Laboratory analyses

Breastfeeding Help And Baby Care For New Parents

The Secret to Pain Free Breastfeeding

Get Instant Access

Milk sodium and potassium were measured by flame photometry as described previously.10 Milk IL8 was measured using a commercial ELISA

kit (CLB, Eurogenetics, Middlesex, UK). Lactulose and mannitol in urine were measured enzymatically using an autoanalyser.22 Milk viral load was measured in the aqueous cell free fraction by RNA PCR (Amplicor, Roche Diagnostics). The method had a detection limit of200 copies/ml. Infant HIV infection was determined by positive blood antibody tests at 9 months or greater and/or by detection of RNA in samples taken at earlier time points according to aprotocol describedpreviously.20

2.2.2 Statistical analyses

Biochemical data were log-transformed to normalise distributions. Na/K ratios <= 0.6 were considered normal, 0.6 to 1.0 were considered moderately raised, and > 1.0 greatly raised. These cutoffs were based partly on histograms ofthe distribution ofNa/Kratios, which were remarkably similar to those in our other studies,10,11 and partly on published values of milk electrolytes. Na/K ratios in milk ofhealthy women at one month postpartum generally averaged 0.6 or less 23 which is about twice the value expected if milk electrolyte levels reflect intracellular levels.12Na/K >1.0 is approximately equivalent to 18 mmol/L sodium which, after the first few days postpartum and in the absence ofweaning, is indicative ofmastitis.15

2.3 Results and Discussion

The trial is ongoing so results presented are based on preliminary cross-sectional analysis of an incomplete set. The subsample of women from the main study who were examined for subclinical mastitis did not differ in general from those in the whole study except that women who did not breastfeed at all were excluded. Ofthe 108 infants (only 104 mothers since there were 4 pairs of twins) 26 infants were HIV-infected by 3 months of age.

The prevalence at each time point of raised milk Na/K ratio in one or both breasts, for women from whom paired samples were available, is shown in Table 1. Results are similar to those of the cross-sectional survey. Unilateral raised milk sodium was more common than bilateral, except at 1 week where it is likely that in some women lactation was not fully established and the mammary epithelium tight junctions had not yet closed completely. Na/Kratio and milk viral load were not significantly affected by infant birth weight or gestational age or mode of delivery.

Table 2. Subclinical mastitis among South African women with HIV infection

Na/K level in both breasts*

1 week(n = 94)

6 weeks (n = 83)

3 months (n

bilateral low

35 (37%)

44 (53%)

33 (54%)

unilateral raised


17 (18%)

19 (23%)

10 (16%)


11 (12%)

9 (11%)

12 (20%)

bilateral raised

both moderate

11 (12%)

5 (6%)

3 (5%)

one moderate, one severe


4 (5%)

3 (5%)

both severe

8 (8%)

2 (2%)

0 (0%)

*low: Na/K <= 0.6; moderate: Na/K > 0.6 and <=

1 .0; severe: Na/K> 1.0

Since left and right breasts seemed to behave independently, for some analyses we used breast, rather than woman, as the sampling unit. Milk IL8 was always highly correlated with Na/K ratio as we have shown in other populations previously (Table 2). Milk IL8 was also correlated with viral load at all times but viral load was significantly correlated with Na/K ratio only at 1 and 14 weeks. A similar relation between milk sodium and viral load has been found in HIV-infected Malawian women.24 IL8 acts to recruit leukocytes and we actually measured cell-free virus so we do not believe the relationship between IL8 and viral load is directly causal but that both are related to the changes that occur during subclinical mastitis. It is slightly surprising that Na/K ratio, which indicates non-specific permeability, is not more closely related to cell-free virus than is IL8 concentration.

Table 3. Correlations between logarithmic values of milk Na/K ratios, HIV viral loads and

IL8 concentrations in samples from both breasts

Table 3. Correlations between logarithmic values of milk Na/K ratios, HIV viral loads and

IL8 concentrations in samples from both breasts

Correlations (r, n, P)

1 week

6 weeks

14 weeks

Na/K with IL8 Na/K with viral load IL8 with viral load

0.56, 189, <0.001 0.33, 128, <0.001 0.28, 127,0.001

0.50, 166, <0.001 0.13, 137, 0.12 0.24, 138,0.005

0.56, 120, <0.001 0.32, 110, 0.001 0.45, 116, <0.001

At 1 and 14 weeks viral load in milk samples with Na/K<=0.6 was significantly lower than in samples with Na/K>l (Table 3). Interestingly, the proportion of samples with undetectable virus decreased with increasing Na/K ratio category at 6 and 14 weeks but not at 1 week. It appears that the cause of subclinical mastitis and its relation to inflammation and viral load differs at 1 week from later times.

In the overall trial, women who exclusively breastfed were less likely to transmit HIV to their infants than women who mixed breast milk and other foods. Therefore we analysed milk viral loads (Table 4) according to feeding practice.

Table 4. Subclinical mastitis and breast milk HIVviral load in SouthAfrican women1

Na/K ratio

] week

6 weeks

14 weeks

<= 0.6

Sil (570- 1155) 18/66 (27%)

1018(707- 1465)

35/98 (36%)

829(553 - 1245) 31/77 (40%)

0.6 - 1.0

1372 (809 - 2329) 10/39 (26%)

1550 (576-4172) 8/26 (31%)

1772 (639-4910) 5/18 (28%)

> 1.0

2751 (867- 8729)* 9/25 (36%)

1565 (699 -3504) 1/14 (7%)

1/15 (7%)

1 Geometric mean (95% CI); proportion (%) undetectable, i.e. < 200 copies/ml.

* Different from group with Na/K <=0.6

1 Geometric mean (95% CI); proportion (%) undetectable, i.e. < 200 copies/ml.

* Different from group with Na/K <=0.6

Mixed feeding was associated with increased viral load at 1 week only. The proportions ofbreasts with undetectable viral load were significantly different across groups at 6 and 14 weeks but this was mainly because ofthe high proportion among the small numbers in the breast milk plus water group. IL8 showed the same pattern as viral load andNa/Kratio was similar but not statistically significantly different even at 1 week, unlike in our previous cross-sectional study. However, when analysed as proportions in the different Na/K categories, there were significantly fewer samples in the moderately and greatly raised Na/K groups in the exclusive breast feeding group compared to the other groups. Feeding mode likely affects Na/K, IL8 and viral load, rather than vice versa, since there was no association ofinitial feeding choice with measures of maternal health (eg blood CD4 count or hemoglobin during pregnancy) or infant health (birth weight, gestational age) other than the perhaps spurious associations that women who gave breast milk plus water were slightly older and had slightly heavier babies than women in other groups.

Table 5. Breast milk HIV viral load in both breasts according to infantfeeding practice1

1 week

6 weeks

14 weeks

Exclusive breast feeding

936(692 - 1265)" 31/100 (31%)

1267 (893 - 1799)a 34/115 (30%)

1288(886 - 1875) 30/98 (31%)

Breast milk + water

1707 (706 -4128)^ 2/18(11%)

334 (173 - 646)b 8/12 (67%)

200 (200 - 200) 4/4 (100%)

Mixed feeding

6557 (618 - 69626)b 4/12(33%)

1420 (356-5667)111 3/13 (23%)

765 (295 - 1984) 7/16(44%)

1 Geometric mean (95% CI); proportion (%) undetectable, i.e. <200 copies/ml. Means in a column not followed by the same superscript are significantly different by Duncan's multiple range test, P<0.05.

1 Geometric mean (95% CI); proportion (%) undetectable, i.e. <200 copies/ml. Means in a column not followed by the same superscript are significantly different by Duncan's multiple range test, P<0.05.

Finally we investigated whether subclinical mastitis was associated with increased infant gut permeability. There was no significant association between gut permeability and either milk Na/K or feeding mode. The latter observation is not surprising since, in a larger cohort from the same trial, we found significantly increased intestinal permeability in infants given formula only but not in those who were mixed fed (Rollins, Filteau et al, unpublished). It appears that the amount ofmilk consumed by the mixed fed group was sufficient to promote gut integrity and that complementary foods in this middle income (by African standards) peri-urban community with piped water supply were not grossly contaminated.

2.4 Conclusions

Mother-to-infant transmission occurs in utero, during delivery and postpartum through breast feeding. Recent progress in reducing transmission during delivery25'26 means that even greater attention needs to be paid to breast feeding transmission since more women are likely to consent to testing and more of their infants will be born uninfected. Thus the results from the main study from which the present results were taken are especially exciting showing decreased transmission up to 3 months of age in infants who were exclusively breastfed, compared to mixed fed, until that age. The present substudyprovides some evidence ofmechanisms for this association.

Subclinical mastitis was more prevalent among those who mixed fed, rather than exclusively breastfed, their infants. Subclinical mastitis was associated with increased milk viral load at 1 and 14 weeks postpartum but, interestingly, increased viral load was associated with mixed feeding only at 1 week. The mechanisms may differ at 1 week from those at later time points. Subclinical mastitis had no effect on infant gut permeability. This may be because high milk concentrations ofinflammatory cytokines such as IL8 are balanced by high concentrations ofanti-inflammatory cytokines such as transforming growth factor-13Therefore, although results are preliminary, it appears that subclinical mastitis may increase risk of HIV transmission during early lactation and this risk is mediated through increased milk viral load. This supports the epidemiological data suggesting that early lactation is the major period ofpostnatal transmission.27

Extrapolating from causes of clinical mastitis,28 subclinical mastitis is likely to result from systemic infection, breast infection, mixed feeding such that milk production decreases below about 400 ml/day15 or poor lactation practices which lead to milk stasis and mammary gland involution. Infections, usually from skin commensals such as Staphylococcus aureus, are unlikely to take hold without other factors inducing milk stasis.29 Lactation counselling to optimise breast feeding may reduce the incidence of milk stasis and raised milk sodium concentration.13

Policies for optimal infant feeding by African women with HIV infection are continuallybeingrevised. The present study provides mechanistic data to support the observation ofa protective effect ofexclusive breast feeding in a non-randomised cohort. Further studies should be conducted to determine whether promotion of exclusive breast feeding, which is cheap and safe even for women and infants unaffected by HIV, can decrease HIV transmission among other populations.

Was this article helpful?

0 0
New Mothers Guide to Breast Feeding

New Mothers Guide to Breast Feeding

For many years, scientists have been playing out the ingredients that make breast milk the perfect food for babies. They've discovered to day over 200 close compounds to fight infection, help the immune system mature, aid in digestion, and support brain growth - nature made properties that science simply cannot copy. The important long term benefits of breast feeding include reduced risk of asthma, allergies, obesity, and some forms of childhood cancer. The more that scientists continue to learn, the better breast milk looks.

Get My Free Ebook

Post a comment