Skip to main content
. Author manuscript; available in PMC: 2022 Apr 13.
Published in final edited form as: Nutr Cancer. 2018 May 21;70(5):804–820. doi: 10.1080/01635581.2018.1470653

Table 4.

Exposure Assessment.

Author year Serum 25(OH)D Reference Cut Points Level of Detection (LOD) Adjustments made for season at time of 25(OH)D sample donation Timing of 25(OH)D sample donation in cases Multiple 25(OH)D assessments Method of 25(OH)D assessment Star rating

Abbas 2009 Apriori cut points
<30 nmol/L
30–34 nmol/L
45–60 nmol/L
≥60 nmol/L

<12 ng/mL
12–18 ng/mL
18–24 ng/mL
≥24 ng/mL
No data Yes – covariate in statistical analysis Average 189 days after diagnosis No Immunodiagnostic Systems Limited enzyme immunoassay
1 batch
CV = 3.1% intra-assay
CV = 3.3% interassay
10% duplicate measures with average absolute deviation = 1.7%
***
Abulkhair 2015 Apriori cut points
≤25 nmol/L
>25 nmol/L

≤10 ng/mL
>10 ng/mL
No data No Initial visit with oncologist No DiaSorin Liaison CLIA
(No CV data provided)
**
Crew 2009 Apriori cut points
<20 ng/mL
20–29 ng/mL
30–39 ng/mL
≥40 ng/mL
Range of measures in study was 0–103 ng/mL. Only one subject had a value of 0 ng/mL. Yes – covariate in statistical analysis Variable – average of 60 days between time of diagnosis and blood draw. 20% of cases received chemotherapy before blood collection. No DiaSorin RIA
Cases and controls assayed in each of 8 batches
Respective accuracy and precision:
@ 17 ng/mL = +4%, 15%
@48 ng/mL = +7, 18%
@57 ng/mL = +15%, 15%
**
Eliassen 2011 Quartile cut points
<18.4 ng/mL
18.4–< 24.6 ng/mL
24.6–<30.6 ng/mL
≥30.6 ng/mL
No cases below LOD Yes – covariate in statistical analysis Mean 57 months (Range
1–127) before diagnosis
No RIA
CV = 10.7% or 6.0%, based on sample batch
***
Eliassen 2016 Apriori cut points
<30 ng/mL
≥30 ng/mL
No cases below LOD Yes – stratified statistical analysis by season (summer/winter) of blood draw Mean 7.9 years (Range 1 month to 20 years) before diagnosis Yes – 413 cases and controls contributed two measures of 25(OH)D Analyzed in 7 batches
Batches 1 & 2 – PBA
Batch 3 – RIA
Batches 4–7 CLIA
CV for batches 1–3 = 18.1%–21.8%
CV for batches 4–7 = 6.5%–9.4%
****
Kuhn 2013 Season standardized apriori cut points
≤39.3 nmol/L
>39.3 – 50.9 nmol/L
>50.9 – 63.0 nmol/L
>63.0 nmol/L

≤15.7 ng/mL
>15.7 – 20.4 ng/mL
>20.4 – 25.2 ng/mL
>25.2 ng/mL
Lower LOD <6.7 nmol/L (2.7 ng/mL)
5 cases excluded with 25(OH) D below LOD
Yes – used season standardized 25(OH)D values Mean 4.1 years (+/−2.8 years) before diagnosis No OCTEIA 25(OH)D enzyme immunoassay
CV @ 23.0 nmol/L = 3.7–6.6%
CV@ 42.6 nmol/L = 5.6–10.9%
***
McCullough 2009 Tertile cut points
<45.9 nmol/L
45.9–64.2 nmol/L
≥64.2 nmol/L

<18.4 ng/mL
18.4–25.7 ng/mL
≥25.7 ng/mL
No Lower LOD. No cases excluded for 25(OH)D level. Lowest case reported was 7.5 nmol/L (3 ng/mL) and lowest control was 7.8 nmol/L (3.1 ng/mL). Yes – covariate in statistical analysis Range <1 month to 6.9 years before diagnosis No DiaSorin Liaison CLIA
CV range = 6.4–8.8%
****
Park 2015 Apriori cut points
<20 ng/mL
≥20 ng/mL
No lower LOD established. No cases excluded on this basis. No Time of diagnosis No DiaSorin RIA used for both cases and controls. However, cases and controls were analyzed at different laboratories.
(No CV data provided)
**
Peppone 2012 Apriori cut points
<20 ng/mL
20–31 ng/mL
<32 ng/mL
≥32 ng/mL
Lower LOD = 5 ng/mL
One case was below LOD and assigned a value of 5 ng/mL
Yes – covariate in statistical analysis Blood sample taken prior to surgery, mean 30.1 days No January – June 2009 – CLIA
After June 24, 2009 – LC-MS/MS
LC-MS/MS levels were 14% higher than CLIA levels, controlling for age, race, and month of test.
(No CV data provided)
***
Rejnmark 2009 Tertile cut points
<60 nmol/L
60–84 nmol/L
>84 nmol/L

24 ng/mL
24–33.6 ng/mL
33.6 ng/mL
Lower LOD <10 nmol/L (4 ng/mL). Persons with levels below the LOD were assigned a value of 9 nmol/L (3.6 ng/mL) No Prior to mammogram/study enrollment – average time unclear No ID-LC-MS/MS
CV@ 41.2 nmol/L = 8.8%
CV@ 25.3 nmol/L = 9.4%
***
Scarmo 2013 NSMSC season-adjusted quintiles+
<40.09 nmol/L
40.09–49.57 nmol/L
49.58–57.37 nmol/L
57.38–67.95 nmol/L
>67.95 nmol/L

<16.0 ng/mL
16.0–19.8 ng/mL
19.8–22.9 ng/mL
23.0–27.2 ng/mL
>27.2 ng/mL

NYUWHS season-adjusted quintiles+
<38.18 nmol/L
38.18–50.14 nmol/L
50.15–60.77 nmol/L
60.78–73.64 nmol/L

>73.64 nmol/L
<15.3 ng/mL
15.3–20.1 ng/mL
20.1–24.3 ng/mL
24.3–29.5 ng/mL
>29.5 ng/mL
No cases below LOD Yes – used season standardized 25(OH)D values At study enrollment. Mean 8.7 years before diagnosis. Yes – 678 cases and 1208 controls contributed two measures of 25(OH)D, at least 1 year apart. Average time between sample donations was 2.1 years in the NYUWHS cohort and 4.4 years in the NSMSC cohort. DiaSorin Liaison CLIA
CV (NYUWHS)
 Intra-assay = 9.5%
 Inter-assay = 11.4%
CV (NSMSC)
 Intra-assay = 7.4%
 Inter-assay = 9.0%
*****
Shirazi 2016 Tertile cut points
≤ 76 nmol/L
≥77–≤97 nmol/L
≥98 nmol/L

≤ 30.4 ng/mL
≥ 30.8–≤38.8 ng/mL
≥39.2 ng/mL
No data Yes – covariate in statistical analysis Mean 7.0 years (SD = 3.8) before diagnosis No HPLC
CV@ 70 nmol/L = 8.5%
CV@ 210 nmol/L = 7.1%
****
Yao 2011 Apriori cut points
<26.2 ng/mL
>26.2 ng/mL
No cases below LOD Yes – used season standardized 25(OH)D values Prior to surgery or any adjuvant treatment (27-day average between diagnosis and blood draw) No DiaSorin Liaison CLIA
CV = 8.8%
****
Yao 2014 Apriori cut points
<20 ng/mL
20–29.9 ng/mL
≥30 ng/mL
No cases below LOD Yes – covariate in statistical analysis Prior to surgery or any adjuvant treatment No DiaSorin Liaison CLIA
CV = 8.8%
****

Abbreviations: CLIA – Chemiluminescent Immunoassay, CV – Coefficient of Variation, HPLC – High Performance Liquid Chromatography, ID-LC-MS/MS – Isotope Dilution Liquid Chromatography tandem Mass Spectrometry, LC-MS/MS – Liquid Chromatography tandem Mass Spectrometry, PBA – Protein Binding Assay, RIA – Radioimmunoassay.

+

Quintile values were provided through communication with the author and did not appear in the published article.