- Email: [email protected]
Maternal-infant vitamin D relationships during breast-feeding
ORIGINAL ARTICLES
Maternal-infant vitamin D relationships during breast-feeding To evaluate the effect o f maternal dietary vitamin D intake on infant vitamin D status in a country with a temperate climate, but where the commercial milk supply is not vitamin D fortified, this randomized, double-blind study was conducted on term mother-infant pairs during the winter months. Well-nourished, white nursing mothers were given a placebo, 500 IU vitamin D/day or 1,000 IU vitamin D/day; their infants were not given supplemental vitamin D. After six weeks, mothers receiving supplemental vitamin D had higher levels o f 25-hydroxyvitamin D than had mothers receiving placebo. A direct relationship was observed between maternal and infant levels o f 25-hydroxyvitamin D at six weeks, implying that maternal vitamin D intake directly affects the vitamin D concentration in breast milk. A control group o f infants who had received 400 IU vitamin D/day had even higher concentrations o f 25-hydroxyvitamin D, suggesting that infant supplementation with vitamin D is more efficacious than maternal supplementation. Despite the favorable climate in South Africa, during winter breast-fed infants have low serum vitamin D values i f maternal dietary vitamin D intake is low.
Alan D. Rothberg, M.B., B.Ch., F.C.P.(SA),* John M. Pettifor, M.B., B.Ch., F.C.P.(SA), Ph.D.(Med), Desmond F. Cohen, M.B., B.Ch., D.C.H.(SA), Ernst W. W. Sonnendecker, M.B., B.Ch., M.Med.(O et G), F.R.C.O.G., and F. Patrick Ross, Ph.D.,
Johannesburg, South Africa
A QUESTION remains regarding the need for vitamin D supplementation in the term, exclusively breast-fed infant, with much of the controversy revolving around both the content and bioactivity of D in human milk. 1-6The results of two recent studies showed adequate 25-hydroxyvitamin D concentrations in nonsupplemented breast-fed infants during the first three to four months of life. l, 2 However, in both studies, all nursing mothers supplemented their daily
From the Departments o f Pediatrics and Obstetrics and Gynecology, Johannesburg Hospital, and the Metabolic and Nutrition Research Unit, University o f the Witwatersrand. Supported in part by grants from the South African Medical Research Council and the Atomic Energy Board. *Reprint address: Department of Pediatrics, Johannesburg Hospital, Private Bag X39, Johannesburg 2000, South Africa.
500
The Journal of PEDIATRICS Vol. 101, No. 4, pp. 500-503
diets with D, and most presumably drank D-fortified milk. The American Academy of Pediatrics Committee on Nutrition has recently commented on the role of maternal intake of D in fortified cow milk in preventing rickets in Abbreviation used D: vitamin D 25-OHD: 25-hydroxyvitamin D (calcidiol) Ca: calcium PO4: phosphate breast-fed infants2 In most countries of the world cow milk is not routinely supplemented with D. The present study was therefore undertaken to evaluate the need for supplemental D in term, breast-fed infants in a country with a temperate climate but where the milk supply is not D fortified. The absence of D fortification also enabled us to control maternal dietary intake and to examine the relationship between maternal D intake and infant D status.
0022-3476/82/100500+04500.40/0 @ 1982 The C. V. Mosby Co.
Volume 101 Number 4
Maternal-infant vitamin D relationships in breast-feeding
50 1
Table I. Maternal blood values (mean _+ S D) 6 wk Day 4: all groups (n = 40)
Maternal blood 25-OHD (ng/ml) Ca (mmol/L) PO~ (mmol/L) Alkaline phosphatase (1U/L)
11.9 2.1 1.4 111
I I I
• 6.0 _+ 0.1 • 0.4 • 24
Placebo (group P, n = 10)
500 U/day (group A, n = 9)
10.0 2.5 1.4 61
13.8 2.4 1.4 69
• 5.5 _+ 0.3 • 0.1 • 17
1,000 U/day (group B, n=9)
_+ 5.4 + 0.1 • 0.2 • 17
14.7 2.5 1.4 74
• • • •
4.0 0.3 0.1 24
Control (group C, n = 12) 11.0 2.3 1.3 75
_+ 3.5 -+ 0.1" _+ 0.2 • 16t
Conversion of SI to traditional units: serum Ca, I mmol/L ~ 4 rag/100 ml; PO4, 1 mmol/L .~ 3.1 mg/100 ml. *All six-weekvalues > baseline (P < 0.05). tAll six-weekvalues < baseline (P < 0.01).
Table II. Blood values in infants (mean + SD) 6 wk Infant blood 25-OHD(ng/ml) Ca (mmol/L) PO4 (mmol/L) Alkaline phosphatase (1U/L)
Day 4: all groups (n = 40}
(group P, n-- lO)
8.9 2.2 2.1 122
1.I 2.4 2.2 267
+ 7.1 • 0.3 __+0.3 • 30
-- 1.4 _+ 0.1" + 0.3 _+ 106
(group ,4. n=9) 10.2 2.4 2.3 325
-+ 5.5 _+ 0.2 + 0.2 • 75
(group B, n=9) 9.4 2.3 2.4 318
• 2.I + 0.1 -+ 0.2 • 98
(group C, n=12) t5.2 2.4 2.3 286
_ 3.7 _+ 0.1 _+ 0.2t _+ 71:~
*Six-week value > baseline for placebo group (P < 0.05) by paired t test. ~'Six-weekvalue > baseline for Control subjects (P < 0.05) by paired t test. ~:AI1six-weekvalues > baseline (P < 0.001) by paired t test.
MATERIALS
AND METHODS
This randomized, double-blind study was conducted on term mother-infant pairs at the Johannesburg Hospital, between the months of June (midwinter) through August, 1981. Seventy-seven well-nourished white nursing mothers and their infants were enrolled. Mothers were give n placebo (group P, n = 20), 500 I U D / d a y (group A, n = 20), or 1,000 I U D / d a y (group B, n = 20) for six weeks following delivery. These mothers received no additional D in other vitamin preparations or milk, and their infants did not receive supplemental D. A control group of mothers (group C, n = 17) received no D, but their infants were given 400 I U D / d a y . Mothers received only iron and folate supplements during pregnancy. Serum calcium, phosphorus, alkaline phosphatase, and 2 5 - O H D were measured in mothers and babies on the fourth postnatal day and six weeks later, A t the conclusion of the study all infants were given 400 I U supplemental D daily. After completion of the biochemical analyses, the code was broken, and the results were analyzed. Serum Ca was measured by atomic absorption speetrophotometry. Alkaline phosphatase and PO4 were measured
by standard techniques by means of a Technicon AutoAnalyzer II. Serum 2 5 - O H D was measured by the method of Haddad and Chyu. 7 The limit of sensitivity of this method was 2 ng/ml. In our laboratory the normal pediatric range for 2 5 - O H D is 8 to 40 n g / m l . The two-tailed paired and non-paired Student t tests were used in the analysis of data. Informed consent was obtained, and the study was approved by the hospital ethics committee.
RESULTS O f the 77 infants, 40 (57%) were still being completely breast-fed at six weeks and thus form the study group (group P, n = 10; group A, n = 9; group B, n = 9; group C, n -- 12). Maternal data (Table I). There were no intergroup differences for 25-OHD, Ca, PO4 or alkaline phosphatase on day 4. After six weeks, there were again no intergroup differences for Ca, PO4, or alkaline phosphatase; however, Ca concentrations had increased significantly from baseline values (P < 0.05), and alkaline phosphatase had decreased (P < 0.01). The mean 2 5 - O H D concentration for nonsupplemented mothers (groups P + C) at six weeks
502
Rothberg et al.
The Journal of Pediatrics October 1982
-E c
30
C)
c 25 E 0 20
9~
o
~---..o Mothers r =-.41 p = <0'05
3O
~
25
Infants
08 o
s
15
"-~-.*.,~oo
9
9
r =-.44
s
p =<005
.-~ 20 E
o
-- Day 4 (Placental transfer) r = 0.50 p= < 0 0 1
~ - - - o 6 Weeks (Milk transfer) r=0.58 p = < O.OI
i
?
Cg
E
A~.
5
o o
9
0 9
0
9 9
9
*
t-"
o
I
U')
0 0
l 10
June I , July, ! ' August 20 ~0 10 20 5 0 I0 20 Date of entry into study (4th postnatal day)
I
50
c~
5
_~
0
0
5
iO
t5
2o
25
50
Maternal 25-hydroxy vitamin D (ng/ml)
Fig. 1. Relationship between month ofdelivery and maternal and infant 25-OHD concentrations.
Fig. 2. Relationship between maternal and infant 25-OHD concentrations at four days and at six weeks.
Table IlL Meteorologic data
trations of 25-OHD than those infants in the maternally supplemented groups (P < 0.01). At six weeks, 25-OHD concentrations in group A, B, and~P infants were significantly related to maternal 25-OHD values (P < 0.01; Fig. 2).
Mean maximum daytime temperature 1981"
April May June July August
(~ 25.9 21.7 19.7 21.5 23.3
Mean hr sunlight per day
9.2 8.4 9.1 9.5 7.5
*Data from the South African Weather Bureau (Department of Transport).
was 10.6 _ 4.3 ng/ml, significantly lower than that of supplemented mothers, 14.0 _+ 4.5 ng/ml (groups A + B), P < 0.05. Mothers delivering in June had significantly higher 25-OHD values than those delivering in August (P < 0.05; Fig. 1). Infant data (Table II). There were no intergroup differences for 25-OHD, Ca, PO4, or alkaline phosphatase on day 4. The day 4 25-OHD concentrations in the neonate were directly related to the month of entry into the study (P < 0.05; Fig. 1) and to the maternal 25-OHD concentrations (P < 0.01; Fig. 2). After six weeks there were no intergroup differences for Ca, PO4, or alkaline phosphatase despite the finding of low 25-OHD concentrations in the placebo group. Group P infants had a mean 25-OHD concentration significantly below that of infants in the other groups ( P < 0.01), although no biochemical evidence of rickets was observed. Control infants (group C) had significantly higher concen-
DISCUSSION The mean day 4 maternal 25-OHD concentration in our study (11.9 +-6.0 ng/ml) was lower t h a n the thirdtrimester concentrations previously observed in white American mothers during the winter months (15.4 +_ 5.9 ng/ml), 8 possibly as a consequence of lower dietary D intake in our mothers. This finding~would be compatible with the conclusion of Hi!lman and Haddad g that, although ultraviolet exposure was the major determinant of maternal serum 25-OHD concentration, dietary D intake was important during periods of decreased exposure. Our mothers' diets were not supplemented with D during pregnancy, and they did not receive D-fortified milk. Despite the generally favorable climate in Johannesburg (Table III), when day 4 maternal 25-OHD concentrations were analyzed in terms of date of delivery, we observed a progressive decrease from June through August (Fig. 1). Maternal concentrations of 25-OHD correlated with concentrations in the infants at birth as reported previously.6,9 No infant had clinical or biochemical sequelae of low serum 25-OHD values during the immediate neonatal period. We have demonstrated a relationship between maternal D supplementation and maternal 25-OHD concentrations at six weeks. Furthermore, maternal 25-OHD concentrations correlate directly with infant 25-OHD values over the
Volume 101 Number 4
Maternal-infant vitamin D relationships in breast-feeding
same period, implying that maternal D intake directly affects the D concentration in human milk and thus the D status of the infant. This has been suggested previously~~ but could not be confirmed in a recent study,~ perhaps because of the small sample size. Maternal and infant 25-OHD concentrations were unaffected by doubling the maternaI D supplementation from 500 IU to 1,000 IU/day, whereas infants receiving 400 IU/day directly had significantly higher values at six weeks. Supplementation of the infant's diets is apparently a more efficacious method of maintaining adequate D status than supplementation of the maternal diet. These data show that in winter, in the nonsupplemented mother, milk concentrations of 25-OHD and transfer of D or one of its metabolites to the infant are low, even in a healthy, well-nourished, white population living in a favorable climate. Thus, D supplementation to mothers or infants would seem prudent, since under the study conditions, nonsupplemented infants had extremely low serum concentrations of D within six weeks. Recommendations based on the results of studies into the need for infant D supplementation in countries where milk is D fortified may not be relevant for populations not consuming Dfortified milk or milk products. 2'3 Certainly, a reductionin the incidence of rickets has been associated with milk fortification.I~ Recent American reports of nutritional rickets have included breast-fed infants of strict vegetarians and food faddists who have excluded milk from their diets] thus emphasizing the importance of D fortification in the prevention of rickets during breast-feeding. Similarly, the role of vitamin preparations taken by the mother during the postnatal period Should be considered when an attempt is made to answer the question of the need to supplement the diet of the breast-fed infant.2 Indeed, in two recent studies, nonsupplemented infants of mothers who received at least 400 IU D/day had serum 25-OHD concentrations in the range of 20 ng/ml at six weeks. I,~ Despite the absence of biochemical evidence of rickets, even in those infants with very low 25-OHD concentrations, the present study has demonstrated the potential for the development of rickets during Winter in nonsupple-
503
mented infants. However, it is possible that 25-OHD concentrations rise fairly rapidly in these infants after the first six weeks of life, as sun exposure of mothers and infants increases. Greater sun exposure of mothers and infants may explain the difference between our findings and those of Birkbeck and Scott, ~ who studied mothers and breast-fed infants in New Zealand. Pending further studies, we recommend routine D supplementation of breast-fed infants whose mothers have low intake of Vitamin D, irrespective of climate conditions. We thank Mr. Gopal Moodley and Miss M. Pentopoulosfor the biochemical evaluations. REFERENCES
1. Greer FR, SearcY JE, Levin RS, Steichen JJ, Asch PS, and Tsang RC: Bone mineral content and serum 25-hydrovitamin D concentration in breast-fed infants with and without supplemental vitamin D, J PEDIATR98:696, i981. 2. Roberts CC, Chan GM, Folland D, Rayburn C, and Jackson R: Adequate bone mineralization in breast-fed infants, J PEDIATR99:192, 1981. 3. Finberg L: Human milk feeding and vitamin D supplementation-1981, J PEDIATR99:228, 1981. 4. Greer FR, Ho M, Dodson D, and Tsang RC: Lack of 25-hydroxyvitaminD and 1,25-dihydroxyvitaminD in human milk, J PEDIATR99:233, 1981. 5. Nutrition and lactation--Committee on Nutrition, American Academy of Pediatrics, Pediatrics 68:435, 1981. 6. Birkbeck JA, and Scott HF: 25-hydroxycholecalciferolserum levels in breast-fed infants~ Arch Dis Child 55:691, 1980. 7. Haddad J, and Chyu K: Competitive protein-binding radioassay for 25-hydroxycholecaleiferol, J Clin Endoerinol Metab 33:992, 1971. 8. Hillman LS, and Haddad JG: Perinata[ vitamin D metabo[ism. II. Factors influencing late gestationa[ human serum 25-hydroxyvitaminD, Am J Obstet Gynecol 125:196, 1976. 9. Hillman LS, and Haddad JG: Human perinatal vitamin D metabolism. I. 25-Hydroxyvitamin D in maternal and cord blood, J PEDIATR84:742, 1974. 10. Weech AA: The influence of the administration of cod-liver oil to the mother on the development of rickets in the infant. Johns Hopkins Hosp Bull 40:244, 1927. 11. Rudolf M, Arulanantham K, and Greenstein RM: Unsuspected nutritional rickets, Pediatrics 66:72, 1980.
Maternal-infant vitamin D relationships during breast-feeding To evaluate the effect o f maternal dietary vitamin D intake on infant vitamin D status in a country with a temperate climate, but where the commercial milk supply is not vitamin D fortified, this randomized, double-blind study was conducted on term mother-infant pairs during the winter months. Well-nourished, white nursing mothers were given a placebo, 500 IU vitamin D/day or 1,000 IU vitamin D/day; their infants were not given supplemental vitamin D. After six weeks, mothers receiving supplemental vitamin D had higher levels o f 25-hydroxyvitamin D than had mothers receiving placebo. A direct relationship was observed between maternal and infant levels o f 25-hydroxyvitamin D at six weeks, implying that maternal vitamin D intake directly affects the vitamin D concentration in breast milk. A control group o f infants who had received 400 IU vitamin D/day had even higher concentrations o f 25-hydroxyvitamin D, suggesting that infant supplementation with vitamin D is more efficacious than maternal supplementation. Despite the favorable climate in South Africa, during winter breast-fed infants have low serum vitamin D values i f maternal dietary vitamin D intake is low.
Alan D. Rothberg, M.B., B.Ch., F.C.P.(SA),* John M. Pettifor, M.B., B.Ch., F.C.P.(SA), Ph.D.(Med), Desmond F. Cohen, M.B., B.Ch., D.C.H.(SA), Ernst W. W. Sonnendecker, M.B., B.Ch., M.Med.(O et G), F.R.C.O.G., and F. Patrick Ross, Ph.D.,
Johannesburg, South Africa
A QUESTION remains regarding the need for vitamin D supplementation in the term, exclusively breast-fed infant, with much of the controversy revolving around both the content and bioactivity of D in human milk. 1-6The results of two recent studies showed adequate 25-hydroxyvitamin D concentrations in nonsupplemented breast-fed infants during the first three to four months of life. l, 2 However, in both studies, all nursing mothers supplemented their daily
From the Departments o f Pediatrics and Obstetrics and Gynecology, Johannesburg Hospital, and the Metabolic and Nutrition Research Unit, University o f the Witwatersrand. Supported in part by grants from the South African Medical Research Council and the Atomic Energy Board. *Reprint address: Department of Pediatrics, Johannesburg Hospital, Private Bag X39, Johannesburg 2000, South Africa.
500
The Journal of PEDIATRICS Vol. 101, No. 4, pp. 500-503
diets with D, and most presumably drank D-fortified milk. The American Academy of Pediatrics Committee on Nutrition has recently commented on the role of maternal intake of D in fortified cow milk in preventing rickets in Abbreviation used D: vitamin D 25-OHD: 25-hydroxyvitamin D (calcidiol) Ca: calcium PO4: phosphate breast-fed infants2 In most countries of the world cow milk is not routinely supplemented with D. The present study was therefore undertaken to evaluate the need for supplemental D in term, breast-fed infants in a country with a temperate climate but where the milk supply is not D fortified. The absence of D fortification also enabled us to control maternal dietary intake and to examine the relationship between maternal D intake and infant D status.
0022-3476/82/100500+04500.40/0 @ 1982 The C. V. Mosby Co.
Volume 101 Number 4
Maternal-infant vitamin D relationships in breast-feeding
50 1
Table I. Maternal blood values (mean _+ S D) 6 wk Day 4: all groups (n = 40)
Maternal blood 25-OHD (ng/ml) Ca (mmol/L) PO~ (mmol/L) Alkaline phosphatase (1U/L)
11.9 2.1 1.4 111
I I I
• 6.0 _+ 0.1 • 0.4 • 24
Placebo (group P, n = 10)
500 U/day (group A, n = 9)
10.0 2.5 1.4 61
13.8 2.4 1.4 69
• 5.5 _+ 0.3 • 0.1 • 17
1,000 U/day (group B, n=9)
_+ 5.4 + 0.1 • 0.2 • 17
14.7 2.5 1.4 74
• • • •
4.0 0.3 0.1 24
Control (group C, n = 12) 11.0 2.3 1.3 75
_+ 3.5 -+ 0.1" _+ 0.2 • 16t
Conversion of SI to traditional units: serum Ca, I mmol/L ~ 4 rag/100 ml; PO4, 1 mmol/L .~ 3.1 mg/100 ml. *All six-weekvalues > baseline (P < 0.05). tAll six-weekvalues < baseline (P < 0.01).
Table II. Blood values in infants (mean + SD) 6 wk Infant blood 25-OHD(ng/ml) Ca (mmol/L) PO4 (mmol/L) Alkaline phosphatase (1U/L)
Day 4: all groups (n = 40}
(group P, n-- lO)
8.9 2.2 2.1 122
1.I 2.4 2.2 267
+ 7.1 • 0.3 __+0.3 • 30
-- 1.4 _+ 0.1" + 0.3 _+ 106
(group ,4. n=9) 10.2 2.4 2.3 325
-+ 5.5 _+ 0.2 + 0.2 • 75
(group B, n=9) 9.4 2.3 2.4 318
• 2.I + 0.1 -+ 0.2 • 98
(group C, n=12) t5.2 2.4 2.3 286
_ 3.7 _+ 0.1 _+ 0.2t _+ 71:~
*Six-week value > baseline for placebo group (P < 0.05) by paired t test. ~'Six-weekvalue > baseline for Control subjects (P < 0.05) by paired t test. ~:AI1six-weekvalues > baseline (P < 0.001) by paired t test.
MATERIALS
AND METHODS
This randomized, double-blind study was conducted on term mother-infant pairs at the Johannesburg Hospital, between the months of June (midwinter) through August, 1981. Seventy-seven well-nourished white nursing mothers and their infants were enrolled. Mothers were give n placebo (group P, n = 20), 500 I U D / d a y (group A, n = 20), or 1,000 I U D / d a y (group B, n = 20) for six weeks following delivery. These mothers received no additional D in other vitamin preparations or milk, and their infants did not receive supplemental D. A control group of mothers (group C, n = 17) received no D, but their infants were given 400 I U D / d a y . Mothers received only iron and folate supplements during pregnancy. Serum calcium, phosphorus, alkaline phosphatase, and 2 5 - O H D were measured in mothers and babies on the fourth postnatal day and six weeks later, A t the conclusion of the study all infants were given 400 I U supplemental D daily. After completion of the biochemical analyses, the code was broken, and the results were analyzed. Serum Ca was measured by atomic absorption speetrophotometry. Alkaline phosphatase and PO4 were measured
by standard techniques by means of a Technicon AutoAnalyzer II. Serum 2 5 - O H D was measured by the method of Haddad and Chyu. 7 The limit of sensitivity of this method was 2 ng/ml. In our laboratory the normal pediatric range for 2 5 - O H D is 8 to 40 n g / m l . The two-tailed paired and non-paired Student t tests were used in the analysis of data. Informed consent was obtained, and the study was approved by the hospital ethics committee.
RESULTS O f the 77 infants, 40 (57%) were still being completely breast-fed at six weeks and thus form the study group (group P, n = 10; group A, n = 9; group B, n = 9; group C, n -- 12). Maternal data (Table I). There were no intergroup differences for 25-OHD, Ca, PO4 or alkaline phosphatase on day 4. After six weeks, there were again no intergroup differences for Ca, PO4, or alkaline phosphatase; however, Ca concentrations had increased significantly from baseline values (P < 0.05), and alkaline phosphatase had decreased (P < 0.01). The mean 2 5 - O H D concentration for nonsupplemented mothers (groups P + C) at six weeks
502
Rothberg et al.
The Journal of Pediatrics October 1982
-E c
30
C)
c 25 E 0 20
9~
o
~---..o Mothers r =-.41 p = <0'05
3O
~
25
Infants
08 o
s
15
"-~-.*.,~oo
9
9
r =-.44
s
p =<005
.-~ 20 E
o
-- Day 4 (Placental transfer) r = 0.50 p= < 0 0 1
~ - - - o 6 Weeks (Milk transfer) r=0.58 p = < O.OI
i
?
Cg
E
A~.
5
o o
9
0 9
0
9 9
9
*
t-"
o
I
U')
0 0
l 10
June I , July, ! ' August 20 ~0 10 20 5 0 I0 20 Date of entry into study (4th postnatal day)
I
50
c~
5
_~
0
0
5
iO
t5
2o
25
50
Maternal 25-hydroxy vitamin D (ng/ml)
Fig. 1. Relationship between month ofdelivery and maternal and infant 25-OHD concentrations.
Fig. 2. Relationship between maternal and infant 25-OHD concentrations at four days and at six weeks.
Table IlL Meteorologic data
trations of 25-OHD than those infants in the maternally supplemented groups (P < 0.01). At six weeks, 25-OHD concentrations in group A, B, and~P infants were significantly related to maternal 25-OHD values (P < 0.01; Fig. 2).
Mean maximum daytime temperature 1981"
April May June July August
(~ 25.9 21.7 19.7 21.5 23.3
Mean hr sunlight per day
9.2 8.4 9.1 9.5 7.5
*Data from the South African Weather Bureau (Department of Transport).
was 10.6 _ 4.3 ng/ml, significantly lower than that of supplemented mothers, 14.0 _+ 4.5 ng/ml (groups A + B), P < 0.05. Mothers delivering in June had significantly higher 25-OHD values than those delivering in August (P < 0.05; Fig. 1). Infant data (Table II). There were no intergroup differences for 25-OHD, Ca, PO4, or alkaline phosphatase on day 4. The day 4 25-OHD concentrations in the neonate were directly related to the month of entry into the study (P < 0.05; Fig. 1) and to the maternal 25-OHD concentrations (P < 0.01; Fig. 2). After six weeks there were no intergroup differences for Ca, PO4, or alkaline phosphatase despite the finding of low 25-OHD concentrations in the placebo group. Group P infants had a mean 25-OHD concentration significantly below that of infants in the other groups ( P < 0.01), although no biochemical evidence of rickets was observed. Control infants (group C) had significantly higher concen-
DISCUSSION The mean day 4 maternal 25-OHD concentration in our study (11.9 +-6.0 ng/ml) was lower t h a n the thirdtrimester concentrations previously observed in white American mothers during the winter months (15.4 +_ 5.9 ng/ml), 8 possibly as a consequence of lower dietary D intake in our mothers. This finding~would be compatible with the conclusion of Hi!lman and Haddad g that, although ultraviolet exposure was the major determinant of maternal serum 25-OHD concentration, dietary D intake was important during periods of decreased exposure. Our mothers' diets were not supplemented with D during pregnancy, and they did not receive D-fortified milk. Despite the generally favorable climate in Johannesburg (Table III), when day 4 maternal 25-OHD concentrations were analyzed in terms of date of delivery, we observed a progressive decrease from June through August (Fig. 1). Maternal concentrations of 25-OHD correlated with concentrations in the infants at birth as reported previously.6,9 No infant had clinical or biochemical sequelae of low serum 25-OHD values during the immediate neonatal period. We have demonstrated a relationship between maternal D supplementation and maternal 25-OHD concentrations at six weeks. Furthermore, maternal 25-OHD concentrations correlate directly with infant 25-OHD values over the
Volume 101 Number 4
Maternal-infant vitamin D relationships in breast-feeding
same period, implying that maternal D intake directly affects the D concentration in human milk and thus the D status of the infant. This has been suggested previously~~ but could not be confirmed in a recent study,~ perhaps because of the small sample size. Maternal and infant 25-OHD concentrations were unaffected by doubling the maternaI D supplementation from 500 IU to 1,000 IU/day, whereas infants receiving 400 IU/day directly had significantly higher values at six weeks. Supplementation of the infant's diets is apparently a more efficacious method of maintaining adequate D status than supplementation of the maternal diet. These data show that in winter, in the nonsupplemented mother, milk concentrations of 25-OHD and transfer of D or one of its metabolites to the infant are low, even in a healthy, well-nourished, white population living in a favorable climate. Thus, D supplementation to mothers or infants would seem prudent, since under the study conditions, nonsupplemented infants had extremely low serum concentrations of D within six weeks. Recommendations based on the results of studies into the need for infant D supplementation in countries where milk is D fortified may not be relevant for populations not consuming Dfortified milk or milk products. 2'3 Certainly, a reductionin the incidence of rickets has been associated with milk fortification.I~ Recent American reports of nutritional rickets have included breast-fed infants of strict vegetarians and food faddists who have excluded milk from their diets] thus emphasizing the importance of D fortification in the prevention of rickets during breast-feeding. Similarly, the role of vitamin preparations taken by the mother during the postnatal period Should be considered when an attempt is made to answer the question of the need to supplement the diet of the breast-fed infant.2 Indeed, in two recent studies, nonsupplemented infants of mothers who received at least 400 IU D/day had serum 25-OHD concentrations in the range of 20 ng/ml at six weeks. I,~ Despite the absence of biochemical evidence of rickets, even in those infants with very low 25-OHD concentrations, the present study has demonstrated the potential for the development of rickets during Winter in nonsupple-
503
mented infants. However, it is possible that 25-OHD concentrations rise fairly rapidly in these infants after the first six weeks of life, as sun exposure of mothers and infants increases. Greater sun exposure of mothers and infants may explain the difference between our findings and those of Birkbeck and Scott, ~ who studied mothers and breast-fed infants in New Zealand. Pending further studies, we recommend routine D supplementation of breast-fed infants whose mothers have low intake of Vitamin D, irrespective of climate conditions. We thank Mr. Gopal Moodley and Miss M. Pentopoulosfor the biochemical evaluations. REFERENCES
1. Greer FR, SearcY JE, Levin RS, Steichen JJ, Asch PS, and Tsang RC: Bone mineral content and serum 25-hydrovitamin D concentration in breast-fed infants with and without supplemental vitamin D, J PEDIATR98:696, i981. 2. Roberts CC, Chan GM, Folland D, Rayburn C, and Jackson R: Adequate bone mineralization in breast-fed infants, J PEDIATR99:192, 1981. 3. Finberg L: Human milk feeding and vitamin D supplementation-1981, J PEDIATR99:228, 1981. 4. Greer FR, Ho M, Dodson D, and Tsang RC: Lack of 25-hydroxyvitaminD and 1,25-dihydroxyvitaminD in human milk, J PEDIATR99:233, 1981. 5. Nutrition and lactation--Committee on Nutrition, American Academy of Pediatrics, Pediatrics 68:435, 1981. 6. Birkbeck JA, and Scott HF: 25-hydroxycholecalciferolserum levels in breast-fed infants~ Arch Dis Child 55:691, 1980. 7. Haddad J, and Chyu K: Competitive protein-binding radioassay for 25-hydroxycholecaleiferol, J Clin Endoerinol Metab 33:992, 1971. 8. Hillman LS, and Haddad JG: Perinata[ vitamin D metabo[ism. II. Factors influencing late gestationa[ human serum 25-hydroxyvitaminD, Am J Obstet Gynecol 125:196, 1976. 9. Hillman LS, and Haddad JG: Human perinatal vitamin D metabolism. I. 25-Hydroxyvitamin D in maternal and cord blood, J PEDIATR84:742, 1974. 10. Weech AA: The influence of the administration of cod-liver oil to the mother on the development of rickets in the infant. Johns Hopkins Hosp Bull 40:244, 1927. 11. Rudolf M, Arulanantham K, and Greenstein RM: Unsuspected nutritional rickets, Pediatrics 66:72, 1980.