Even longer exposures needed for vitamin D sufficiency in winter?
With potentially important health implications ...
Following up from my post from last year on the subject. This could be important. Tell your friends.
We all know that too much UV causes skin damage, though we do need some to maintain healthy levels of vitamin D in our bodies. But how much? I’ve written about that in the past, but science moves on. It seems I may have to review the numbers, at least for conditions when the sun is lower in the sky - as in winter at mid-latitudes.
Most of our body’s vitamin D production is initiated by photo-chemical reactions involving UV radiation from the sun that convert cholesterol in the skin to previtamin D. The relative importance of the different UV wavelengths involved is shown by the blue curve in the upper panel below, which is called the ‘action spectrum’ for its production (as reported by the CIE in 2006). The graph is from a paper we published on the subject back in 2009. There have been questions about the validity of that action spectrum ever since we drew attention to a problem with it in that paper.
The amount of previtamin D produced depends on the overlap between the action spectrum and the sun spectrum, as shown by the blue curves in the lower panel. It is proportional to the area under those curves (noting that the dashed curve for winter has been scaled up by a factor of 10 ).
In our paper we showed there that if the action spectrum were correct, the amount of vitamin D produced near noon in winter would be ‘about 5 percent’ of that in summer at latitude 45S. But it had long been known that there is essentially no production in winter at latitudes poleward of about 40N - so presumably none either in winter at higher latitudes in the southern hemisphere. There was a problem. We suggested that the action spectrum may not be correct.
Twelve years later, it seems our suspicions have been vindicated. A clever experiment undertaken last year by my colleague Antony Young and his team based at Kings College London showed that more consistent results are obtained if the action spectrum is displaced by about 5 nm towards shorter wavelengths.
We don’t yet know if this is the end of the story. The paper didn’t consider changes in the shape of the action spectrum. All they did was progressively shift it by different amounts to find which wavelength shift gave best agreement with their measured vitamin D increases in subjects exposed to UV from sunbed lamps instead of real sunlight. If all that’s required is a simple shift of 5nm, it goes some way towards resolving the inconsistency we raised in 2009. The implications for health advice may also be important.
Using the previously published action spectrum, the ratio of previtamin D production calculated from UV spectra at the summer and winter solstices at Lauder (45S, similar to the black curves above) is 16, meaning that if it takes 2 minutes of near-noon exposure per day for vitamin D sufficiency in summer, it would take just over 30 minutes in winter. In practice it would - of course - probably be much longer because less skin is normally exposed in winter.
On the other hand, using the shifted version of the action spectrum, that ratio increases from 16 to 25, showing that it would be much more difficult to achieve vitamin D sufficiency in winter. The time required would be closer to 50 minutes, rather than 30. It’s moving in the right direction for consistency.
If this shift turns out to be correct, I may have to ask my friend Jerry Burke to update our Global UV and UVNZ smartphone apps, which include estimates of the exposure times necessary for vitamin D sufficiency (In addition to their dependence on UV, these times also depend on skin type and coverage, which can selected in the App). The exposure times reported there will still be correct for high-sun conditions, but as the sun gets lower in the sky they would become progressively longer than currently shown by the apps.
In the plot below, the calculated effect of this 5 nm wavelength shift on the exposure times required for vitamin D sufficiency is shown as a function of SZA, ranging from 0 (overhead sun) to 90 degrees (sunrise or sunset). The calculations assume an ozone amount of 300 DU, which is close to the global average.
The good news is that differences are less than 10 percent for summer conditions, for example when the SZA is less than 35 degrees (corresponding to UVI > ~7). The effect is largest for SZA = 80 degrees. For that SZA, the time required would be longer by a factor of 1.8.
From a sunburn or vitamin D perspective, what happens at SZA = 80 degrees is of little consequence, because of the low amounts of UV that penetrate the long path through the atmosphere at those times. For example, the UVI for that SZA is less than 0.3, which means that it would take more than 10 hours of that exposure to cause skin damage, and much longer to make sufficient vitamin D.
But there are still consequences. The graph below shows the calculated exposure times required for skin damage (erythema) compared with those for sufficient vitamin D using the shifted and unshifted versions of the action spectrum. In both cases, it is assumed that just hands and face are exposed, as is likely to be the case in winter when problems with vitamin D insufficiency arise. As the SZA increases (i.e., as the sun gets closer to the horizon), the calculated times increase markedly. So markedly in fact, that I had to plot them on a log y-axis (as introduced in my last post).
As you can see, the gap between the red curve and others gets smaller and smaller at larger SZAs. It becomes harder and harder to maintain vitamin D sufficiency without inducing skin damage, especially for the shifted spectrum.
The bad news is that the curves actually cross. From that point on, the exposure times required for vitamin D sufficiency with realistic attire - just hands and face exposed - would then exceed those for skin damage. Using the unshifted action spectrum for previtamin D that threshold occurs for SZA greater than about 78 degrees. But with the shifted version, that crossover occurs when the sun is 10 degrees higher in the sky, at a SZA of 68 degrees, which the minimum reached at midwinter where I live near latitude 45S. At that SZA, the exposure periods required for both outcomes (sunburn or vitamin D sufficiency) is about 2 hours. Of course there are few places in the world where the UVI would remain the same for such a long period (can you guess where?), but hopefully you get the idea. The main point is that unless you expose more, of your skin’s area you wouldn’t be able to get enough UV for vitamin D sufficiency without increasing your risk of skin cancer. If most of you skin is exposed to sunlight, the time required would reduce substantially (if you survive the cold).
Troublingly, it also reinforces the fact that even in summer, vitamin D production is most efficient when the sun is highest in the sky. Unfortunately, that’s also the time when its most damaging to our skin. A careful balancing act is needed.
Clearly, more money is needed for research. ☹ We won’t change the UV apps until the dust has settled. Just be aware of that limitation in winter, when exposure time required for vitamin D sufficiency could be underestimated by more than 50 percent.
Thanks for reading this. Previous posts on the intersection between Ozone, UV, Climate, and Health can be found at my UV & You area at Substack.