How much collagen per day is worthwhile?
Five grams of hydrolysed collagen per day produces measurable improvements in skin moisture and structure in most people after four to eight weeks; for knee complaints, a completely different type has been studied (40 mg of undenatured type II collagen), so look carefully at which product you are buying.
Five grams of hydrolysed collagen per day is the dose that proved effective in most studies on the skin. At that amount, multiple studies observed measurable improvements in skin moisture, dermal density and wrinkle visibility after four to eight weeks. Studies used doses between 2.5 and 10 grams; there is no clear evidence that a higher dose produces more effect than five grams.
The grams are not the whole story, however. One study using a much lower total amount (less than two grams per day) still saw an effect, because the product contained a high proportion of specific active peptides. Another study showed that collagen with a richer peptide profile worked better than collagen of a comparable weight but poorer composition. In other words, the quality and composition of the product matter at least as much as the dose stated on the label.
Combining vitamin C with collagen appears sensible, since vitamin C is required for the skin's own collagen production. In a study in which women aged 40 to 65 received five grams of collagen plus 80 mg of vitamin C, dermal density and texture improved after 16 weeks. Adding hyaluronic acid to the formula offered no additional benefit over collagen without hyaluronic acid.
Completely different rules apply to joints. Undenatured type II collagen, at a dose of just 40 mg per day, proved more effective for knee pain caused by osteoarthritis after six months than both placebo and the common combination of glucosamine and chondroitin. This type of collagen most likely works through the immune system, not as a building block. It is a fundamentally different product from the grams of hydrolysed collagen used for the skin; the two cannot be used interchangeably.
No side effects were reported in any of the studies. The effects on the skin persisted for two weeks after stopping in one study. The groups studied were relatively small, and most studies lasted no longer than three to six months, so little is yet known about longer-term effects.
This answer is based on eight randomised studies and one systematic review of eleven such studies. The systematic review included a total of 805 participants. One of the review articles used was written by authors affiliated with a collagen manufacturer, which may introduce bias in that source; the remaining studies are of a more independent design. Most studies lasted eight to sixteen weeks and examined middle-aged women; whether the effects also apply to men or younger people cannot be determined from these studies.