Is Vitamin D Deficiency a Public Health Concern for Low Middle Income Countries?: a Systematic

Is vitamin D deficiency a public health concern for Low Middle Income Countries?: A systematic literature review

European Journal of Nutrition

1,2,3Kevin D. Cashman, 2Tony Sheehyand1Colette M. O’Neill

1Cork Centre for Vitamin D and Nutrition Research, 2School of Food and Nutritional Sciences, and 3Department of Medicine,University College Cork, Cork, Ireland.

Contact details:

Professor Kevin Cashman

Cork Centre for Vitamin D and Nutrition Research

School of Food and Nutritional Sciences,

University College Cork,

Cork,Ireland.

Email:

ORCID:orcid.org/0000-0002-8185-1039

Supplemental materials and methods

Additional inclusion and exclusion criteria for study selection

Studies included in the present systematic review were those observational studies published in English after 1990 and which met the following requirements:

(i) conducted with randomly selected subjects from the general population, population subgroups or with apparently healthy control subjects from case-control studies, within one of our a prioridefined LMICs (as per Table 1).

(ii) reported mean values and standard deviation (or median and interquartile range [IQR]/range) of serum/plasma 25(OH)D concentrations and/or prevalence data on serum/plasma 25(OH)D below one or more of our specified thresholds (namely <12.5, <25, <50 and <75 nmol/L) or other broadly cognate serum 25(OH)D thresholds, if selected and stipulated a priori by the authors.

When available, the details of type of assay used for assessment of serum 25(OH)D concentrations and/or month(s)/season(s) in which blood sampling within the study occurred, were reported.

In situations where population-based and/or cross-sectional studies were identified for a country and for a specific age-group(s), these were reported in preference to control groups of case-control studies, but the latter were still used (if available) for those age-groups where population-based and/or cross-sectional study data were absent within a LMIC or alternatively, where it provides data on another city/region with the country or combined male and female data over and above that provided by included population-based and/or cross-sectional studies for that LMIC.

Studies were excluded if they were intervention studies, or studies in clinical inpatients groups or case reports/series on the basis that these were of seriously ill subjects (e.g. hospitalised for cancer, cardiovascular disease, TB or other infections). Studies of women screened for osteoporosis were included on the basis these were only as outpatients. Studies were also excluded if the population group focussed on a typically non-resident sample (e.g., overseas military personnel stationed in the country, groups of individuals originally from a LMIC but relocated to a non-LMIC). As per Hilgeret al. [1], studies published before 1990 were excluded.

Additionalinformation on search strategy, quality assessment of eligible studies and data presentation

Study selection was independently conducted by 2 authors (CO’N and KDC), first by a screen of the titles and abstracts, followed by a review of the full text of potentially relevant studies. In addition, all of the included studies were checked against the reference lists of the two systematic reviews on global vitamin D status [1,2] to identify if they contained any additional articles that should be collected and assessed for inclusion. In addition, electronic searches were also performed to identify any reports/publications other than peer-reviewed papers which report data from national nutrition/health surveys in LMICs.

While inclusion of a quality assessment step in a systematic review is common place, we decided to apply only in the context of those LMICs for which our electronic searches identified more than 8 studies in total and/or 3 studies per population subgroup(s) within a LMIC. This allowed us prioritize those studies within such a LMIC for which the data should be presented in the review. Our quality assessment was based on a modified scoring system proposed by Loneyet al. [3] in which a minimum score a study could reach was 0, while the maximum score was 8. A quality assessment step was not used for those LMICs in which the number of studies identified were less than 8 studies in total for a LMIC or less than 3 for a particular population sub-group, as this amount of data was considered too limited to apply an elimination step. We did, however, apply the scoring system post hoc to these included studies to allow us to comment on their quality.

Circulating 25(OH)D concentrations were reported as mean ± SD or median (IQR/range). In all cases, serum/plasma 25(OH)D concentrations were expressed in nmol/L, following their conversion from ng/mL (i.e. ng/mL multiplied by 2.496), if required. Data on the prevalence of serum 25(OH)D <12.5, 25, 50, 75 nmol/L (or alternative cognate thresholds), are presented as available from 85 from the total collection of 89 included studies (4 studies only reported summary statistics data).

References

1. Hilger J, Friedel A, Herr R, Rausch T, Roos F, Wahl DA, Pierroz DD, Weber P, Hoffmann K (2014) A systematic review of vitamin D status in populations worldwide. Br J Nutr 111:23-45
2. Palacios C, Gonzalez L (2014) Is vitamin D deficiency a major global public health problem? J Steroid BiochemMolBiol144 Pt A:138-45
3. Loney PL Chambers LW, Bennett KJ, Roberts JG, StratfordPW (1998) Critical appraisal of the health research literature: prevalence or incidence of a health problem. Chronic Dis Can 19:170-6.
   

Supplemental Table 1. Exemplarsearch strategy for systematic literature review of serum 25(OH)D status in LMICs.

# / Search History1 / Results*
1 / “epidemiologic studies” OR population-based OR population OR survey OR representative OR cross-sectional OR observational NOT (“clinical trials”OR reviews) / 2424303
2 / Human OR children OR child OR adolescent OR adolescence OR adults OR adult OR elderly / 8412609
3 / #1 AND #2 / 1482045
4 / “hypovitaminosis D" OR “vitamin D insufficiency" OR “vitamin D deficiency" OR “vitamin D deficiency" OR “vitamin D" OR “vitamin D2" OR “vitamin D3" OR “25-hydroxyvitamin D” OR “25-hydroxyvitamin D2” OR “25-hydroxyvitamin D2" OR “25-hydroxyvitamin D3" OR “25(OH)D" OR “25(OH)D2" OR “25(OH)D3" OR Calcidiol OR Cholecalciferol OR Ergocalciferol / 82265
5 / #3 AND #4 / 10072
6 / Cameroon OR Cameroonian / 8133
7 / #5 AND #6 / 5

1All fields used. *Search results correct as of July 3rd 2017.

Supplemental Table 2.Mean and/or median circulating 25(OH)D concentrations and prevalence of circulating 25(OH)D concentrations below selected thresholds reported for studies from Low Middle Income Countries (LMICs), identified as part of the present systematic review.

Reference / Country/
region
(Latitude) / Sample size (n) / Population group/
Age1 / Sampling months / Serum 25(OH)D
(nmol/L)2
Mean ± SD
or Median (IQR/range) / Prevalence below serum 25(OH)D threshold (%)3 / Method
(manufacturer)
<12.5
nmol/L / <25
nmol/L / <50
nmol/L / 75
nmol/L
World region: Europe & Central Asia
Jehanet al. 2010[37] / Moldova
(47o N) / 204 / Children /Adolescents
11.4 ± 1.7 y / Jan-April / 45.6 ± 15.0 (m)
40.3 ± 14.8 (f) / - / 15%
(<30 nmol/L) / 45%
(<40 nmol/L) / - / CBP
World region: East Asia & Pacific
8th National Nutrition Survey Report 2013 [42] / Philippines
/Cebu, Davao del Sur, Benguet, Cagayan(14.6oN) / NR / Adults
20+ y / NR / 76.2 ± 0.9 (SEM)
Highest to lowest mean:
107.8 [Cagayan]
73.3 [Benguet] / - / - / - / NR
Rasoet al.2009 [43] / Philippines
/Manila (14.6oN) / 70 (f)
(pm and osteopor) / Older adults
70.0 ± 8.0 y / May to Oct / 87.7 ± 20.0 / - / 0% / - / 36%
(25-79 nmol/L) / NR
Pohet al. 2016 [44] / Vietnam
(14.0oN) / 384 / Children
8.8 ± 2.0 y / All seasons / 56.3 ± 33.3 / - / 11.1% / 48.2% / 77.6% / HPLC
Laillouet al. 2013 [45] / Vietnam
/19 provinces (14.0oN) / 485
Children [C]
542
Women [W] / Children/Adults
45.3 ± 20.3 mo [C]
32.9± 7.2 y [W] / All seasons / 44.5 (95%CI: 42.3-46.8) [W]
43.4 (95%CI: 40.8-46.1) [W] / - / 17.3% [W]
20.6% [C] / 57.7% [W]
57.3% [C] / 92.0% [W]
88.8% [C] / HPLC
Ho-Pham et al. 2011[46] / Vietnam
/Ho Chi Minh City(10.8oN) / 637 / Adults/Older adults
18-87 y / N/A / 92.0 ± 25.5 (m)
75.3 ± 14.7 (f) / - / - / 1% (m)
3% (f) / 20% (m)
46% (f) / ECLIA
(Roche Diagnostics)
Nguyen et al. 2012[47] / Vietnam
/Hanoi and Hanam (21oN) / 491 / Adults/Older adults
13-83 y / N/A / 71.5 ± 22.2 (m)
58.0 ± 18.5 (f) / - / 15.8% (m)
30.5% (f) / ECLIA
(Roche Diagnostics)
Ho-Pham et al. 2012 [48] / Vietnam
/Ho Chi Minh City(10.8oN)) / 210 (f)
105
vegans [V]
105 omnivores [O] / Older adults
60.0± 9.0 y [V]
61.0± 9.0 y [O] / April-Aug / 65.2 ± 18.5 [V]
79.0 ± 17.3 [O] / - / - / 27.3% [V]
6.5% [O] / 72.7% [V]
46.2% [O] / ECLIA
(Roche Diagnostics)
Smith et al. 2016[49]
and
Cambodian Demographic Health Survey 20144[50] / Cambodia
/Mutiplesites (12.6oN) / 781
Children [C]
725
women [W] / Children/Adults
0.5-5+ y [C]
30 (25-34) y [W] / N/A / 69.7 ± 31.2 [W]
91.1 ± 37.3 [C] / - / 4.1% [W]
2.0% Urban(U)
4.7% Rural(R)
2.9% [C]
3.7% (U)
2.8% (R) / 29% [W]
32.8% (U)
27.8% (R)
13.4% [C] / 64.6% [W]
67.6% (U)
63.8% Rural
34.8% [C]
41.5% (U)
33.1% (R) / EIA
(Roche Diagnostics)
Pohet al. 2016 [44] / Indonesia
(0.8oS) / 276 / Children
6.6 ± 3.3 y / All seasons / 52.7 ± 14.9 / - / 0% / 44.0% / 94.3% / Immuno-activity detection system
Greenet al. 2008[51] / Indonesia
/Jakarta
(6oN)5 / 504 (f) / Adults
18-40 y / Jan-March (Jakarta) / 48(95% CI: 46-49) (p) / - / 0.8%
(<17.5 nmol/L) / 61% / - / RIA
(Diasorin
Sarmidiet al. 2008 [52] / Indonesia
/Jakarta(6oN) / 42 (f)
(pm and osteopor) / Older adults
67.6
(Range: 53-77)y / Oct-Dec / 52.4 (Range 17.5-154.9) / - / - / 61.9% / - / EIA
(NR)
Rinaldiet al. 2007[53] / Indonesia
/Jakarta(6oN) / 62
innursing homes / Older adults
71.1± 7.2 y / Jan / 68.2± 21.6 / - / - / 22.6% / - / EIA
(NR)
World region: South Asia
Haugen et al. 2016 [62] / Nepal
/Bhaktapur
(27.7oN) / 466
breastfeed infants [IF]
500 lactating women [LW] / Infants/Adults(mothers)
7 (4-9) mo [BIF]
25.8 ± 4.2 y [LW] / All seasons / 47.4 ± 16.4 (LW) (p)
82.0 ± 21.4 (BIF) (p) / - / 14.0% (LW)
0.6% (BIF)
(<30 nmol/L) / 59.8% (LW)
3.6% (BIF) / 95.2% (LW)
41.0% (BIF) / LC-MS/MS
Schulze et al. 2014 [63] / Nepal
/Rural(28.4oN) / 1,000 / Children
7.5 ± 0.4 y / All seasons / 66.3 ± 18.5 (p) / - / - / 17.2% / - / EIA
(IDS)
Jiang et al. 2005 [64] / Nepal
/District of Sarlahi(27.0oN) / 1,163
pregnant
(T 1) / Adolescents/adults
15-45 y / All seasons / 51.1 ± 24.6 / - / 13.9%
[4.3% summer;
24.4% winter] / - / - / RIA
(Nichols)
Perumalet al. 2015 [65] / Bangladesh
/Dhaka (23.8oN) / 55 infants
(followed for up to 6 months)6 / Infants
0.5 y / N/A / 77.5 ± 29.1
(at 6 months; n 34) / - / 6%
(<30 nmol/L) / 21% / 53%
(<80 nmol/L) / LC-MS/MS
Islam et al. 2006 [66] / Bangladesh
/Dhaka city
(23.8oN) / 36 healthy (f)
30 veiled (f) / Adults
22.3 ± 1.9 y
(healthy)
47.7 ± 9.4 y
(veiled) / June-July / 30.3 ± 22.6 (healthy)
31.0 ± 11.0
(veiled) / - / 38.9%
(healthy)
30.0%
(veiled) / 77.9% (healthy)
83.0%
(veiled)
(<40nmol/L) / - / RIA
(Diasorin)
Islam et al. 2002 [67] / Bangladesh
/Dhaka city and Mymensingh
(23.8 & 24.7oN, and high and low s-e, respectively) / 189 (f)
99 low s-e
90 high s-e
Shari dress code7 / Adults
16-40 y / Feb-March
(low s-e)
April-May (high s-e) / 39.9 ± 12.5low s-e
40.7 ± 10.6high s-e
(NPW)
44.8 ± 20.0low s-e
49.0 ± 16.7 high s-e
(PW) / - / 17%low s-e (t)
12% high s-e (t) / 50% low s-e 38% high s-e
(<37.5 nmol/L) / - / RIA
(Nichols)
Marasingheet al. 2015 [82] / Sri Lanka
/Ragama,Westernprovince(~7oN) / 340 / Children
2-5 y / N/A / 58.7 ± 22.4 / - / 5% / 34.1% / - / Liaison 25(OH)D total assay (Diasorin)
HettiarachchiLiyanage 2010 [83] / Sri Lanka
/Galle District, Southern Province(~6oN) / 248 / Children
3-5 y / N/A / 87.0 ± 33.5 (m)
81.6 ± 37.0 (f) / - / - / 26% (m)
25% (f)
(<35 nmol/L) / - / RIA
(in house)
Meyer et al. 2008 [84] / Sri Lanka
/Kandy,Central Province(~7o N) / 196 / Adults
30-60 y / Aug-Dec / 54.2 (t)
63.0 ± 22.8 (m)
47.4 ± 15.8 (f) / - / 0% (m)
6.3% (f) / 34.1% (m)
58.6% (f) / - / RIA
(Diasorin)
World region: Middle East & North Africa
Loudyiet al.2016[91] / Morocco
/Rabat
(34oN) / 102
Mother [W]-infant [N] pairs
(78.9% W veiled) / Pregnancy/ Neonates
28.3 ± 6.7 y [W]
- [N] / All seasons
(Jan-Dec) / 27.7 ± 14.8 [W]
25.3 ± 17.0 [N] / - / - / 90.1% [W]
92.9% [N] / - / CIA
El Maghraouiet al.2012[92 / Morocco
/Rabat
(34oN) / 178(f) / Older adults
58.8 ± 8.2 y / June-Sept / 39.5 ± 29.0 / - / 51.6% / 65.7% / 85.3% / ECIA
(Roche Diagnostics)
El Maataouiet al.2016[93] / Morocco
(~31oN) / 254 / Older adults
>50 y / All seasons / 51.8 ± 16.5 (m)
50.2 ± 23.2 (f) / - / 7.1% (t)
4.4% (m)
8.6% (f) / - / 86.6% (t)
89.6% (m)
86.0% (f) / ECIA
(Roche Diagnostics)
El Lithyet al. 2014[94] / Egypt
/Cairo
(~30.0oN) / 80 (f)
healthy control pregnancy(T3)
(veiled) / Pregnancy
24.4 ± 3.8 y / All seasons / 46.6 ± 6.1 / - / - / - / - / RIA
(NR)
El Rifaiet al. 2014[95] / Egypt
/Cairo
(~30.0oN) / 135
Mothers [W]-neonate [N] pairs
(86.7%veiled) / Neonates/Adults(Mothers)
26.0 ± 5.8 y [W]
- [N] / All seasons / 81.5 ± 53.5 [W]
41.7 ± 25.0 [N] / - / - / 40% [W]
60% [N] / 68.9% [W]
92.6% [N]
(<80 nmol/L) / ELISA (DRG 25(OH) vitamin D)
Abd-Allah et al. 2014[96] / Egypt
/Zagazig
(~30.6oN) / 120
Healthy controls / Adolescents
11.1 ± 2.6 y / N/A / 46.6 ± 13.5 / - / 30%
(<37.5 nmol/L) / 45% / - / EIA
(Biosource
Europe S.A)
Abu Shady et al. 2016[97] / Egypt
/Giza(~30.0oN) / 200 / Adolescents
10.4 ± 0.6 y / April-May / 103 ± 34.9 / - / - / 11.5% / 26.5% / EIA
(NR)
Botroset al. 2015[98] / Egypt
/Cairo and Port Said
(~30.0 and 31.3oN, respectively) / 404 (f)
51 lactating
50 pregnant
208 child-bearing age
38 elderly
57 geriatric / Adults
26.3 ± 5.0 y
25.7 ± 5.4 y
31.5 ± 8.2 y
58.0 ± 4.2 y
76.0 ± 6.7 y / Spring and Summer / 30.0 (12.5-52.5)8
36.5 (25.0-95.0)
27.5 (4.0-62.5)
66.2 (35.0-120.0)
27.7 (13.2-47.7) / - / - / 72.6%
54.0%
72.0%
39.5%
77.2% / 86.3%
64.0%
77.3%
52.6%
78.9% / RIA
(NR)
Ragabet al. 2013 [99] / Egypt
/Cairo (30.0°N) andFayoum (23.9°N) / 105 (f)
35 Pre-menopausal [Pre]
70 Post-menopausal [pm] / Adults/Older adults
51.9 ± 12.3 y / All seasons / 49.8 ± 10.8 (t)
54.5 ± 10.3 [Pre]
47.3 ± 10.3 [pm] / 25.7% [Pre]
68.6% [pm] / RIA
(NR)
Aly et al. 2014[100] / Egypt
/Dakahlia(Rural)(~31.1oN) / 176 / Older adults
67.7 ± 6.7 y / Winter / - / - / - / - / 26.0%
37.9% (m)
15.3% (f) / EIA
(Immunodiagnostics AG)
World region: Latin America & Caribbean
Velarde Lopézet al. 2016[104] / Guatemala
/Guatemala city (14.6oN) / 68
Healthy
Controls / Children
24.8 (12.2-39.9) mo / Sept-Dec / 68.8 (53.5-80.8) / - / - / 19.7% / 67.6% / HPLC
Naqviet al. 2016[105] / Guatemala
/Izabal Province(~15oN) / 43 (Mayan [M])
43 (Afro-Caribe [A-C]) / Adolescent
13.0 ± 1.4 y
14.0 ± 1.6 y / Feb. / 64.5 ± 14.8 [M]
74.5 ± 19.8 [A-C] / - / - / 21% [M]
5% [A-C] / 72% [M]
56% [A-C] / Liaison 25(OH)D total assay (Diasorin)
Sudet al. 2010 [106] / Guatemala
/Quetzaltenango Province
(14.5o N) / 108 / Older adults
69.0 ± 7.2 y / Summer / 53.3 ± 15.0 / - / - / 46.3% / 96.3%
(<80 nmol/L) / RIA
(Nichols)
World region: Sub-Saharan Africa
Braithwaiteet al. 2012 [107] / The Gambia
/West Kiang
(13.3oN) / 382 / Children
8.9± 2.4 y / All seasons
(excluding summer) / 61.8± 15.8 (p) / - / 0.5% / - / - / RIA
(Diasorin)
Asprayet al. 2005[108] / The Gambia
/West Kiang
(13.3oN) / 113 (f) / Adults/Older adults
25-75+ y / All seasons / Mean range 72.8-113.3 (depending on age group) (p)
80.9± 22.8 (p) for 22-44 y
91.4± 24.5 (p) for 55-64 y
82.9± 25.7 (p) for 65+ y / - / - / - / - / RIA
(Incstar)
Wejseet al. 2007[109] / Guinea-Bissau
/Western Africancoastline
(~12 oN) / 494
Healthy
controls / Adults
37.3 ± 12.9 y / All seasons / 85.3 ± 34.8 / - / 4.9% / 13.2% / 39.1% / LC-MS/MS
Kosket al. 2016 [116] and
Durazo-Arvizu et al. 2014 [117] / Ghana
Kumasi
(~6o N) / 482 / Adults
34.2 ± 6.7 y / All seasons / 75.8 ± 17.3 / - / 0.2%
(30 nmol/L) / 4.6% / 50.1% / LC-MS/MS
Njeminiet al. 2002[118] / Cameroon
/South-west province
(~5oN) / 152 / Older adults
>60 y / N/A / 52.8 ± 19.3 / - / - / 24%
(<37.5 nmol/L) / - / RIA
(Diasorin)
Njeminiet al. 2011 [119] / Cameroon
/South-west province
(~5oN) / 137 / Older adults
66.8 ± 6.1 y / N/A / 53.3 ± 19.0 / - / - / 22.6%
(<40 nmol/L) / - / RIA
(Diasorin)
Ramos et al. 2015 [120] / Uganda
/Kampala
(0.4oN) / 31 pregnant [PW]
25
Non-pregnant [NPW] / Pregnancy/Adults
18-31 y [PW T3]
20-25 y
[NPW] / All seasons / 68.6 ± 25.0
(pregnancy T3) / - / - / 25% [PW]
36% [NPW] / - / Liaison 25(OH)D total assay (Diasorin)
Nabetaet al. 2015 [121] / Uganda
/Kampala
(0.4oN) / 41
Non-malnourished [NM]
117
Malnourished [Mal] / Children
6-24 mo / All seasons / - / 0% [NM]
0% [Mal] / 2.4% [NM]
2.6% [Mal] / 14.6% [NM]
12.0% [Mal] / 36.6% [NM]
43.6% [Mal] / ECIA
(Roche Diagnostics)
Friiset al. 2013 [122] / Tanzania
/Mwanzacity
(2.5oS) / 347
community non-TB controls / Adolescents/Adults
>15 y / All seasons / 84.4 ± 25.6 / - / - / 4.3% / 39.6% / Liaison 25(OH)D total assay (Diasorin)
Luxwoldaet al. 2013 [123] / Tanzania
/Variouspopulations
(2-4oS) / 138 pregnant [PW]
57
Cord bloods [Cb]
88 (m+ NPW) / Pregnancy/Cord/Adults
33.0 ± 10.0 y [PW]
26.0 ± 6.0 y
- / N/A / 115.1 ± 27.0
138.5 ± 35.0 [PW]
79.0 ± 26.4 [Cb] / - / 0%
0%
0% / 0%
1%
9% / 13%
2%
56%
(<80 nmol/L) / LC-MS/MS
Boillat-Blanco et al. 2016 [124] / Tanzania
/Kinondoni
(6.8oS) / 358 Healthy controls / Adults
36.1 ± 13.0 y / March to Sept / 89.6 ± 26.9 / - / - / 5.9% / 31.0% / ECIA
(Roche Diagnostics)
Turneret al. 2016[125] / Zimbabwe
(~19oS) / 571 (f) / Adults
26 (22-30) y / N/A / 74.5 (61.8-85.8)9 / - / - / 9% / 51.5% / Liaison 25(OH)D total assay (Diasorin)
Mastalaet al. 2013[126] / Malawi
/Blantyre
(15.8oS) / 56
Non-TB
outpatients / Adults
38.9 (18-80) y / June-July / 92.6 ± 29.7 / - / 0% / 5.4% / 28.6% / EIA
(IDS)
Felekeet al. 1999[113 / Ethiopia
/Addis Ababa
(9oN) / 31
Full-term pregnant women [PW]
30
adults / Pregnancy/Adults
20-22 y
22-28 y [PW] / July
Aug-Sept / 23.5 (18-29)
25 (17-46) [PW] / - / 77%
55% [PW]
(<30 nmol/L) / 97%
81% [PW] / - / HPLC
Herradoret al. 2014 [114] / Ethiopia
/Amhara Regional State
(~11oN) / 627 / Children
9.0± 3.2 y / May-Dec / 80.1 ± 26.4 / - / - / - / 49.0% / Liaison 25(OH)D total assay (Diasorin)
GebreegziabherStoecker 2013 [115] / Ethiopia
/Rift Valley
(7.3oN) / 196 (f) / Adults
30.8± 7.8 y / July / 39.7 ± 9.7 (p) / - / 14.8%
(<30 nmol/L) / 84.2% / - / EIA
(IDS)

Abbreviations: t: total sample of both males and females; f, females; m, males; RIA, Radioimmunoassay; EIA, Enzyme immunoassay; CPB, Competitive Protein Binding; ECIA, Electro chemiluminescence immunoassay; HPLC, High-performance liquid chromatography method; LC-MS/MS, liquid chromatographic with tandem mass spectrometry; CIA, chemiluminescence immunoassay; W, women; IF, infant; TB, tuberculosis; HIV, human immunovirus; M, Mayan;A-C, Afro-Caribe; y, years; mo, months; CI, confidence interval; N/A, not specified; NR, method assay not reported; s-e, socio-economic; T3, trimester 3; pm, postmenopausal; osteopor, osteoporotic. SEM, standard error of mean.

1Unless otherwise specified.

2Serum 25(OH)D unless measured in plasma, which will be represented as (p).

3Some authors used alternate serum 25(OH)D thresholds, where this occurred they have been reported in italics and with the cut-off indicated.

4Based on a micronutrient component of the Cambodian Demographic Health Survey which collected data for women and children <5 y in the 2014.

5Data for this country reported as part of a wider study with more than one country.

6Infants born to women randomised throughout their pregnancy to placebo, and then followed up for 6 months.

7Shari dress code which is a covered-up style dress but face and hands are exposed.

8Units reported as ng/dl in the paper, but presumed to be ng/ml.

9Serumfrom a decadebeforebutwasstored at -80oC untilanalysis.

A

B

Supplemental Figure 1. Panel A: The average per capita plant-based oil supply (g/day) in the 26 Low middle income countries for which FAO balance sheet data was available (averaged over the period 2003 to 2013); Panel B: The estimated average per capita supply of vitamin D (g/day) arising from theper capita plant-based oil supply in panel A fortified with 0.05 ppm [blue bars] and 0.125 ppm [blue bars] vitamin D representing the spread in national standards in these countries (see Table 3). National standards in these countries allow voluntary addition* of mandatory** addition of vitamin D to vegetable oils.

Supplemental Table 3. Changes in UV radiation levels (as UV Index*) with season and latitude (World Health Organisation)

*Maximal UV Index values are given for a range of cities in different countries, calculated for the 21st of each month. UV Index readings of 1-2, 3-5, 6-7, 8-10, and 11+ are low, moderate, high, very high and extreme, respectively.