Table 2.
Responsivity of breast milk vitamin content to maternal diet
| Ref | Participants Risk of bias |
Study | Ethnicity | Breast milk collection timing | Effects on breast milk vitamin content |
|---|---|---|---|---|---|
| Vitamin A | |||||
| Experimental studies | |||||
| Bahl, 2002 | Healthy women 18–42 d PP Ghana, India and Peru RCT Some concerns |
Single dose (60 mg) retinyl palmitate (n 322) or control (soyabean oil, n 309). Milk retinol levels at 0, 2, 6 and 9 months | Hispanic, Indian and African | Usually collected between 09.00 and 12.00 | Significantly higher retinol in treatment group at 2 months only (difference in means 7·1 nmol/g fat, 91 % CI 3·4, 10·8, P < 0·05) |
| Basu, 2003 | Healthy lactating women after delivery India RCT High Risk of bias |
Single dose (209 μmol, 200 000 μg) retinol within 24 h of delivery (n 139), and untreated control group (n 132). Milk retinol levels at 0 and 24 h, and 1–6 months | Asian | Colostrum collected by manual breast pump, no information in the collection timing | Treatment group had significantly higher retinol levels up to 4 months (P < 0·01) |
| Canfield, 2001 | Lactating women 7 months PP Honduras RCT Some concerns |
90 mg β-carotene as red palm oil (n 32), β-carotene supplements (n 36) or placebo (n 18). Six doses over 10 d. Milk retinol, lutein, β-cryptoxanthin, lycopene, α- and β-carotene at 0 and 10 d | Hispanic | Mid-morning collection by manual expression | No significant difference in retinol, but palm oil supplementation led to greater increases in lutein, lycopene, α- and β-carotene v. control. Increases in β-carotene concentrations were greater for the palm oil group (2·5-fold, P < 0·0001) than for the β-carotene concentrations supplement group (1·6-fold, P < 0·006) relative to placebo |
| Canfield, 1998 | Healthy lactating women < 6 months PP USA RCT High Risk of bias |
Participants received a single dose of 60 mg (group 1, n 6) or 210 mg (group 2, n 6) of β-carotene. Milk retinol. α-Tocopherol and carotenoids were monitored for 8 d | Not provided | Mid-afternoon | Data show that a single 60 mg supplement of β-carotene sustained elevated β-carotene concentrations in milk for > 1 week in healthy mothers but did not affect concentrations of other carotenoids, retinol and α-tocopherol β-Carotene concentration in BM group 1 v. group 2, mean (se): 36·1(5·5) nmol/g lipid v. 50·4 (16·8) nmol/g lipid. α-Carotene concentration in BM group 1 v. group 2, mean (se): 10·4 (2·4) nmol/g lipid v. 14·0 (5·0) nmol/g lipid Lycopene concentration in BM group 1 v. group 2, mean (s e): 18·7 (3·4) nmol/g lipid v. 34·7 (5·6) nmol/g lipid (P < 0·05) |
| Ding, 2021 | Healthy lactating women 30–45 d PP China RCT Low risk of bias |
Supplementation for 2 months with 1800 μg vitamin A and 600 μg vitamin D (n 117), or placebo (n 128) | Asian | BM collected between 07.30 and 09.00 through breast pump | After 2 months vitamin A in supplemented group was M = 1 (sd = 0·5) µmol/l v. control group M = 0·8 (sd = 0·5) µmol/l, P < 0·05. |
| Gossage, 2002 | Healthy lactating women < 32 d PP USA RCT Some concerns |
Subjects (n 21) received 30 mg/d of β-carotene or placebo from day 4 to day 32 PP. BM samples analysed for concentrations of carotenoids, α-tocopherol, and retinol. Eight diet records and eight BM samples | Two African American, two Hispanic and 17 European American | Not provided | No significant effects of β-carotene supplementation on BM concentrations of lutein, β-cryptoxanthin, lycopene or α-carotene. Milk concentrations for all four carotenoids decreased over time (P < 0·01 for all). Milk concentration in retinol andα- tocopherol were unaffected by supplementation and decreased over time (P < 0·0050. Mean (sd) retinol concentration were 4944 (539) µmol/l initially and 2079 (207) µmol/l at the end of the study. α-Tocopherol concentrations were 31 (4·6) µmol/l initially and 9·4 (1·2) µmol/l at the end of the study |
| Grilo, 2016 | Healthy puerperal women 24 h PP Brazil RCT Some concerns |
Single dose (60 mg, 200 000 μg) retinyl palmitate after first colostrum collection (n 30), and untreated control group (n 27). Milk retinol and α-tocopherol levels at 0 and 24 h and 30 d | Not provided | Collection after an overnight fast from 08.00 until 12.00 | Intervention significantly increased retinol levels at 24 h (+157 %, P < 0·001); however, retinol levels of both groups did not differ at 30 d. After supplementation, colostrum α-tocopherol decreased significantly (–16·4 %, P < 0·05), but no significant difference in α-tocopherol levels between groups at 30 d. |
| Johnson, 1997 | Healthy lactating women < 8 months PP USA RCT High Risk of bias |
Subjects given either placebo (n 4) or naturally occurring β-carotene (64 mg all-trans BC and 69 mg 9-cis BC; n 8) for 8 d | Not provided | Not provided | For supplemented group, significant increase in concentration observed at day 3 (P < 0·001) and steadily increased to six times the baseline level by the end of the supplementation period (day 8, P < 0·001). After 1 month, BM concentration of all-trans BC decreased but was still significantly higher than day 1 (P < 0·022). |
| B vitamins | |||||
| Experimental studies | |||||
| Chang, 2002 | Healthy lactating women < 1 month PP USA RCT High risk of bias |
Four groups of lactating mothers (n 47) receiving 2·5, 4, 7·5 and 10 mg/d of PN-HCl, 24 h dietary record. | Not provided | Not provided | BM vitamin B6 responsive to maternal supplementation from 1–6 months PP. Mean BM B6 significantly lower for women supplemented with 2·5 mg PN-HCl/d than for those supplemented with 4·0, 7·5 or 10·0 mg/d. Mean (sem) range from 1–6 months for groups supplemented with 2·5 mg/d (891 (29·9) to 1·316 (74·8) nmol/l), 4 mg/d (1184 (40·8) to 1944 (74·8)), 7·5 mg/d (1·752 (86·7) to 2278 (86·7)), and 10 mg/d (1704·3 (38·9) to 2338·1 (104·7) nmol/l). |
| Hampel, 2017 | Healthy lactating women 2–4 months PP Bangladesh RCT Low risk of bias |
3-d supplementation study (n 18). Day 1: fasting no supplement, day 2: one time the US Canadian RDA for vitamins, day 3: Twice US Canadian RDA for vitamins. BM vitamin A, B1, B2, B3, B6, B12 and E measured | Not provided | Breast milk collected at each feeding through a breast pump form the same breast for 24 h | BM vitamin A, B1, B2 and B6 significantly increased in BM-supplemented group, median increases > 180 % (B2 and B6) and 120–130 % for B1 and A. B3 and E levels significantly lower on days supplements were consumed (P < 0·05). No significant effect of supplementation for B12 |
| Nail, 1980 | Healthy lactating women at parturition USA RCT Some concerns |
Two groups, one (n 7) received vitamin B1 and B2 supplementation (B1 1·7 mg/d, B2 2·0 mg/d), second group (n 5) received no supplementation. BM samples after 1 and 6 weeks supplementation | Not provided | Not provided | BM mean (sd) B1 of both groups after 1 and 6 weeks, respectively were: non-supplemented 138 (18) µg/l and 220 (27) µg/l; and supplemented 133 (27) µg/l and 238 (21) µg/l. There were significant increases in B1 both groups (P < 0·05). BM mean (sd) B2 of both groups after 1 and 6 weeks, respectively, were non-supplemented 367 (128) µg/l and 485 (123) µg/l and supplemented 880 (168) µg/l and 710 (187) µg/l |
| Styslinger, 1985 | Healthy lactating women 2–3 months PP USA RCT High risk of bias |
0, 2·5, 10·0 or 20·0 mg pyridoxine-HCl for three consecutive days in addition to dietary sources (n 6). BM samples at baseline and 3 d | Not provided | Not provided | Significant positive correlation (r = 0·80, P < 0·001) between supplemental intake and vitamin B6 content |
| Thomas, 1979 | Healthy lactating women 1 week PP USA RCT High risk of bias |
No supplement (n 7), or multivitamin and multimineral supplement (n 10), containing 4 mg vitamin B6. BM measured for 3 d periods at 1 and 6 weeks | Caucasian | Milk samples expressed four times a day at 4 h interval | Vitamin B6 differed significantly (P < 0·05) at 5–7 d. At 43–45 d, content significantly increased (P < 0·05) in supplemented group. Content in supplemented group remained constant at 43–45 d, and the difference between groups at 43–45 d was not significant |
| Thomas, 1979 | Healthy lactating women 1 week PP USA RCT High risk of bias |
Non-supplement (n 7) and multivitamin and multimineral supplement (n 10), containing 8 μg of vitamin B12. BM measured for 3 d periods at 1 and 6 weeks | Caucasian | Milk samples expressed four times a day at 4 h interval | No significant differences between the groups at 5–7 d. Decline in both groups at 43–45 d. However, levels in non-supplemented group were significantly lower than both the 5 to 7 d levels. Mean (sd) values for non-supplemented were 1·22 (0·41) µg/l at 5–7 d and 0·61 (0·17) µg/l at 43–45 d, (P < 0·05), and supplemented group at 43–45 d PP; and supplemented were 1·65 (0·63) µg/l at 5–7 d, 1·10 (0·57) µg/l, (P < 0·05). |
| Vitamin C | |||||
| Experimental studies | |||||
| Byerley, 1985 | Healthy lactating women 11 weeks PP Canada RCT High risk of bias |
Five vitamin C groups (1) 0 mg, 0 mg 1 d + 90 mg per for 2 d, (3) 90 mg for 1 d + 250 mg per d for 2 d, (4) 90 mg for 1 d + 500 mg per d for 2 d, (5) 90 mg for 1 d + 1000 mg per d for 2 d, all n 5 | Not provided | BM collected at each feeding either manually or by use of breast pump | Mean vitamin C ranged from 44 to 158 mg/l but not correlated with intake and not significantly different between groups |
| Daneel Otterbech, 2005 | Healthy lactating women Switzerland and Republic of Ivory Coast 8 mPP RCT High risk of bias |
Five separate studies: 1. Baseline milk ascorbic acid levels European (n 142) and African (n 171) women. 2. 1000 mg ascorbic acid per d for 10 d, European (n 10) and African (n 18) women. 3. European women (n 17), 1000 mg ascorbic acid per d for 5 d, followed for 35 d. 4. African women (n 11), 100 mg ascorbic acid per d for 10 d. 5. 1, 3 or 5 servings of orange juice (100 mg ascorbic acid/serving) per week for six weeks, African women (n 15) |
Caucasian and African | All samples collected between 07.00 and 12.00 | BM vitamin C 50 % lower (P < 0·001) from African women. Supplementation (1000 mg/d for 10 d) increased levels in both African and European women, from 19 to 60 mg/kg (P < 0·001) and 60 to 70 mg/kg (P < 0·03), respectively |
| Thomas, 1979 | Healthy lactating women 1 week PP USA RCT High risk of bias |
Non-supplement (n 7) and multivitamin and multimineral supplement (n 10), containing 90 mg vitamin C from parturition. BM analysed for 3 d periods at 1 and 6 weeks | Caucasian | BM samples expressed four times a day at 4 h intervals | No significant differences in levels between groups at any time points |
| Vitamin D | |||||
| Experimental studies | |||||
| Ala-Houhala, 1988 | Healthy lactating women 8 weeks and 20 weeks PP USA RCT High risk of bias |
Daily supplementation for 8 weeks (Winter) and 15 weeks (Spring) with 2000 μg (n 15) vitamin D, 1000 μg vitamin D (n 15), or no supplementation (n 15) | Not provided | Morning milk | Oral maternal supplementation of vitamin D had no significant effect on milk vitamin D levels, but 25(OH) D levels of mothers receiving either 1000 or 2000 μg (25 or 50 micrograms) vit D/d were significantly higher than those of non-supplemented mothers in February and April |
| Basile, 2006 | Healthy lactating women 1 month PP USA RCT Low risk of bias |
Subjects received (n 12) 2000 μg/d vitamin D or (n 13) 4000 μg/d for 3 months. BM samples collected to measure vitamin D | 16 White and 9 African-American | Not provided | 25 (OH) D increased from 1 to 4 months in both group (mean (sd)): (+11·5 (2·3)) ng/ml for group 2000 (P = 0·002) and (+14·4 (3·0)) ng/ml for group 4000 (P = 0·0008). The 4000 μg/d regimen was more effective in raising BM antirachitic activity than the 2000 μg/d supplementation. Decline in BM was not associated with vitamin D dose (P = 0·73) or maternal 25(OH)D (P = 0·94) |
| Ketha, 2018 | Healthy lactating women < 6 months PP USA RCT Low risk of bias |
Subjects (n 40) received either a single dose 150 000 μg or 5000 μg daily of vitamin D3 for 28 d. BM vitamin D measured at 1, 3, 7, 14 and 28 d. Outcome was the temporal changes in 24,25 (OH)2D3/25(OH)D3 ratio |
Not provided | Not provided | BM vitamin D3 values in the single-dose group were inversely associated with 24,25 (OH)2D3/25(OH)D3 ratio (r2 = 0·14, P < 0·001), but not with daily dosing |
| Niramitmahapanya, 2017 | Healthy lactating women < 6 weeks PP Thailand RCT Low risk of bias |
Subjects received either 800 μg/d vitamin D supplement (n 35) for 6 weeks or a placebo (n 33) | Asian | Not provided | BM vitamin D at baseline mean (sd) supplemented group v. non-supplemented group: 79·86 (18·27) nmol/l v. 88·33 (21·28) nmol/l, (P = 0·183). Vitamin D BM concentration at 6 weeks supplemented group v. non-supplemented group: 97·49 (19·32) nmol/l v. 88·92 (22·42) nmol/l, (P = 0·076) |
| Oberhelman, 2013 | Healthy lactating women < 6 months PP USA RCT Some concerns |
Single dose 150 000 μg cholecalciferol (n 20) or 5000 μg/d cholecalciferol (n 20) for 28 d. BM cholecalciferol and 25(OH)D measured on 0, 1, 3, 7, 14 and 28 d | Not provided | Not provided | BM mean cholecalciferol reached peak of 40 ng/ml at 1 d in single-dose group, whereas in the daily supplemented group levels remained at approximately 8 ng/ml from 3 to 28 d |
| Vitamin E | |||||
| Experimental studies | |||||
| Clemente, 2015 | Healthy lactating women 12 h PP Brazil RCT Low risk of bias |
Non-supplemented control (n 36), single dose 400 μg RRR-α-tocopherol (n 40), or single dose 400 μg all-rac-α-tocopherol synthetic (n 33). BM colostrum α-tocopherol measured 12 h PP and 24 h after supplementation | Not provided | Colostrum collected by manual expression, no information on timing | No change in control group at 24 h, whereas RRR α-tocopherol and all-rac α-tocopherol increased by 57 % and 39 %, respectively (significantly different between control group and α-tocopherol synthetic group and control group and α-tocopherol natural group, P < 0·001); significantly different between α-tocopherol synthetic group and α-tocopherol natural group, P = 0·04) |
| Gaur, 2017 | Healthy lactating mothers < 4–6 weeks PP USA RCT Low risk of bias |
3 groups (n 89), group 1 (n 29) received 45·5 mg all-rac-α-tocopherol acetate, group 2 (n 30) 22·8 mg all-rac-α-tocopherol acetate + 20·1 mg RRR-α-tocopherol, group 3 (n 30) 40·2 mg RRR α-tocopherol for 6 weeks | Not provided | Not provided | In group 3, % of RRR-α-tocopherol increased in BM (mean (sem): 78 % (2·3 %) compared with 82 % (1·7 %) (P < 0·05). In contrast, the % of RRR-α-tocopherol decreased in the group 2 (P < 0·05) and group 1 (P < 0·0001) |
| Kanno, 1989 | Healthy lactating mother 70 d parturition Japan RCT High risk of bias |
d-α-Tocopherol (1·1 g) in a capsule was orally administrated once with ice cream to a mother | Not provided | BM expressed 2–3 times daily with manual pump | The transfer of α-tocopherol into BM reached a maximum value of 414 µmol/100 g after 3 d and then declined to the baseline level after 5 d. The amount of α-tocopherol recovered in BM was 0·11 %. The α-tocopherol equivalent/PUFA ratio (mg/g) was increased from 0·25 to value between 0·7 and 1·7 |
| Pires Medeiros, 2016 | Healthy lactating women < 30 d PP Brazil RCT High risk of bias |
Non-supplemented control (n 51), of single dose 400 μg RRR-α-tocopherol (n 38). BM α-tocopherol measured after delivery, 24 h, 7 and 30 d | Not provided | Samples collected after an overnight fast | BM α-tocopherol increased by 60 %, 24 h after, but not control group (P < 0·001). At 7 d supplemented group levels 35 % higher than control group, but no difference at 30 d |
| De Souca Reboucas, 2019 | Healthy lactating women 30–90 d PP Brazil RCT Some concerns |
Single dose 800 μg (588 mg, RRR-α-tocopherol, (n 39), or non-supplemented control (n 40). BM α-tocopherol measured at supplementation and next day. | Not provided | BM collected manually from a single breast that had not been collected previously | No difference between control and supplemented groups at baseline. One day after supplementation, supplemented group levels significantly increased by 124 % (mean = 15·00 µmol/l, sd = 5·1 µmol/l, P < 0·001), with no change in control group (mean = 6·94 µmol/l, sd = 2·0 µmol/l) |
| Observational studies | |||||
| Antonakou, 2011 | Healthy lactating mothers < 1 month PP Greece Obs Good quality 8 |
BM samples (n 64), 3 d dietary record at 1, 3 and 6 months PP | Not provided | Morning hour BM collection by electric pump | BM mean (sd)α tocopherol was 8·3 (3·4) µmol/l, 8·1 (4·2) µmol/l and 8·5 (4·7) µmol/l at 1, 3 and 6 months PP, respectively; while total tocopherol values were 8·9 (3·6) µmol/l, 8·7 (4·6) µmol/l and 9·5 (5·6) µmol/l, respectively. No significant differences observed between the time points. Mean (sd) maternal vitamin E dietary intake was 7·2 (3·7) mg/d, 6·8 (3·5) mg/d and 10·9 (5·2) mg/d at 1, 3 and 6 months PP, respectively. Though, vitamin E dietary intake was lower than the recommended one. Correlation of dietary intake parameters with the concentration of vitamin E in mature milk at first month of lactation: total fat (% total fat), r = 0·244, P = 0·047, PUFA (% total fat) r = 0·092, P = 0·387; MUFA (% total fat) r = 0·195, P = 0·062 |
| Vitamin K1 | |||||
| Experimental studies | |||||
| Bolisetty, 1998 | Healthy lactating women with preterm births, 28–32 weeks RCT High risk of bias |
2·5 mg phylloquinone (vitamin K1) daily for 2 weeks (n 6). BM phylloquinone measured daily for 14 d | 4 Caucasian and 2 Asian | BM extracted either manually or with electric pump at 5 h intervals | Mean (sd) BM levels increased from baseline: 3 (2·3) ng/ml to 22·6 (16·3) ng/ml, (P < 0·05) after the first dose, with continual increase until plateau at 64·2 (31·4) ng/ml after sixth day |
| Greer, 1997 | Healthy lactating mothers < 3 d PP USA RCT Low risk of bias |
2 groups: either 5 mg of phylloquinone (n 11) or placebo (n 11), daily supplementation for 12 weeks. Placebo was glucose. BM collected after 2 weeks, 6 weeks and 12 weeks supplementation |
Not provided | Not provided | Mean (sd) BM vitamin K supplemented v. placebo: 1·10 (0·75) ng/ml v. 0·69 (0·39) ng/ml at baseline; 76·53 (26·98) ng/ml v. 1·17 (0·70) ng/ml after 2 weeks (P < 0·01); 75·27 (46·23) ng/ml v. 1·14 (0·46) ng/ml after 6 weeks (P < 0·01) |
| Von Kries, 1987 | 1 lactating mother Germany < 5 weeks PP RCT High risk of bias |
Single dose of 0·5 mg, 1 mg and 3 mg vit K1 (n 1 per group). Dose response (0·1 mg, 0·5 mg, 1 mg and 3 mg) measured over 24 h (n 1). Single dose study measured BM vitamin K1 over 50 h, and dose–response study measured over 24 h | Not provided | Complete expression of both breasts using an electric pump | 0·5 mg, 1 mg and 3 mg produced increases in milk levels, peaking at 12–24 h. Dose–response relationship observed, with the lowest dose (0·1 mg) producing 2-fold increase in vitamin K1 content. 100 µg dose of vitamin K1 raised BM vitamin K1 from 2·5 to 4·9 ng/ml after 16 h, this then declined to 1·9 ng/ml after 24 h |
PP, postpartum; RCT, randomised control trial; BM, breast milk; AM, ante meridiem; PM, post meridiem; M, median; Obs, observational study.