Table 3.
Responsivity of breast milk mineral, amino acid and protein content to maternal diet
| Ref | Participants Risk of Bias |
Study | Ethnicity | Breast milk collection timing | Effects on breast milk content |
|---|---|---|---|---|---|
| Iodine | |||||
| Experimental studies | |||||
| Leung, 2012 | Healthy lactating women over 3 months PP USA RCT High risk of bias |
600 µg oral potassium iodide (456 µg iodine) after overnight fast (n 16). Iodine measured in BM at baseline and hourly for 8 h after intake and dietary iodine recorded | Not provided | Not provided | Following supplementation, there was a significant median increase in BM iodine levels (280·5 μg/l; IQR 71·5–338·0) above baseline (P < 0·01); the median peak iodine was 354 μg/l (IQR 315–495) |
| Mulrine, 2010 | Healthy lactating women after delivery New Zealand RCT Low risk of bias |
Placebo (n 56), 75 µg iodine/d (n 27) or 150 µg iodine per d (n 26) for 24 weeks. BM iodine measured at 1, 4, 8, 12, 16, 20 and 24 weeks | Most likely White Caucasian | All BM samples collected from 09.00 and 12.00 | Iodine decreased by 40 % over 24 weeks in placebo group (P < 0·001) but was 1·3 times higher in 75 µg supplemented group (P = 0·003) and 1·7 times higher in 150 µg supplemented group (P < 0·001) |
| Nazeri, 2016 | Healthy lactating mothers 3–5 d PP Iran RCT Low risk of bias |
Either 150 µg iodine/d (n 42), or no supplementation, but recommendation to used only iodised salt for cooking (control) (n 42). Iodine measured at 0, 7, 10, 14 and 30 d | Persian | Manual expression but no information on timing | At baseline: Median (IQR) 176·0 µg/l (133·7–218·7 µg/l) in supplemented group and 215 µg/l (168·5–315·5 µg/l) in control group, (P = 0·027). d7:191·0 µg/l (105·0–245·0) in supplemented group and 176·0 µg/l (140·0–286·0 µg/l) in control; d10:217·0 µg/l (148·7–339·0 µg/l) in supplemented group and 162·0 µg/l (120·0–206·5 µg/l) in control; d14:242·0 µg/l (156·2–355·7 µg/l) supplemented group and 160·0 µg/l (115·2–199·2 µg/l) in control; d30:210·0 µg/l (100·0–286·0 µg/l) supplemented group and 142·0 µg/l (92·2–197·2 µg/l) in control |
| Observational study | |||||
| Ureta-Velasco | Milk donors Spain Obs Good quality 8 |
BM iodine level from milk donors (n 113) measured and analysed together with five dietary records | Caucasian | Not provided | Iodine positively correlated with total iodine intake (rho = 0·0499, P < 0·001), but not iodine intake from food only (rho = 0·046, P = 0·628). Iodine positively correlated with consumption of > three dairy products per d |
| Se | |||||
| Experimental studies | |||||
| Dodge, 1999 | Healthy lactating women from delivery New Zeeland RCT Some concerns |
50 µg Se per d (n 12), or placebo (n 10) during pregnancy and lactation for 3 months. | Caucasian | First morning milk was collected using a breast pump | Se increased by 37 % (P = 0·003) following supplementation, glutathione activity was unchanged. Supplementation increased PUFA levels by 41 %; (placebo: mean = 9·7, sd = 1·24, v. supplemented: mean = 13·70, sd = 1·02), including LA and ARA, all P ≤ 0·05. SFA levels were correspondingly decreased by 11 % (placebo: mean = 50·1, sd 2·0, v. supplemented: mean = 44·4, sd = 1·6 g/100, P ≤ 0·04). Fatty acid values all expressed as g/100 g fatty acids |
| Dylewski, 2002 | Healthy lactating women 3 months PP USA RCT High risk of bias |
20 µg Se per d for 3 months (n 23). Dietary data and milk at 3 and 6 months | Not provided | Not provided | BM mean (sd) Se dietary intake was 111 (40) µg/d and did not change over the study. Supplementation significantly increased levels by 41 % from 3 (23 (7) ng/ml) to 6 months (32 (14) ng/ml), P ≤ 0·01 |
| Trafikowska, 1996 | Healthy lactating women 3–4 weeks PP Poland RCT High risk of bias |
Subjects (n 16) supplemented with 200 µg/d Se in the form of yeast-rich-Se for 3 months | Not provided | Not provided | After 1 month of Se supplementation, the Se concentration in milk increased significantly (P < 0·001) by 73 % to a plateau of 14–16 ng/ml |
| Trafikowska, 1998 | Healthy lactating women 3–5 weeks PP Poland RCT High risk of bias |
3 groups; group 1 (n 24) supplemented with 200 µg/d Se (yeast-rich Se, group 2 (n 30) supplemented with 200 µg/d selenite mixed with baker’s yeast, group 3 (n 13) supplemented with plain brewer’s yeast without Se. Supplementation lasted 3 months | Not provided | BM collected by manual expression prior to the first morning feeding | Baseline BM (sd) Se 8·9 (2·8) µg/l. In the control group, it remained constant during the 3-month period. Group 1 and 2; BM Se increased significantly reaching a plateau of 14–16 µg/l after 1 month of supplementation. The difference was significantly higher than controls in the yeast-rich Se (P < 0·001) and the selenite-Se-supplemented group (P < 0·01) |
| Observational studies | |||||
| Bianchi, 1999 | Healthy lactating mothers < 210 d PP Brazil Obs Fair quality 6 |
BM from (n 30) mothers, 24-h recall FFQ. BM collected at 7 d and 270 d. BM Se analysed | Nor provided | Not provided | No significant correlation between BM Se and maternal BMI (r2 = −0·0654, P = 0·7351); Se intake (r2 = −0·103, P = 0·594) and stage of lactation (r2 = −0·2981, P = 0·1095) |
| Valent, 2011 | Healthy lactating mothers 3 months PP Italy Obs Good quality 8 |
BM from (n 100) mothers, semi-structured dietary questionnaires. BM Se analysed. | Not provided | Milk collected at any time | BM Se significantly correlated with current fresh fish consumption, r = 0·21, P = 0·04. No significant correlation between intake of multivitamin supplements during pregnancy and Se in BM (mean 10·9 (3·3) ng/g, median 9·6 v. mean 12·3 (2·9) ng/g. median 11·4 for women who did not consume supplement, P = 0·11) |
| Zn/Cu/Fe | |||||
| Choi, 2016 | Healthy women 5–15 d PP Korea Obs Good quality 9 |
n 79 participants completed 3 d dietary record. | Asian | Not provided | Mean (sd) Fe significantly higher in BM from those taking daily Fe supplements (7·36 (9·10) mg/ml, n 64) v. those not (2·83 (6·36) mg/l, n 15), (P = 0·002). No significant difference from those taking daily Zn supplementation (0·36 (0·18) mg/l, n 64) v. those not (0·40 (0·17) mg/l, n 15). No significant difference from those taking daily Cu supplementation (0·69 (0·27) mg/l, n 64) v. those not (0·70 (0·22) mg/l, n 15) |
| Leotsinidis, 2005 | Healthy lactating mothers 3 d PP Greece Obs Fair quality 5 |
BM samples collected (n 180) to measure Cd, Cu, Fe, Pb, Mn and Zn at 3 and 14 d PP, FFQ | Not provided | Morning BM 2 h after previous breast-feeding | Mean (sd) values of colostrum samples: Cd, 0·190 (0·150) µg/l; Cu, 381 (132) µg/l; Fe, 544 (348) µg/l; Pb, 0·48 (0·60) µg/l; Mn, 4·79 (3,23) µg/l; Zn, 4905 (1725) µg/l. All metals with exception of Cu were found in lower concentrations in transitory milk |
| Vuori, 1980 | Healthy lactating mothers 6–8 weeks PP Finland Obs Fair quality 6 |
BM samples collected to measure Fe and Zn at 6–8 weeks PP and 17–22 weeks PP, 7 d FFQ n 15 |
Not provided | BM collected at beginning and end of each breast-feeding | Correlation coefficient (r) between enery intake (Kcal) and BM minerals (mg): Fe, 0·478 (P < 0·01) and Zn, 0·554 (P < 0·01) |
| Fe | |||||
| Experimental studies | |||||
| Yalcin, 2009 | Healthy lactating women 2 weeks PP Turkey RCT Low risk of bias |
Placebo (n 23) or 80 mg/d Fe (n 24) for 16 weeks. BM Fe and Zn at baseline, and Fe at 2 and 16 weeks | Turkish | Collection in morning before feeding infant, 2 h after previous breast-feeding | Fe supplementation to non-anaemic women did not change BM Fe content. BM mean (sd) supplemented Fe concentration was 579 (219) µg/l in week 2 and 372 (163) µg/l in week 16; whereas placebo 512 (178) µg/l week 2, and 385 (207) µg/l, in week 16 |
| Tyrosine | |||||
| Experimental studies | |||||
| Downlati, 2014 | Healthy lactating women from 2 to 24 months PP Canada RCT Some concerns |
Single dose 0, 2, 5 and 10 g oral tyrosine (n 24). Free and total tyrosine measured before and 2, 4 and 6 h after supplementation | Not provided | Not provided | Significant rise only in free tyrosine. 10 g of tyrosine group had significantly higher free tyrosine concentration compared with other groups (P < 0·001). Peak free tyrosine in BM after 10 g dose occurred at 4 h, whereas for the 2 g and 5 g tyrosine doses, maximum free tyrosine levels occurred at 6 h |
| Protein, amino acids, ovalbumin | |||||
| Experimental studies | |||||
| Forsum, 1980 | Healthy lactating women 13–20 weeks PP Sweden RCT Some concerns |
4 d of low protein diet, 1 d wash out and then 4 d of high-protein diet (n 3) | Not provided | Milk collected before and after each nursing by hand expression or manual pump | Mean (sd) protein content in BM (g/d) for low protein diet v. high protein diet: true protein 7·31 (0·74) v. 8·83 (0·44) (P < 0·05); lactoferrin (g/24 h), 2·52 (0·17) v. 3·01 (0·36); α-lactalbumin (g/24 h) 1·50 (0·20) v. 1·75 (0·12); lactose (g/24 h), 58·1 (13·2) v. 63·5 (5·6). Differences in lactoferrin, α-lactalbumin and lactose between the two groups not significant |
| Metcalfe, 2016 | Healthy lactating women first 6 weeks PP Australia RCT High Risk of Bias |
Groups comprised high-egg diet (> 4 eggs per week, n 40), low-egg diet (1–3 egg per week, n 44), and egg-free diet (n 36). BM measured at 2, 4 and 6 weeks. | 85 % Caucasian | BM collected between 2 and 6 h after previous breast-feeding | Mean egg consumption associated with ovalbumin concentration, whereby each additional egg ingested per week led to 25 % increase in ovalbumin levels (95 % CI 5, 48 %, P = 0·01). Ovalbumin significantly higher in high egg group compared with egg-free group. One third of women had no ovalbumin detected |
| Palmer, 2016 | Healthy lactating women 11–14 weeks PP Australia RCT Some concerns |
Groups comprised, no egg, one raw egg, half a cooked egg and one cooked egg (all n 41). BM samples collected every 2 h for 8 h. | Not provided | Not provided | Direct dose–response between amount of cooked egg ingested and peak of ovalbumin in BM (no egg 0·05 ng/ml (95 % CI 0·01, 0·1), half a cooked egg 2·24 ng/ml (95 % CI 0·57, 3·91), one cooked egg 3·16 ng/ml (95 % CI 1·41, 4·91), P < 0·05. No difference between raw and cooked eggs. No ovalbumin detected in BM of 24 % of women |
| Observational studies | |||||
| Rana, 1986 | Healthy lactating women 4–6 weeks PP England Obs Poor quality 2 |
Group 1 followed vegan diet for a year, no medication, no contraceptives (n 14), group 2, control (n 14) omnivore diet. FFQ 7 consecutive d | Not provided | Not provided | Mean taurine concentration in the vegan BM 35 mg/l was significantly lower than in BM of omnivores (55 mg/l) (P < 0·01) |
| Choline | |||||
| Observational studies | |||||
| Perrin, 2020 | Healthy lactating women > 2 weeks PP USA Obs Fair quality 7 |
FFQ to classify as non-vegetarian, vegetarian or vegan (n 74). Single BM sample measured free choline, PC and GPC. | Not provided | Samples collected in morning during first or second breast-feeding of day and last 2 h after previous breast-feeding | Wide range in free choline (4–301 mg/l), with no significant differences between groups. Significantly higher GPC in vegan (mean = 62·7 mg/l, sd = 25·3 mg/l) than vegetarian (mean = 47·7 mg/l, sd = 21·2 mg/l) and non-vegetarian (mean = 42·4 mg/l, sd = 14·2 mg/l), P = 0·005. Significantly lower PC in vegan (mean = 32·5 %, sd = 18·3 %) than vegetarian (M = 46·1 %, sd = 18·3 %) and non-vegetarian (Mean = 44·8 %, sd = 15·7 %), P = 0·01. |
RCT, randomised control trial; PP, postpartum; BM, breast milk, obs, observational study; IQR, interquartile range; GPC; glycerol-phosphocholine; PC, phosphocholine.