April 8, 2025 • 27 mins
Let us know what you think about Health Affairs podcasts at communications@healthaffairs.org. If you have 30 minutes to spare, let us know and we'll set up a 30-minute chat for the first 20 listeners that reach out. Coffee will be on us.
Health Affairs' Senior Deputy Editor Rob Lott interviews Seth Berkowitz of the UNC School of Medicine to discuss his recent paper that explores a new approach to help guide research and policy at the intersection of income, food, nutrition, and health.
Order the April 2025 issue of Health Affairs.
Currently, more than 70 percent of our content is freely available - and we'd like to keep it that way. With your support, we can continue to keep our digital publication Forefront and podcast
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Rob Lott (00:31):
Hello, and welcome to A Health Podicy. I'm your host,
Rob Lott. We know that dietrelated disease is a big threat
to population health. Diabetes,cardiovascular disease, and many
cancers all can be attributed inpart to things like people's
(00:54):
diet quality and their nutritionsecurity. And yet statements
like that, while true, arewholly inadequate.
There are numerous otherinterrelated factors
contributing to those and otherdiseases, and there are also
numerous other interrelatedfactors upstream too. Factors as
varied as income and environmentand countless other drivers of
(01:18):
things like diet quality, whichas we've said, in turn drives
health. And so, yes, althoughwe've known about the food and
health connection forgenerations, less clear are how
those connections, and there areso many, relate to each other.
And also less clear is wheresomeone like a researcher or a
(01:39):
policymaker might start when itcomes to developing and
evaluating interventions aimedat improving diet and health.
That's the subject of today'sHealth Odyssey.
I'm here with Doctor. Seth A.Berkowitz, a general internist
and associate professor ofmedicine at the University of
(02:00):
North Carolina, Chapel Hill.Together with Drs. Hilary
Seligman and Darish Mazafarian,Doctor.
Berkowitz is the coauthor of anarticle titled, quote, A New
Approach to Guide Research andPolicy at the Intersection of
Income, food, nutrition, andhealth. It's one of the key
(02:21):
framing articles in the brandnew April 2025 issue of Health
Affairs, which is dedicatedentirely to the subject of food
and health. This article verymuch sets the stage so nicely
for many of the other empiricalstudies in the issue, now
available at HealthAffairs.org.Doctor. Seth Berkowitz, we're
(02:42):
lucky to have you here to talkabout the article.
Welcome to A Health Odyssey.
Seth Berkowitz (02:46):
Yeah. Thanks for having me on. I really
appreciate the opportunity totalk about it.
Rob Lott (02:49):
Great. Well, let's dive right in. Your article
offers a a quote, new conceptualmodel for understanding the
interrelationships among income,food security, nutrition
security, diet quality, andhealth. Before we talk about the
new model, would you paint us apicture of sort of what came
before? What was the traditionalinterpretation of these elements
(03:14):
and where perhaps does it fallshort that you felt the need to
develop a new model?
Seth Berkowitz (03:21):
Yeah. I think that's a great question. I think
in in some cases, there are, youknow, sort of older models that
people have and a new one istrying to supplant it. In this
case, I think it's more that,there was, like, a a general
sort of nonspecific sense thatthese things were kind of
related, but there wasn't a lotof specificity to it. And so the
(03:41):
newness of this model is istrying to add some specificity
to these interrelationships.
We know, overall, I think that,you know, the income you have
might affect the, you know, thefood you can buy, and that
affects the overall quality ofyour diet, and that can affect
health. But as we've, you know,sort of moved more into this
research field, we've realizedthat, you know, it's actually
worth making distinctionsbetween some of these constructs
(04:04):
that might, tend to blendtogether, you know, food
insecurity as being separatethan nutrition insecurity, both
of those being separate fromdiet quality, income, you know,
influencing those, incomeinfluencing health through other
ways that don't go through that.And so our our goal in in
putting the model togetherwasn't necessarily to say, oh,
the old way we thought aboutthese things is wrong. Here's a
new way to think about it, butrather to say, you know, people
(04:26):
are paying attention to theseconcepts now. They're
interrelated.
We all know that. But but,specifically, how does that
actually work and try to givesort of a scaffolding or or, you
know, framework in that sensefor people to to use as they're
thinking about these differentinterrelated ideas.
Rob Lott (04:41):
Got it. Okay, well before we sort of dig into the
model, one other backgroundquestion. I'm wondering if you
can say a little bit about whythis kind of research is so
hard. The sort of obviousexample I think about often is,
one day you'll read a headlinethat says, drinking red wine is
good for your health and thenthe next day it's bad, same with
(05:02):
chocolate. And I think the samealso applies to sort of the more
complicated dynamics at playhere.
Is that just the nature ofsocial science and empirical
research, or is there somethingparticular about this field that
(05:23):
is just really hard?
Seth Berkowitz (05:24):
Yeah. No. I do I do think there are some some
elements that are particularlychallenging when it comes to
nutrition and things like that.I will take a step back and say
that I I think big picture whenthought about in, you know you
know, what I think at least isthe right way or sort of, what
what might be a helpful way, Ido think we have pretty sound,
nutrition science when when wetalk about overall dietary
(05:46):
patterns. Meaning that I thinkwe know overall, what are more
versus less health ful, ways toeat.
I think we know that, you know,having, you know, a higher
proportion of your foods comefrom plants, a higher proportion
of, you know, the fats that youconsume be unsaturated, you
know, kind of more fat, plantderived fats rather than, animal
(06:06):
fats. I think how consumingcalories in moderation. I think
not consuming too much sugar. Ithink not consuming too much
alcohol. Those kind of bigpicture things, I think, are
relatively well studied.
And we know that there are acouple different dietary
patterns that all kind of dothat but are helpful. You know,
a Mediterranean diet pattern ishelpful. A DASH diet pattern is
helpful. A lot of traditional,patterns from different
(06:26):
countries around the the worldhave been studied and found to
be, helpful. It gets morecomplicated, as you say, when
you start getting into, like,specific foods or maybe even,
like, specific nutrients.
Like, is this one specific food,you know, the best one for you,
or or are foods with thisparticular phytochemical, good
for you and things like that.That's that's a little bit
(06:48):
trickier and harder to say, asyou say. You know, it's
definitely the case that in thepopular press because I think we
love you know, everyone likesloves to eat, and and so we love
articles about eating and thingslike that. And so it's it's
always catchy to say, oh, thisthis, you know, do this eat this
one food and you'll live to be ahundred or, like, drop this one
food and you'll live to be ahundred or whatever else. So
that that does happen.
I think the other thing that'sjust tricky about, like,
(07:10):
nutrition and diet in general isyou're you're generally talking
about exposures that happen overlong time frames for their, you
know, for their effects toreally manifest, you know,
years, if not decades, if not,you know, you know, double
digit, you know, you know,twenty, thirty, forty years,
things like that. And that'sjust scientifically hard to
study. Right? It's much easierto study something, you know, an
antibiotic treats your pneumoniain a week or it doesn't, and,
(07:32):
like, you know that right awayversus, you know, a dietary
pattern that you have to adhereto for twenty years or something
like that. That's just a harderthing to study.
I think also just the sheervolume and variety of foods
around the world. I mean, youknow, we I think of blood
pressure medicines as a as aclass of drugs where we have a
relatively high number oftherapeutic agents compared to
many things, but that's probablyreally only, like, 10 or maybe
(07:54):
15 in widespread use. Right?There are thousands, tens of
thousands of foods that eachhave their own, you know,
molecular composition, all thatstuff. So there's just a lot of
variety, and people eat a lot.
You know? We're doing itmultiple times a day and
everything. And so measuringprecisely what people are
consuming is just sort of adifficult scientific task as
well. Multiply that over longperiods of time. So it is a I
think it is a more difficultarea to study in some ways,
(08:16):
which which adds to some of thecomplexity around these issues
as well.
Rob Lott (08:21):
Yeah. I get that. I I keep trying to convince my
colleagues at health affairs topublish an article about the 10
superfoods that will turn yourhair blue or something like
that.
Seth Berkowitz (08:32):
Yeah, get those clicks All
Rob Lott (08:35):
right, well against that backdrop, tell me a little
bit about the model in yourpaper. Can you walk us through
it briefly?
Seth Berkowitz (08:42):
Sure. Yeah. So I think I think the underlying so
a couple of sort of underlyingpoints, that that we make in
trying to relate and, again, Iguess, it's kind of a mouthful
to say all these, but just tosay, it's a model that relates
income, food security, nutritionsecurity, diet quality, and
health. So those are sort of thefive key constructs that we're
that we're dealing with. One ofthe key underlying points is
(09:05):
that health itself is a verymultidimensional construct.
So there's physical health.There's mental health. Even
within physical health, there's,you know, diabetes and
cardiovascular disease, there's,you know, kidney disease,
there's malignancy, and allthose kind of things. So so it's
a very multidimensionalcontract. And one of the key
things that we wanted to, getacross is that how these
(09:25):
different other constructs, thethe income, the food insecurity,
interest insecurity, dietquality, interact with health
has to do a lot with thespecific measure or
manifestation of health you'retalking about.
And so there might be, you know,even though both physical health
and mental health are two keyparts of health, diet might have
a very different interactionwith, you know, some physical
health outcomes than otherphysical health outcomes or
(09:47):
physical health versus mentalhealth and these kinds of
things. And so we wanna relatethat in that sense. The other,
kind of key thing that we wantedto to get across is that, you
know, income is sort of thisenabling factor that, you know,
it it provides purchasing power,so it lets people buy various
things that they might need tobe healthy and can affect their
(10:08):
health in different ways. Andsome of those are food related,
and so that's how some of thoseother things get into. But
others have nothing to do withfood.
Right? It lets you buy higherquality housing so you're not
exposed to, you know,respiratory, pathogens like mold
or, air pollution. It lets youbuy, your medicine so you can
take those or access higherquality health care. Lets you
buy transportation so you canget to where you need to be and
all those kind of things. Sothere there are plenty of
(10:29):
pathways that income affectshealth that don't go through any
of the, any of the kind ofnutrition related ideas.
And then finally, there thereare these kind of three,
distinct but interrelated,ideas, food insecurity,
nutrition insecurity, and dietquality. And food insecurity,
(10:49):
you know, people usually thinkabout as, access to the food
needed for an active healthactive healthy life, That's sort
of a USDA definition of it. It'stightly tied to financial,
conditions, and so we expectthat there's gonna be a strong
relationship between income andfood insecurity. If you have too
little income, you're you'remore likely to be food insecure.
We also know that foodinsecurity is related to diet
(11:12):
quality.
If you look on average, peoplewith food insecurity have lower
diet quality than, than peoplewho are food secure, probably
because healthier foods are moreexpensive. However, the
magnitude of the difference is,you know, is relatively modest.
It's real and measurable andpersistent over time, but it's
not huge. And in the overall UScontext, we know that there are
(11:33):
a lot of things that affect dietquality that aren't related to
food insecurity. The overalllevel of diet quality in The US
across all income groups and andacross both, people who
experience food insecurity andpeople who are food secure is
relatively low.
And and so, you know, basically,in The US could there's probably
some room for for improvement orat least every income group and
(11:53):
and the different food securitystatuses, there's room room for
improvement there. So foodinsecurity isn't the only thing
that determines diet, quality.And then the and then, you know,
diet quality, you know, refersto sort of the overall
healthfulness of the diets thatpeople are eating and, again,
has many influences. So, youknow, income and food insecurity
(12:14):
are are one of those, but thereare others, and there are even
some paradoxical findings where,you know, for example, in
epidemiologic studies, we knowthat there's greater junk food
consumption amongst, high incomechildren than low income
children, which is sort of anopposite pattern of what people
might expect. And so we knowthat there are plenty of other
determinants of diet qualitythat don't go through sort of
(12:35):
income and food insecurity andhave to do with culture and
preferences and the food that'savailable to us and food
marketing and, you know, allthose different things there.
And then finally, kind of thenew kid on the block is,
nutrition security. And so sothat's a construct that I think
is still kinda taking shape, andwhat the, you know, final form,
it takes and how we think aboutit is is a little bit up in the
(12:55):
air right now. There's a lot ofinterest in it. On one hand, we
sort of know that that some ofour traditional measures of food
insecurity, don't really get atthe healthfulness of the foods
that people have access to. Andso, so there's some
justification for wanting to,you know, measure not just, do
people have access to foodoverall, but what is the
(13:17):
healthfulness of the food thatpeople have access to?
And so that's sort of onejustification for this this idea
of nutrition security. Anotherone is, you know, the diet
quality kind of talks about, theoverall patterns that of food
that people are eating, but itdoesn't necessarily get at what
someone's nutritional status isin any given point of time. Do
(13:37):
you actually consume enough ofthe macronutrients and
micronutrients that you need tostay healthy? And so another way
to think about nutritionsecurity is, you know, as a as
an indicator of the nutritionalstatus that that people have.
And there are kinda differentcamps within the field who some
prefer this sort of access tohealthy food interpretation of
nutrition security, some preferthis, nutritional status and and
(13:59):
overall healthfulness of dietsinterpretation of it.
And so we kind of go into alittle bit of the these,
different ideas within thisemerging construct of nutrition
security as well.
Rob Lott (14:08):
So it sounds like a little bit of what you're
getting at is that there arethese ideas or factors, but then
there's also some complexityaround how different people
interpret those factors, or Iguess the term you use is
construction. And so how did youattempt to navigate that tension
(14:28):
between sort of the facts of thereality, as well as sort of
perhaps the nature of the debatearound those facts. It seems
like you're sort of trying to doboth, and I and it's probably
the only way to do it, but I'mcurious how you thought about
that.
Seth Berkowitz (14:47):
Yeah. No. Totally. And, I mean, I think
there's think there's a tendencyin academia overall, and this is
probably not a good tendency to,you know, hyperfocus on
terminology, to makedistinctions even if those
distinctions maybe don't have adifference and things like that.
And and, you know, I'm certainlynot gonna say that we're immune
to that or or those kind ofthings, but really the guiding
impulse of the this kind of workhere was for interventional
(15:10):
reasons.
And in particular saying, youknow, because these are
different but interrelatedconstructs, interventions that
affect, you know, either onlyone and not the other of these
constructs or or affect theconstructs to different extents,
you know, maybe more of one thananother, are likely to have
different impacts on health. Andso so the way we really wanted
(15:33):
to sort of organize thisapproach overall is, you know,
in in constructing thisframework, how might it be used
to, you know, one, synthesizesort of existing evidence, make
sense of what we know from priorstudies, but also provide a way
for people to think about,alright. If we wanna change
health, we wanna make people'shealth better in some ways.
Which of these differentconstructs would be the levers
(15:55):
that we that we wanna use? Andthat that might be an
intervention that focuses onlyon one.
It might be one that that triesto synergistically work on
multiple ones at the same time.But having a sense of how
they're related, we think willlet let us do better at, at
addressing each of these issuesor at some of these issues to to
improve health. And so that'sreally sort of the underlying
name of the game, is how thesemight be applied to designing
(16:19):
interventions or understanding,you know, why interventions that
have been done had the resultsthey did have.
Rob Lott (16:25):
Well, that's a great segue going on to interventions.
In the paper, you basically sortof catalog a number of different
interventions in this space, aswell as the evidence about their
effectiveness, things as variedas cash assistance on one hand,
(16:46):
steps to improve diet quality onthe other. I'm curious as you
were looking at those interinterventions, did you encounter
any surprises that maybe, caughtyou off guard?
Seth Berkowitz (16:58):
I don't know if there were it was so much,
surprises, but, I mean, I thinkjust the variety of different
health outcomes that, thatdifferent interventions, you
know, saw effects on was, reallyspoke to the, sort of the
complexity of the situation, butalso the potential of using
interventions that focus ondifferent, you know, constructs
(17:20):
within here to achievedifferent, different outcomes.
Just as you know, kind of anexample, you know, for in for
interventions that focusparticularly on providing income
but in a relatively unstructuredway, so not, you know, saying,
oh, you have to use, you know,this income in this way or
whatever, but sort of broadincome support. You know, we
found impacts on mental health.We found acts impacts on health
(17:43):
care utilization, like reducedemergency department visits and
things like that. But we didn'tfind as much on, you know, some
of the, like, biomarkers, likehemoglobin a one c, for people
with diabetes or things likethat.
On the other hand, you know,some things that were, you know,
maybe a little closer on the,you know, on what you might
imagine is sort of the causalmap, between these things, so
(18:05):
interventions that are trying toaffect diet quality in
particular did that, but butthen they were the ones that
that, impacted some of thesebiomarkers a little bit more. So
we had a a, you know, food asmedicine study, that made health
food more affordable for forpeople with diabetes and that
found a a reduction in inhemoglobin a one c for for
people with diabetes. And so itjust sort of spoke to the
(18:27):
different, you know, aspects ofhealth that you may be able to
affect by sort of mixing andmatching, so to speak, between
which of these differentconstructs your intervention is
targeted towards.
Rob Lott (18:38):
Great. Well, in a moment, I wanna ask you a little
bit about putting all that intopractice. What does it look like
to mix and match, if you will?But first, let's take a quick
break. And we're back.
(19:07):
I'm talking with Doctor. SethBerkowitz about his paper from
the April 2025 issue of HealthAffairs offering a new model at
the intersection of income,food, nutrition, and health.
Okay, let's put this intopractice. I'll admit when I hear
the term conceptual model, itfeels very academic and
(19:28):
theoretical to me. I know thatthat's intentional.
That's the work we do here. ButI'm wondering if you can
convince me why something likethis might be useful in the real
world of policymaking. Let's sayyou're in the halls of Congress,
your congressman is walking downthe hall, you're able to grab
(19:52):
him for sixty seconds before heducks into the cloakroom, and
and you can put this paper inhis hands and tell him why you
think it's important for him toknow about it. Where do you
begin?
Seth Berkowitz (20:05):
So, I mean, I think the real reason that a
that a framework like this isimportant is because, you know,
it it helps you, home in on thethe interventional approach
that's most likely to achievethe outcome that you want. So,
again, you know, health is avery multidimensional construct.
We can say, yeah, we wannaimprove health. But your
immediate next question afterthat is like, well, what aspect
(20:26):
of health, is it? And so, fromthere, that you know, once
you've identified that, I wannaimprove hemoglobin a one c for
people with diabetes, or Iwanna, reduce emergency
department visits, or I wannaimprove mental health.
Once you've honed in on that,that's when, you know, sort of
the planning starts. You say,okay. That's the goal. How do we
do that? And I think a modellike this is useful because, you
(20:48):
know, there are all thesepossible intervention targets.
You know, you could try toprovide income support. You
could try to reduce foodinsecurity. You could try to,
provide dietary education andcounseling and those kind of
things. And you wanna have asense of which one is likely or
or what combination of those arelikely to achieve the outcome
you want. And so that that'sreally where I think the utility
is.
You have a problem, you wannahave, you know, some rational
(21:10):
way to, come up with anintervention that's likely to
improve it, and to do that, youknow, we've synthesized a lot of
information together that canprovide some guidance along the
way.
Rob Lott (21:20):
Okay. Certainly for different populations, there are
disparities and factors like youmentioned, income distribution,
access to services and supports,even health outcomes. But so
give us a picture of what itlooks like when you, see these
factors through an equity lens,and how did you include that in
(21:42):
your development of this model?
Seth Berkowitz (21:45):
So, I mean, health equity is really, an
important topic to to all of us,to Hillary and Dari, as well.
For for me, you know, I had abook on health equity come out,
last year. So, I mean, it'ssomething that I spend a lot of
time, thinking thinking aboutand really sort of informs, what
we're doing here. You know, forme, the way I think about health
(22:06):
health inequity as injusticethat harms health. And what what
to me that means is when we havesocial structures and
institutions that sort of aren'tproviding people with with what
they need and what they're dueas, you know, sort of an an
equal member of society.
And we see that play out in alot of things here, the way that
(22:27):
income is distributed in The US,the the prevalence of food in
insecurity and the distributionof it, who experiences food
insecurity, who has low dietquality, all that is really
patterned by social structures,institutions, practices that
distribute, resources, in TheUS. So, you know, you think of
the the educational system andand whether discrimination and
(22:48):
opportunity within that isequally distributed. It's not.
And so that that leads to, youknow, different labor market
outcomes with further aspects ofdiscrimination. And all of that
might mean that, you know, basedon, you know, sort of factors
that should not, play into this,you know, you see lower income
distribution, for certain, youknow, people in The US.
(23:10):
You see higher risk of foodinsecurity for for people in The
US. And so thinking about thesethese structures that sort of
put people in, these positions,I think, is really important.
And that's why we really, the inthe paper, emphasized, you know,
not only having specificintervention to sort of mitigate
the consequences of this, I. E.Something that, you know, takes
(23:31):
someone who, say, isexperiencing, you know, diabetes
and, you know, low access tohealthy foods and provide, say,
a food as medicine interventionto to address that, but also at
the same time sort of thinkingabout the the structures and and
sort of background socialpolicies that put people in that
situation in the first place.
And so that's why you might useincome support interventions, to
(23:51):
provide a baseline of,purchasing power for everyone,
but then on top of that, add a afood as medicine intervention,
specifically for people who are,you know, in worse circumstances
at the present moment. And so wewanted people to think both
structurally, in a big picturesense, but also be able to think
clinically about, you know, saythe patient in front of you if
you're a physician or, you know,the client in front of you if
(24:13):
you're a human serviceorganization or something like
that, and provide specificinterventions that way as well.
Rob Lott (24:19):
I wanted to ask about another big factor that we've
seen on our radar a fair bit athealth affairs, relatively new
when it comes to food andhealth, and that's the rapid
uptake and embrace of GLP-1s.And let's say that that
continues to expand and grow.Have you thought about how that
(24:44):
fits sort of into this model orwhat someone looking at this
model might apply to the therise of GLP ones in in health
care.
Seth Berkowitz (24:57):
Yeah. No. I think GLP ones are really a very
complementary intervention to alot of the stuff that we're
talking about here. So, thereare a lot of health benefits to
GLP ones, you know, althoughthere there's ever more research
into the mechanism of action ofGLP ones. You know, a key way
that they work is they they tendto reduce appetite, and so
people tend to lose, a goodamount of weight, while on them.
(25:19):
But that weight loss is, youknow, kind of a blunt instrument
in some way. So so some of theweight that people lose is
adipose tissue, which is, youknow, endocrinologically active
and causes some of the harms ofexcess adiposity. But some of
the tissue that people loo orsome of the weight people lose
is muscle tissue, and so thatcan cause something called
sarcopenia, you know, too lowmuscle mass, and that can be
(25:40):
associated with frailty andother problems, as well. And so
having tools that can reallycause weight loss like GLP ones
is great, but that doesn'tobviate the need for people to
consume healthful diets, to tryto optimize where that weight
loss comes from, I e, can it bemore adipose tissue and less
muscle tissue when when you'relosing weight? We also know that
(26:01):
that at least right now, becauseof the current, patent regimes
and everything, GLP ones arevery expensive.
And we know that GLP ones have alot of side effects, and so
there there are a number ofpeople who just can't really
tolerate, being on them. And soI think there are plenty of ways
to use nutrition interventions,to to sort of complement some of
that therapy. You know, my mycoauthor, Dari Mozaffarian, had
(26:22):
a great viewpoint, you know, Ithink at last year or maybe the
year before about using food asmedicine interventions initially
while people are on GLP ones,and then and then further using
it to help people transition offGLP ones. That time of weight
loss might be a time to toestablish some healthy patterns
in terms of how people areeating and things like that. And
then, ultimately, you know, oncea a good amount of weight loss
(26:42):
has been achieved and everythingelse, maybe people can stop and
maintain a more healthier,lifestyle and get sort of the
benefits of that treatment, butbe assisted by, say, a food as
medicine or other kind ofnutrition intervention, to, to
maintain that weight loss, havehealthy diet patterns, and those
kind of things.
So I think there are a lot ofways that that, you GLP ones
and, you know, food andnutrition interventions can be,
(27:04):
complementary, but but they'realso doing different things, and
I wouldn't wanna have only oneor only the other.
Rob Lott (27:10):
Great. Well, that might be a good place to stop.
Doctor. Seth Berkowitz, thankyou so much for taking the time
to chat with us today.
Seth Berkowitz (27:18):
Yeah, thanks so much for having me on.
Rob Lott (27:20):
To our listeners and our readers, check out Doctor.
Berkowitz's article in the April2025 issue of Health Affairs.
And if you enjoyed this podcast,please recommend it to a friend,
smash that subscribe button, andtune in next week. Take care
all.
Seth Berkowitz (27:41):
Thanks for listening. If you enjoyed
today's episode, I hope you'lltell a friend about a health
policy.
Hello, and welcome to A Health Podicy. I'm your host,
Rob Lott. We know that dietrelated disease is a big threat
to population health. Diabetes,cardiovascular disease, and many
cancers all can be attributed inpart to things like people's
(00:54):
diet quality and their nutritionsecurity. And yet statements
like that, while true, arewholly inadequate.
There are numerous otherinterrelated factors
contributing to those and otherdiseases, and there are also
numerous other interrelatedfactors upstream too. Factors as
varied as income and environmentand countless other drivers of
(01:18):
things like diet quality, whichas we've said, in turn drives
health. And so, yes, althoughwe've known about the food and
health connection forgenerations, less clear are how
those connections, and there areso many, relate to each other.
And also less clear is wheresomeone like a researcher or a
(01:39):
policymaker might start when itcomes to developing and
evaluating interventions aimedat improving diet and health.
That's the subject of today'sHealth Odyssey.
I'm here with Doctor. Seth A.Berkowitz, a general internist
and associate professor ofmedicine at the University of
(02:00):
North Carolina, Chapel Hill.Together with Drs. Hilary
Seligman and Darish Mazafarian,Doctor.
Berkowitz is the coauthor of anarticle titled, quote, A New
Approach to Guide Research andPolicy at the Intersection of
Income, food, nutrition, andhealth. It's one of the key
(02:21):
framing articles in the brandnew April 2025 issue of Health
Affairs, which is dedicatedentirely to the subject of food
and health. This article verymuch sets the stage so nicely
for many of the other empiricalstudies in the issue, now
available at HealthAffairs.org.Doctor. Seth Berkowitz, we're
(02:42):
lucky to have you here to talkabout the article.
Welcome to A Health Odyssey.
Seth Berkowitz (02:46):
Yeah. Thanks for having me on. I really
appreciate the opportunity totalk about it.
Rob Lott (02:49):
Great. Well, let's dive right in. Your article
offers a a quote, new conceptualmodel for understanding the
interrelationships among income,food security, nutrition
security, diet quality, andhealth. Before we talk about the
new model, would you paint us apicture of sort of what came
before? What was the traditionalinterpretation of these elements
(03:14):
and where perhaps does it fallshort that you felt the need to
develop a new model?
Seth Berkowitz (03:21):
Yeah. I think that's a great question. I think
in in some cases, there are, youknow, sort of older models that
people have and a new one istrying to supplant it. In this
case, I think it's more that,there was, like, a a general
sort of nonspecific sense thatthese things were kind of
related, but there wasn't a lotof specificity to it. And so the
(03:41):
newness of this model is istrying to add some specificity
to these interrelationships.
We know, overall, I think that,you know, the income you have
might affect the, you know, thefood you can buy, and that
affects the overall quality ofyour diet, and that can affect
health. But as we've, you know,sort of moved more into this
research field, we've realizedthat, you know, it's actually
worth making distinctionsbetween some of these constructs
(04:04):
that might, tend to blendtogether, you know, food
insecurity as being separatethan nutrition insecurity, both
of those being separate fromdiet quality, income, you know,
influencing those, incomeinfluencing health through other
ways that don't go through that.And so our our goal in in
putting the model togetherwasn't necessarily to say, oh,
the old way we thought aboutthese things is wrong. Here's a
new way to think about it, butrather to say, you know, people
(04:26):
are paying attention to theseconcepts now. They're
interrelated.
We all know that. But but,specifically, how does that
actually work and try to givesort of a scaffolding or or, you
know, framework in that sensefor people to to use as they're
thinking about these differentinterrelated ideas.
Rob Lott (04:41):
Got it. Okay, well before we sort of dig into the
model, one other backgroundquestion. I'm wondering if you
can say a little bit about whythis kind of research is so
hard. The sort of obviousexample I think about often is,
one day you'll read a headlinethat says, drinking red wine is
good for your health and thenthe next day it's bad, same with
(05:02):
chocolate. And I think the samealso applies to sort of the more
complicated dynamics at playhere.
Is that just the nature ofsocial science and empirical
research, or is there somethingparticular about this field that
(05:23):
is just really hard?
Seth Berkowitz (05:24):
Yeah. No. I do I do think there are some some
elements that are particularlychallenging when it comes to
nutrition and things like that.I will take a step back and say
that I I think big picture whenthought about in, you know you
know, what I think at least isthe right way or sort of, what
what might be a helpful way, Ido think we have pretty sound,
nutrition science when when wetalk about overall dietary
(05:46):
patterns. Meaning that I thinkwe know overall, what are more
versus less health ful, ways toeat.
I think we know that, you know,having, you know, a higher
proportion of your foods comefrom plants, a higher proportion
of, you know, the fats that youconsume be unsaturated, you
know, kind of more fat, plantderived fats rather than, animal
(06:06):
fats. I think how consumingcalories in moderation. I think
not consuming too much sugar. Ithink not consuming too much
alcohol. Those kind of bigpicture things, I think, are
relatively well studied.
And we know that there are acouple different dietary
patterns that all kind of dothat but are helpful. You know,
a Mediterranean diet pattern ishelpful. A DASH diet pattern is
helpful. A lot of traditional,patterns from different
(06:26):
countries around the the worldhave been studied and found to
be, helpful. It gets morecomplicated, as you say, when
you start getting into, like,specific foods or maybe even,
like, specific nutrients.
Like, is this one specific food,you know, the best one for you,
or or are foods with thisparticular phytochemical, good
for you and things like that.That's that's a little bit
(06:48):
trickier and harder to say, asyou say. You know, it's
definitely the case that in thepopular press because I think we
love you know, everyone likesloves to eat, and and so we love
articles about eating and thingslike that. And so it's it's
always catchy to say, oh, thisthis, you know, do this eat this
one food and you'll live to be ahundred or, like, drop this one
food and you'll live to be ahundred or whatever else. So
that that does happen.
I think the other thing that'sjust tricky about, like,
(07:10):
nutrition and diet in general isyou're you're generally talking
about exposures that happen overlong time frames for their, you
know, for their effects toreally manifest, you know,
years, if not decades, if not,you know, you know, double
digit, you know, you know,twenty, thirty, forty years,
things like that. And that'sjust scientifically hard to
study. Right? It's much easierto study something, you know, an
antibiotic treats your pneumoniain a week or it doesn't, and,
(07:32):
like, you know that right awayversus, you know, a dietary
pattern that you have to adhereto for twenty years or something
like that. That's just a harderthing to study.
I think also just the sheervolume and variety of foods
around the world. I mean, youknow, we I think of blood
pressure medicines as a as aclass of drugs where we have a
relatively high number oftherapeutic agents compared to
many things, but that's probablyreally only, like, 10 or maybe
(07:54):
15 in widespread use. Right?There are thousands, tens of
thousands of foods that eachhave their own, you know,
molecular composition, all thatstuff. So there's just a lot of
variety, and people eat a lot.
You know? We're doing itmultiple times a day and
everything. And so measuringprecisely what people are
consuming is just sort of adifficult scientific task as
well. Multiply that over longperiods of time. So it is a I
think it is a more difficultarea to study in some ways,
(08:16):
which which adds to some of thecomplexity around these issues
as well.
Rob Lott (08:21):
Yeah. I get that. I I keep trying to convince my
colleagues at health affairs topublish an article about the 10
superfoods that will turn yourhair blue or something like
that.
Seth Berkowitz (08:32):
Yeah, get those clicks All
Rob Lott (08:35):
right, well against that backdrop, tell me a little
bit about the model in yourpaper. Can you walk us through
it briefly?
Seth Berkowitz (08:42):
Sure. Yeah. So I think I think the underlying so
a couple of sort of underlyingpoints, that that we make in
trying to relate and, again, Iguess, it's kind of a mouthful
to say all these, but just tosay, it's a model that relates
income, food security, nutritionsecurity, diet quality, and
health. So those are sort of thefive key constructs that we're
that we're dealing with. One ofthe key underlying points is
(09:05):
that health itself is a verymultidimensional construct.
So there's physical health.There's mental health. Even
within physical health, there's,you know, diabetes and
cardiovascular disease, there's,you know, kidney disease,
there's malignancy, and allthose kind of things. So so it's
a very multidimensionalcontract. And one of the key
things that we wanted to, getacross is that how these
(09:25):
different other constructs, thethe income, the food insecurity,
interest insecurity, dietquality, interact with health
has to do a lot with thespecific measure or
manifestation of health you'retalking about.
And so there might be, you know,even though both physical health
and mental health are two keyparts of health, diet might have
a very different interactionwith, you know, some physical
health outcomes than otherphysical health outcomes or
(09:47):
physical health versus mentalhealth and these kinds of
things. And so we wanna relatethat in that sense. The other,
kind of key thing that we wantedto to get across is that, you
know, income is sort of thisenabling factor that, you know,
it it provides purchasing power,so it lets people buy various
things that they might need tobe healthy and can affect their
(10:08):
health in different ways. Andsome of those are food related,
and so that's how some of thoseother things get into. But
others have nothing to do withfood.
Right? It lets you buy higherquality housing so you're not
exposed to, you know,respiratory, pathogens like mold
or, air pollution. It lets youbuy, your medicine so you can
take those or access higherquality health care. Lets you
buy transportation so you canget to where you need to be and
all those kind of things. Sothere there are plenty of
(10:29):
pathways that income affectshealth that don't go through any
of the, any of the kind ofnutrition related ideas.
And then finally, there thereare these kind of three,
distinct but interrelated,ideas, food insecurity,
nutrition insecurity, and dietquality. And food insecurity,
(10:49):
you know, people usually thinkabout as, access to the food
needed for an active healthactive healthy life, That's sort
of a USDA definition of it. It'stightly tied to financial,
conditions, and so we expectthat there's gonna be a strong
relationship between income andfood insecurity. If you have too
little income, you're you'remore likely to be food insecure.
We also know that foodinsecurity is related to diet
(11:12):
quality.
If you look on average, peoplewith food insecurity have lower
diet quality than, than peoplewho are food secure, probably
because healthier foods are moreexpensive. However, the
magnitude of the difference is,you know, is relatively modest.
It's real and measurable andpersistent over time, but it's
not huge. And in the overall UScontext, we know that there are
(11:33):
a lot of things that affect dietquality that aren't related to
food insecurity. The overalllevel of diet quality in The US
across all income groups and andacross both, people who
experience food insecurity andpeople who are food secure is
relatively low.
And and so, you know, basically,in The US could there's probably
some room for for improvement orat least every income group and
(11:53):
and the different food securitystatuses, there's room room for
improvement there. So foodinsecurity isn't the only thing
that determines diet, quality.And then the and then, you know,
diet quality, you know, refersto sort of the overall
healthfulness of the diets thatpeople are eating and, again,
has many influences. So, youknow, income and food insecurity
(12:14):
are are one of those, but thereare others, and there are even
some paradoxical findings where,you know, for example, in
epidemiologic studies, we knowthat there's greater junk food
consumption amongst, high incomechildren than low income
children, which is sort of anopposite pattern of what people
might expect. And so we knowthat there are plenty of other
determinants of diet qualitythat don't go through sort of
(12:35):
income and food insecurity andhave to do with culture and
preferences and the food that'savailable to us and food
marketing and, you know, allthose different things there.
And then finally, kind of thenew kid on the block is,
nutrition security. And so sothat's a construct that I think
is still kinda taking shape, andwhat the, you know, final form,
it takes and how we think aboutit is is a little bit up in the
(12:55):
air right now. There's a lot ofinterest in it. On one hand, we
sort of know that that some ofour traditional measures of food
insecurity, don't really get atthe healthfulness of the foods
that people have access to. Andso, so there's some
justification for wanting to,you know, measure not just, do
people have access to foodoverall, but what is the
(13:17):
healthfulness of the food thatpeople have access to?
And so that's sort of onejustification for this this idea
of nutrition security. Anotherone is, you know, the diet
quality kind of talks about, theoverall patterns that of food
that people are eating, but itdoesn't necessarily get at what
someone's nutritional status isin any given point of time. Do
(13:37):
you actually consume enough ofthe macronutrients and
micronutrients that you need tostay healthy? And so another way
to think about nutritionsecurity is, you know, as a as
an indicator of the nutritionalstatus that that people have.
And there are kinda differentcamps within the field who some
prefer this sort of access tohealthy food interpretation of
nutrition security, some preferthis, nutritional status and and
(13:59):
overall healthfulness of dietsinterpretation of it.
And so we kind of go into alittle bit of the these,
different ideas within thisemerging construct of nutrition
security as well.
Rob Lott (14:08):
So it sounds like a little bit of what you're
getting at is that there arethese ideas or factors, but then
there's also some complexityaround how different people
interpret those factors, or Iguess the term you use is
construction. And so how did youattempt to navigate that tension
(14:28):
between sort of the facts of thereality, as well as sort of
perhaps the nature of the debatearound those facts. It seems
like you're sort of trying to doboth, and I and it's probably
the only way to do it, but I'mcurious how you thought about
that.
Seth Berkowitz (14:47):
Yeah. No. Totally. And, I mean, I think
there's think there's a tendencyin academia overall, and this is
probably not a good tendency to,you know, hyperfocus on
terminology, to makedistinctions even if those
distinctions maybe don't have adifference and things like that.
And and, you know, I'm certainlynot gonna say that we're immune
to that or or those kind ofthings, but really the guiding
impulse of the this kind of workhere was for interventional
(15:10):
reasons.
And in particular saying, youknow, because these are
different but interrelatedconstructs, interventions that
affect, you know, either onlyone and not the other of these
constructs or or affect theconstructs to different extents,
you know, maybe more of one thananother, are likely to have
different impacts on health. Andso so the way we really wanted
(15:33):
to sort of organize thisapproach overall is, you know,
in in constructing thisframework, how might it be used
to, you know, one, synthesizesort of existing evidence, make
sense of what we know from priorstudies, but also provide a way
for people to think about,alright. If we wanna change
health, we wanna make people'shealth better in some ways.
Which of these differentconstructs would be the levers
(15:55):
that we that we wanna use? Andthat that might be an
intervention that focuses onlyon one.
It might be one that that triesto synergistically work on
multiple ones at the same time.But having a sense of how
they're related, we think willlet let us do better at, at
addressing each of these issuesor at some of these issues to to
improve health. And so that'sreally sort of the underlying
name of the game, is how thesemight be applied to designing
(16:19):
interventions or understanding,you know, why interventions that
have been done had the resultsthey did have.
Rob Lott (16:25):
Well, that's a great segue going on to interventions.
In the paper, you basically sortof catalog a number of different
interventions in this space, aswell as the evidence about their
effectiveness, things as variedas cash assistance on one hand,
(16:46):
steps to improve diet quality onthe other. I'm curious as you
were looking at those interinterventions, did you encounter
any surprises that maybe, caughtyou off guard?
Seth Berkowitz (16:58):
I don't know if there were it was so much,
surprises, but, I mean, I thinkjust the variety of different
health outcomes that, thatdifferent interventions, you
know, saw effects on was, reallyspoke to the, sort of the
complexity of the situation, butalso the potential of using
interventions that focus ondifferent, you know, constructs
(17:20):
within here to achievedifferent, different outcomes.
Just as you know, kind of anexample, you know, for in for
interventions that focusparticularly on providing income
but in a relatively unstructuredway, so not, you know, saying,
oh, you have to use, you know,this income in this way or
whatever, but sort of broadincome support. You know, we
found impacts on mental health.We found acts impacts on health
(17:43):
care utilization, like reducedemergency department visits and
things like that. But we didn'tfind as much on, you know, some
of the, like, biomarkers, likehemoglobin a one c, for people
with diabetes or things likethat.
On the other hand, you know,some things that were, you know,
maybe a little closer on the,you know, on what you might
imagine is sort of the causalmap, between these things, so
(18:05):
interventions that are trying toaffect diet quality in
particular did that, but butthen they were the ones that
that, impacted some of thesebiomarkers a little bit more. So
we had a a, you know, food asmedicine study, that made health
food more affordable for forpeople with diabetes and that
found a a reduction in inhemoglobin a one c for for
people with diabetes. And so itjust sort of spoke to the
(18:27):
different, you know, aspects ofhealth that you may be able to
affect by sort of mixing andmatching, so to speak, between
which of these differentconstructs your intervention is
targeted towards.
Rob Lott (18:38):
Great. Well, in a moment, I wanna ask you a little
bit about putting all that intopractice. What does it look like
to mix and match, if you will?But first, let's take a quick
break. And we're back.
(19:07):
I'm talking with Doctor. SethBerkowitz about his paper from
the April 2025 issue of HealthAffairs offering a new model at
the intersection of income,food, nutrition, and health.
Okay, let's put this intopractice. I'll admit when I hear
the term conceptual model, itfeels very academic and
(19:28):
theoretical to me. I know thatthat's intentional.
That's the work we do here. ButI'm wondering if you can
convince me why something likethis might be useful in the real
world of policymaking. Let's sayyou're in the halls of Congress,
your congressman is walking downthe hall, you're able to grab
(19:52):
him for sixty seconds before heducks into the cloakroom, and
and you can put this paper inhis hands and tell him why you
think it's important for him toknow about it. Where do you
begin?
Seth Berkowitz (20:05):
So, I mean, I think the real reason that a
that a framework like this isimportant is because, you know,
it it helps you, home in on thethe interventional approach
that's most likely to achievethe outcome that you want. So,
again, you know, health is avery multidimensional construct.
We can say, yeah, we wannaimprove health. But your
immediate next question afterthat is like, well, what aspect
(20:26):
of health, is it? And so, fromthere, that you know, once
you've identified that, I wannaimprove hemoglobin a one c for
people with diabetes, or Iwanna, reduce emergency
department visits, or I wannaimprove mental health.
Once you've honed in on that,that's when, you know, sort of
the planning starts. You say,okay. That's the goal. How do we
do that? And I think a modellike this is useful because, you
(20:48):
know, there are all thesepossible intervention targets.
You know, you could try toprovide income support. You
could try to reduce foodinsecurity. You could try to,
provide dietary education andcounseling and those kind of
things. And you wanna have asense of which one is likely or
or what combination of those arelikely to achieve the outcome
you want. And so that that'sreally where I think the utility
is.
You have a problem, you wannahave, you know, some rational
(21:10):
way to, come up with anintervention that's likely to
improve it, and to do that, youknow, we've synthesized a lot of
information together that canprovide some guidance along the
way.
Rob Lott (21:20):
Okay. Certainly for different populations, there are
disparities and factors like youmentioned, income distribution,
access to services and supports,even health outcomes. But so
give us a picture of what itlooks like when you, see these
factors through an equity lens,and how did you include that in
(21:42):
your development of this model?
Seth Berkowitz (21:45):
So, I mean, health equity is really, an
important topic to to all of us,to Hillary and Dari, as well.
For for me, you know, I had abook on health equity come out,
last year. So, I mean, it'ssomething that I spend a lot of
time, thinking thinking aboutand really sort of informs, what
we're doing here. You know, forme, the way I think about health
(22:06):
health inequity as injusticethat harms health. And what what
to me that means is when we havesocial structures and
institutions that sort of aren'tproviding people with with what
they need and what they're dueas, you know, sort of an an
equal member of society.
And we see that play out in alot of things here, the way that
(22:27):
income is distributed in The US,the the prevalence of food in
insecurity and the distributionof it, who experiences food
insecurity, who has low dietquality, all that is really
patterned by social structures,institutions, practices that
distribute, resources, in TheUS. So, you know, you think of
the the educational system andand whether discrimination and
(22:48):
opportunity within that isequally distributed. It's not.
And so that that leads to, youknow, different labor market
outcomes with further aspects ofdiscrimination. And all of that
might mean that, you know, basedon, you know, sort of factors
that should not, play into this,you know, you see lower income
distribution, for certain, youknow, people in The US.
(23:10):
You see higher risk of foodinsecurity for for people in The
US. And so thinking about thesethese structures that sort of
put people in, these positions,I think, is really important.
And that's why we really, the inthe paper, emphasized, you know,
not only having specificintervention to sort of mitigate
the consequences of this, I. E.Something that, you know, takes
(23:31):
someone who, say, isexperiencing, you know, diabetes
and, you know, low access tohealthy foods and provide, say,
a food as medicine interventionto to address that, but also at
the same time sort of thinkingabout the the structures and and
sort of background socialpolicies that put people in that
situation in the first place.
And so that's why you might useincome support interventions, to
(23:51):
provide a baseline of,purchasing power for everyone,
but then on top of that, add a afood as medicine intervention,
specifically for people who are,you know, in worse circumstances
at the present moment. And so wewanted people to think both
structurally, in a big picturesense, but also be able to think
clinically about, you know, saythe patient in front of you if
you're a physician or, you know,the client in front of you if
(24:13):
you're a human serviceorganization or something like
that, and provide specificinterventions that way as well.
Rob Lott (24:19):
I wanted to ask about another big factor that we've
seen on our radar a fair bit athealth affairs, relatively new
when it comes to food andhealth, and that's the rapid
uptake and embrace of GLP-1s.And let's say that that
continues to expand and grow.Have you thought about how that
(24:44):
fits sort of into this model orwhat someone looking at this
model might apply to the therise of GLP ones in in health
care.
Seth Berkowitz (24:57):
Yeah. No. I think GLP ones are really a very
complementary intervention to alot of the stuff that we're
talking about here. So, thereare a lot of health benefits to
GLP ones, you know, althoughthere there's ever more research
into the mechanism of action ofGLP ones. You know, a key way
that they work is they they tendto reduce appetite, and so
people tend to lose, a goodamount of weight, while on them.
(25:19):
But that weight loss is, youknow, kind of a blunt instrument
in some way. So so some of theweight that people lose is
adipose tissue, which is, youknow, endocrinologically active
and causes some of the harms ofexcess adiposity. But some of
the tissue that people loo orsome of the weight people lose
is muscle tissue, and so thatcan cause something called
sarcopenia, you know, too lowmuscle mass, and that can be
(25:40):
associated with frailty andother problems, as well. And so
having tools that can reallycause weight loss like GLP ones
is great, but that doesn'tobviate the need for people to
consume healthful diets, to tryto optimize where that weight
loss comes from, I e, can it bemore adipose tissue and less
muscle tissue when when you'relosing weight? We also know that
(26:01):
that at least right now, becauseof the current, patent regimes
and everything, GLP ones arevery expensive.
And we know that GLP ones have alot of side effects, and so
there there are a number ofpeople who just can't really
tolerate, being on them. And soI think there are plenty of ways
to use nutrition interventions,to to sort of complement some of
that therapy. You know, my mycoauthor, Dari Mozaffarian, had
(26:22):
a great viewpoint, you know, Ithink at last year or maybe the
year before about using food asmedicine interventions initially
while people are on GLP ones,and then and then further using
it to help people transition offGLP ones. That time of weight
loss might be a time to toestablish some healthy patterns
in terms of how people areeating and things like that. And
then, ultimately, you know, oncea a good amount of weight loss
(26:42):
has been achieved and everythingelse, maybe people can stop and
maintain a more healthier,lifestyle and get sort of the
benefits of that treatment, butbe assisted by, say, a food as
medicine or other kind ofnutrition intervention, to, to
maintain that weight loss, havehealthy diet patterns, and those
kind of things.
So I think there are a lot ofways that that, you GLP ones
and, you know, food andnutrition interventions can be,
(27:04):
complementary, but but they'realso doing different things, and
I wouldn't wanna have only oneor only the other.
Rob Lott (27:10):
Great. Well, that might be a good place to stop.
Doctor. Seth Berkowitz, thankyou so much for taking the time
to chat with us today.
Seth Berkowitz (27:18):
Yeah, thanks so much for having me on.
Rob Lott (27:20):
To our listeners and our readers, check out Doctor.
Berkowitz's article in the April2025 issue of Health Affairs.
And if you enjoyed this podcast,please recommend it to a friend,
smash that subscribe button, andtune in next week. Take care
all.
Seth Berkowitz (27:41):
Thanks for listening. If you enjoyed
today's episode, I hope you'lltell a friend about a health
policy.
Popular Podcasts
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Therapy Gecko
An unlicensed lizard psychologist travels the universe talking to strangers about absolutely nothing. TO CALL THE GECKO: follow me on https://www.twitch.tv/lyleforever to get a notification for when I am taking calls. I am usually live Mondays, Wednesdays, and Fridays but lately a lot of other times too. I am a gecko.