Meanwhile for young people the number of health problems prevented by exercise are actually quite low. Most of the health problems like diabetes and heart disease tend to set in during early middle age. Moroever these health problems tend to be singular. A person has cardiomyopathy or adult onset diabetes. They are ill once, with one condition, albeit chronic. In contrast regular exercise will likely produce several minor acute illnesses every year. So even over a lifetime regular exercise may in fact actually reduce health.
So the answer to the question very much depends on how you define health and how you then measure. If you simply define health as ‘lack of illness or injury” and measure it by how many illnesses/injuries a person suffers in their life then you get some very surprising answers to your question. If you define health in terms of life-threatening/debilitating illness then you get a totally different list.
Then we have to decide when health ceases to be an issue. This might seem obvious too. Health becomes an issue at or shortly before birth and it ends at death. That’s fine, except illness increases with age, especially chronic illlness. So rather ironically the most cost effective way to ensure good health is not to treat any chronic terminal illness. Not only does that mean that a person diagnosed with cardiomypathy at 50 is unlikely to develop glaucoma at 60, it actually has a negative cost. It’s free not to treat these illnesses and die rapidly and expensive to treat them and live a functional life with chronic illness for 30 years.
Taken to extremis if we assume that health encompasses whole-of-lifespan then the single most cost effective way of ensuring good health is to commit suicide as early as possible. If everyone killed themselves as soon as possible they would suffer far, far less illness for the relatively cheap cost of the suicide method.
Now obviously that is not what you are asking. But it does point up the difficulty of ascertaining just what you are asking. How do we say whether getting screened for cancer is really a cheap way of ensuring good health when it has dramatically increased the number of person years spent living with cancer as well as the amount spent on cancer treatments? It is after all cheaper to die of a tumor than it is to treat it.
Another option is to simply discount illnesses contracted in the last X years of life, or after the average lifespan has been reached. That way we can accept that people get ill with age and can measure how healthy they were before illness became inevitable. Or we can measure the number of person years on average saved per dollar for every intervention.
If we look at those things we also get some surprising answers. IIRC wearing a seatbelt (or full face helmet for bikers) is the single most cost effective thing an individual can do to add years to their lives. Because seatbelts are already fitted to most cars by law it’s free to the individual and saves tens of thousands of life-years every single year nationwide. That’s cost effective. We see similar factors for things like home smoke detectors, sunscreen use, child-proof cabinet locks, teaching children to swim and so forth. All either free or less than $10 with the potential to save thousands of person years every year in terms of prevented deaths.
This is where the surprises come from., Things like diet and exercise can ward off potentially fatal chronic illnesses that kill later in life, but there are numerous things that are effectively free that can ward off acutely fatal things like burning to death or drowning.
Then we get to things that are relatively expensive but have almost incalculable benefits. Provision of sewerage to separate drinking water from faeces and treatment of drinking water is fairly expensive, but it has probably saves 90% of the people in every major city and town from severe illness every 12 months. The same for food hygiene standards and so forth. These might seem obvious but they are highly cost effective methods of health promotion.
And of course if we want to cheat by including “not smoking” then we could also include other prohibitions like “Not exceeding the speed limit by more than 10mph”, “not driving tired” “not swimming while drunk” and “Not jumping of tall buildings while on acid”. All those things are far more risky than smoking and just as cheap not to do. But Is spu
I think you really do need to define things a bit more before any answer is possible.