Breastfeeding and Intelligence article has methodologic flaws
The article by Der, Batty and Deary has flawed methodology, making
any conclusion about the effect on breastfeeding and intelligence
premature.
In this study, breastfeeding is not defined. It can mean anything
from one breastfeeding a day, to exclusive breastfeeding (8-12
breastfeedings per day). The authors used duration as a proxy for dose,
which cannot be done. We should treat breastfeeding as we treat all other
drugs or treatments; we would never publish a study on a drug where the
dose was not specified. Duration is not an appropriate proxy for dose; nor
would it be considered so for any other intervention studied.
Furthermore, the median duration of breastfeeding was only 3 months
in this study, when the recommended duration of breastfeeding is 1-2
years. Very few children in this study even reached recommended levels--
the 95th percentile here was only 14 months.
Finally, the study excluded the low-birthweight babies, most of whom
are premature. This is a group known to be affected by breastfeeding, as
breastmilk is especially important for neurologic and eye development in
this group.
It is important that breastfeeding research and its peer review
process be rigorous. It is clear that this study was not reviewed by
people with expertise in breastfeeding research-- such reviewers would
never recommend publishing a study with these serious methodologic flaws.
All we can conclude from this study is that breastfeeding for a very
short duration, with unspecified doses of breastmilk, in non-low
birthweight children, had no effect on intelligence. To really determine
the effect of breastfeeding on intelligence, one must carefully define
breastfeeding, include a population of exclusively breastfed infants, and
study infants whose breastfeeding duration approached recommended levels.
Low birthweight infants should be studied as a separate group.
Rapid Response:
Breastfeeding and Intelligence article has methodologic flaws
The article by Der, Batty and Deary has flawed methodology, making
any conclusion about the effect on breastfeeding and intelligence
premature.
In this study, breastfeeding is not defined. It can mean anything
from one breastfeeding a day, to exclusive breastfeeding (8-12
breastfeedings per day). The authors used duration as a proxy for dose,
which cannot be done. We should treat breastfeeding as we treat all other
drugs or treatments; we would never publish a study on a drug where the
dose was not specified. Duration is not an appropriate proxy for dose; nor
would it be considered so for any other intervention studied.
Furthermore, the median duration of breastfeeding was only 3 months
in this study, when the recommended duration of breastfeeding is 1-2
years. Very few children in this study even reached recommended levels--
the 95th percentile here was only 14 months.
Finally, the study excluded the low-birthweight babies, most of whom
are premature. This is a group known to be affected by breastfeeding, as
breastmilk is especially important for neurologic and eye development in
this group.
It is important that breastfeeding research and its peer review
process be rigorous. It is clear that this study was not reviewed by
people with expertise in breastfeeding research-- such reviewers would
never recommend publishing a study with these serious methodologic flaws.
All we can conclude from this study is that breastfeeding for a very
short duration, with unspecified doses of breastmilk, in non-low
birthweight children, had no effect on intelligence. To really determine
the effect of breastfeeding on intelligence, one must carefully define
breastfeeding, include a population of exclusively breastfed infants, and
study infants whose breastfeeding duration approached recommended levels.
Low birthweight infants should be studied as a separate group.
Competing interests:
None declared
Competing interests: No competing interests