Collard greens: a vitamin K load that warfarin users should keep steady

In a trial of 70 warfarin users whose control was unstable, a steady 150 ug of vitamin K a day, taken as a supplement for 6 months, raised time in the target INR range by 28 percent against 15 percent on placebo. A multicentre Canadian trial of the same dose found no change in time in range (65.1 against 66 percent) and 4 percent fewer extreme INR swings. A systematic review of 2 diet trials and 9 observational studies found no support for telling warfarin users to cut vitamin K rich greens. It advised keeping intake steady, on evidence it called conflicting, and it spoke of leafy greens in general. That matters here because a cup of boiled collards (130 g) carries about 609 ug vitamin K in US survey data, scaled from a single 100 g figure. The same cup gives about 324 mg calcium, 5.6 g fiber and 44 calories.

How much is actually in it

• Raw collards sampled by USDA ranged from 170 to 340 mg calcium and from 64 to 111 mg vitamin C per 3.5 oz (100 g) across 8 samples. — USDA Foundation Foods, Collards, raw, 8 samples (n = 8) Evidence E source
  Calcium mean 276.4 mg (170-340); vitamin C 89.4 mg (64-111); folate 168 ug (118-305); potassium 409.8 mg (335-536)
• A laboratory analysis of an Italian collard cultivar measured 333 mg calcium per 3.5 oz (100 g) fresh leaves, above endive, lettuce and spinach tested in Italy. — collard green leaves, Couve-Manteiga cultivar, Italy Evidence D source
  Calcium 333.09 mg/100 g FW; potassium 215.53 mg; magnesium 29.45 mg

What a real portion looks like

• A cup of boiled collards (130 g) gives about 609 ug vitamin K, 324 mg calcium, 5.6 g fiber and 44 calories in US survey data. — USDA FNDDS, Collards, fresh, cooked, no added fat (n = 1 food code) Evidence E source
  Per 100 g: vitamin K 468.6 ug, calcium 249 mg, vitamin A 256 ug RAE, fiber 4.3 g, energy 34 calories

What your body absorbs

• When 5 men ate 3.5 oz (100 g) of isotope-labeled collard greens with a 24 g fat breakfast, blood vitamin K peaked at 6 to 9 hours and was back to baseline by 24 hours. — 5 men aged 26 to 71, single 3.5 oz (100 g) serving with 396 ug phylloquinone and 24 g fat (n = 5) Evidence C source
  Plasma phylloquinone peak 10.51 nmol/L at 6-9 h, baseline 1.26 nmol/L by 24 h; half-life 22.8 h (n = 3)
• In 36 younger and older adults, vitamin K from about 377 ug a day of broccoli raised blood levels as well as a fortified oil did, a proxy for other Brassica greens. — 36 younger and older men and women, 5 days of supplementation in 15-day metabolic unit periods, crossover (n = 36) Evidence B source
  Plasma phylloquinone rose and %ucOC fell after both broccoli and oil diets; no significant difference between them

What cooking and storage change

• For greens boiled in water to cover and drained, the USDA table keeps 55% of vitamin C and 60% of folate but 95% of calcium. — USDA R6 retention factors, group Vegetables and Vegetable Products, greens boiled, water cover, drained (n = 1 method) Evidence E source
  Retained: vitamin C 55%, folate 60%, thiamin 80%, potassium 85%, calcium 95%, iron 95%

What it does to your health

• Across 5 prospective studies with 80,982 people, the highest dietary vitamin K intake was linked to 22% fewer fractures than the lowest. — 4 cohorts and 1 nested case-control study, 80,982 participants, 1114 fractures (n = 80982) Evidence C source
  RR 0.78 (95% CI 0.56-0.99) highest vs lowest; RR 0.97 (0.95-0.99) per 50 ug/day
• In 356 people with advanced macular degeneration and 520 controls, eating spinach or collard greens more often was linked to substantially lower risk. — 356 cases with advanced AMD and 520 controls aged 55 to 80, five US eye centers (n = 876) Evidence C source
  Highest vs lowest carotenoid intake OR 0.57 (95% CI 0.35-0.92); spinach or collard greens P for trend < .001
• Among 584 older African-American women, eating collard greens or kale more than once a week was linked to 57% lower odds of glaucoma. — 584 older African-American women, Study of Osteoporotic Fractures (n = 584) Evidence C source
  OR 0.43 (95% CI 0.21-0.85) for >1 serving/week vs <=1 serving/month of collard greens/kale

Safety, limits and interactions

• In 70 warfarin users with unstable control, a steady 150 ug vitamin K a day for 6 months raised time in target INR range by 28% against 15% on placebo. — 70 warfarin-treated patients with unstable anticoagulation, 6 months (n = 70) Evidence B source
  SD of INR -0.24 vs -0.11 (P < .001); time in target range +28% vs +15% (P < .01)
• A Canadian multicentre RCT found 150 mcg vitamin K a day did not raise time in therapeutic INR range for warfarin users (65.1% vs 66%) but cut extreme INR swings. — patients on chronic warfarin therapy at four Canadian hospitals, 6 months Evidence B source
  TTR 65.1% vs 66% (p = 0.8); INR excursions absolute reduction 4% (95% CI 2 to 6%)

What people believe that the data does not support

• The NIH calcium fact sheet lists collard greens among foods high in oxalic acid, which can lower calcium absorption. — NIH Office of Dietary Supplements, Calcium fact sheet for health professionals Evidence E source
  Calcium absorption 5% from spinach vs 27% from milk; collard greens named as high in oxalic acid
• A systematic review of 2 dietary trials and 9 observational studies found no support for telling warfarin users to cut vitamin K rich greens, only to keep intake steady. — patients on vitamin K antagonists in 11 clinical studies (n = 11 studies) Evidence C source
  Conflicting evidence; effect on coagulation possibly only above about 150 ug/day dietary vitamin K

Who this works differently for

• In 3143 pregnant US women, a Southern diet pattern with collard greens alongside processed meats, fried foods and sugary drinks was linked to 55% higher odds of preterm birth. — 3143 pregnant women in the PIN study, diet assessed at 26-29 weeks (n = 3143) Evidence C source
  Adjusted OR 1.55 (95% CI 1.07-2.24) quartile 4 vs 1 of this pattern

What is still unknown

• A labeled kale study in 7 volunteers estimated only 4.7% of its vitamin K was absorbed, so estimates for leafy Brassica greens differ widely. — 7 healthy volunteers, single 14 oz (400 g) serving of 13C-labeled kale with 30 g vegetable oil (n = 7) Evidence C source
  Mean bioavailability of phylloquinone from kale 4.7% (n = 6); plasma peak 2.1 nmol/l at 6-10 h

The bottom line

The calcium number is high, and how much of it you absorb is open. Eight raw samples tested by USDA ranged from 170 to 340 mg calcium per 3.5 oz (100 g), and one Italian cultivar measured 333 mg in a single laboratory batch. The NIH calcium fact sheet lists collard greens among foods high in oxalic acid, which can lower calcium absorption. We found no direct measurement of calcium absorbed from collards, so how much that listing matters for them is untested. Vitamin K from collards does reach the blood. When 5 men ate 3.5 oz (100 g) of isotope labeled collards with a breakfast of 24 g fat, blood vitamin K peaked at 6 to 9 hours and was back to baseline by 24 hours. That small study had no control group and did not calculate the share absorbed. For leafy greens of the same family the estimates disagree. A labeled kale study in 7 volunteers put absorption at 4.7 percent, while in a crossover trial of 36 adults about 377 ug a day from broccoli raised blood levels as well as a fortified oil did. Neither food was collards. Boiling and draining keeps 95 percent of calcium and iron, 60 percent of folate and 55 percent of vitamin C, by a USDA factor for greens as a group that was not measured on collards. The health links are observational. Across 5 prospective studies with 80,982 people, the highest vitamin K intake from all foods was linked to 22 percent lower fracture risk than the lowest (RR 0.78, 95% CI 0.56 to 0.99). In a 1994 case-control study of 356 people with advanced macular degeneration and 520 controls, eating spinach or collard greens more often went with a substantially lower risk. Among 584 older African-American women, eating collards or kale more than once a week was linked to 57 percent lower odds of glaucoma (OR 0.43, 95% CI 0.21 to 0.85), from a single survey with self-reported diet. The main caution is warfarin and other vitamin K antagonists, covered above. If you take one, keep your greens steady from week to week and tell whoever manages your INR when that changes. In 3143 pregnant US women, a Southern eating pattern with collard greens alongside processed meats, fried foods and sugary drinks was linked to 55 percent higher odds of preterm birth (OR 1.55, 95% CI 1.07 to 2.24). A pattern like that cannot separate the greens from the rest of the plate.

Sources

1. clgr-r7-02 · composition dataset
   Collards, raw: Calcium, Ca 276.4 mg, range 170.0-340.0 (median 293.0), samples 8; Vitamin C, total ascorbic acid 89.4 mg, range 64.0-111.0 (median 87.0), samples 8; Folate, total 168.0 ug, range 118.0-305.0 (median 148.0), samples 8; Potassium, K 409.8 mg, range 335.0-536.0 (median 401.0), samples 8
   USDA FoodData Central, Foundation Foods, Collards, raw, release 2026-04-30 · checked 2026-09-29
2. clgr-r7-05 · food composition study
   Calcium (333.09 ± 1.020 mg/100 g FW) was the main mineral found in the analyzed samples, followed by potassium (215.53 ± 1.06 mg/100 g FW) and sulfur (108 ± 0.19 mg/100 g FW). ... The calcium content in 100 g of this product is significantly higher than that found in other green leafy vegetables commonly consumed in Italy, such as endive (93 mg/100 g FW), lettuce (45 mg/100 g FW), and spinach (78 mg/100 g FW).
   Nutrients, 2024 · checked 2026-09-29
3. clgr-r7-01 · composition dataset
   Collards, fresh, cooked, no added fat, FNDDS code 72107211, fdc_id 2709577, per 100 g: Vitamin K (phylloquinone) 468.6 ug; Calcium, Ca 249 mg; Vitamin A, RAE 256 ug; Fiber, total dietary 4.3 g; Energy 34 kcal; 1 cup 130 g: Vitamin K 609.2 ug; Calcium 323.7 mg; Fiber 5.6 g; Energy 44.2 kcal
   USDA FNDDS, food code 72107211, FDC 2709577 · checked 2026-09-29
4. clgr-r7-06 · feeding study
   A single bolus of 100 g (396 +/- 28 microg phylloquinone) deuterium-labeled collard greens was fed with a breakfast containing 24 g fat to 5 men (26 to 71 years). ... Plasma total phylloquinone concentrations peaked at 6 to 9 hours (10.51 +/- 4.38 to 8.30 +/- 4.64 nmol/L) and returned to baseline by 24 hours (1.26 +/- 0.38 nmol/L).
   Metabolism, 2004 · checked 2026-09-29
5. clgr-r7-07 · RCT
   In a crossover design with three 15-d residency periods in a metabolic unit, younger and older men and women (n = 36) consumed a mixed diet containing 100 microg phylloquinone/d. During 2 residency periods, the mixed diet was supplemented for 5 d with either broccoli (377 microg phylloquinone/d; broccoli diet) or phylloquinone-fortified oil (417 microg/d; oil diet). ... Overall, the response to broccoli supplementation was not significantly different from the response to the fortified oil in either age group.
   Am J Clin Nutr, 1999 · checked 2026-09-29
6. clgr-r7-03 · retention dataset
   VEG,GREENS,BOILED,WATER COVER DRAIN (Retn_Code 3005): vitamin_c_mg 55%; folate_total_ug 60%; thiamin_mg 80%; potassium_mg 85%; calcium_mg 95%; iron_mg 95%
   USDA Table of Nutrient Retention Factors, Release 6, 2007, Retn_Code 3005 · checked 2026-09-29
7. clgr-r7-12 · meta-analysis of observational studies
   Four cohort studies and one nested case-control study, with a total of 1114 fractures cases and 80,982 participants, were included in our meta-analysis. ... We observed a statistically significant inverse association between dietary vitamin K intake and risk of fractures (highest vs. the lowest intake, RR = 0.78, 95% CI: 0.56-0.99; I = 59.2%, P for heterogeneity = .04).
   Medicine (Baltimore), 2017 · checked 2026-09-29
8. clgr-r7-13 · case-control study
   In particular, a higher frequency of intake of spinach or collard greens was associated with a substantially lower risk for AMD (P for trend < .001).
   JAMA, 1994 · checked 2026-09-29
9. clgr-r7-14 · cross-sectional study
   For vegetables, >1 serving/week compared to ≤1 serving/month of collard greens/kale decreased the odds of glaucoma by 57% (OR = 0.43; 95% CI: 0.21-0.85).
   Am J Ophthalmol, 2012 · checked 2026-09-29
10. clgr-r7-10 · RCT
    Seventy warfarin-treated patients with unstable anticoagulation control were randomly assigned in a double-blinded fashion to receive a daily amount of 150 mug oral vitamin K or placebo orally for 6 months. ... Vitamin K supplementation resulted in a significantly greater decrease in standard deviation of international normalized ratio (INR) compared with placebo (-0.24 +/- 0.14 vs -0.11 +/- 0.18; P < .001) and a significantly greater increase in percentage time within target INR range (28% +/- 20% vs 15% +/- 20%; P < .01).
    Blood, 2007 · checked 2026-09-29
11. clgr-r7-11 · RCT
    There was no significant difference in the final TTR between the two groups (65.1 % vs 66 %, p =0.8). ... The number of INR excursions were significantly decreased in the LDVK group (9.4 % and 5.4 %, absolute difference [pre- minus post-] = 4 %, 95 % CI, 2 to 6 %, p-value <0.001).
    Thromb Haemost, 2016 · checked 2026-09-29
12. clgr-r7-04 · agency fact sheet
    Certain compounds in plants (e.g., oxalic acid, phytic acid) can decrease calcium absorption by forming indigestible salts with calcium [3]. As a result, absorption of calcium is only 5% for spinach, whereas it is much higher, at 27%, for milk [3]. In addition to spinach, foods with high levels of oxalic acid include collard greens, sweet potatoes, rhubarb, and beans [1].
    NIH ODS, Calcium Fact Sheet for Health Professionals · checked 2026-09-29
13. clgr-r7-09 · systematic review
    Two dietary interventional trials and 9 observational studies were included. We found conflicting evidence on the effect of dietary intake of vitamin K on coagulation response. ... The available evidence does not support current advice to modify dietary habits when starting therapy with VKAs. Restriction of dietary vitamin K intake does not seem to be a valid strategy to improve anticoagulation quality with VKAs. It would be, perhaps, more relevant to maintain stable dietary habit, avoiding wide changes in the intake of vitamin K.
    Medicine (Baltimore), 2016 · checked 2026-09-29
14. clgr-r7-15 · prospective cohort
    Increased odds of preterm birth were found for a diet characterized by a high consumption of collard greens, coleslaw or cabbage, red meats, fried chicken and fish, processed meats, cornbread or hushpuppies, eggs or egg biscuits, gravy, whole milk, and vitamin C-rich drinks such as Kool-Aid (Kraft Foods) and Hi-C (Minute Maid Co.) (adjusted OR for quartile 4 vs. quartile 1: 1.55; 95% CI: 1.07, 2.24).
    J Nutr, 2015 · checked 2026-09-29
15. clgr-r7-08 · feeding study
    Seven healthy volunteers ingested a single 400 g serving of kale with 30 g vegetable oil. ... Modelling results demonstrated a mean (n 6) bioavailability of phylloquinone from kale to be 4.7%.
    Br J Nutr, 2010 · checked 2026-09-29