The survey in the article that assessed vitamin D deficiency was a bit odd:

>Because physical exams are performed in mobile vans in NHANES, data could not be collected in northern latitudes during the winter; instead data were collected in northern latitudes during summer and in southern latitudes in winter. To address this season-latitude aspect of the NHANES design, we stratified the sample into two seasonal subpopulations (winter/lower latitude and summer/higher latitude) before examining vitamin D status.

Yeah, I'm not surprised that the rates for vitamin D deficiency were low.

>Less than 1% of the winter/lower latitude subpopulation had vitamin D deficiency (25-OHD <17.5 nmol/L). However, the prevalence of vitamin D insufficiency in this group ranged from 1%–5% with 25-OHD <25 nmol/L /.../, even though the median latitude for this subsample (32°N) was considerably lower than the latitude at which vitamin D is not synthesized during winter months (∽42°N).

and the more northern latitude in summer:

>With the exception of elderly women, prevalence rates of vitamin D insufficiency were lower in the summer/higher latitude subpopulation (<1%–3% with 25-OHD <25 nmol/L)

Now imagine if you lived in northern Europe around the 60th parallel, where the sun doesn't get high enough in winter to produce vitamin D.