The Economist - UK (2022-05-07)

The Economist May 7th 2022 TechnologyQuarterlyThequantifiedself 7

Killerapps, saving lives

I

t is a trendlinethat gives you pause: a spikenearingthetopof

the  blue  band  that  marks  your  normal  blood­sugarrange,fol­

lowed by a sharp dip. The culprit is obvious:thesourdoughtoast

for  breakfast  half  an  hour  before.  A  generouspieceofbanana

bread the following morning leaves a completelydifferenttrace.

The blood­sugar line runs nicely unperturbedrightdownthemid­

dle of the normal zone all the way through tolunchtime.

These  are  the  sort  of  revelations  thatacontinuousglucose

monitor, a coin­sized device attached to theskin,beamstoyour

smartphone’s screen. The device lasts for twoweeksandhasa tiny

needle  that  gets  just  under  the  skin.  Everyfewminutesitmea­

sures  the  concentration  of  sugar  in  the  fluidbetweenthecells

there—a good proxy for what is going on inthebloodstream.

Almost as soon as the first such continuousglucosemonitors

started  replacing  finger­prick  blood  testsfordiabeticsin2014,

they began to show up on the arms of non­diabeticgeeksinSili­

con Valley. They were looking for ways to “hack”theirmetabolism

into  delivering,  for  instance,  more  energyorbrainclarity.Their

ad­hoc  experiments  were  soon  replicatedbyformalresearchers

doing broader studies of metabolism. Those,inturn,haveledto

the possibility of personalised nutrition. 

Such  metabolic  studies  have  changedthescientificthinking

on  what  a  healthy  diet  looks  like.  It  hasturnedoutthatmany

seemingly  healthy  people  often  have  largepost­mealspikesin

blood sugar, which have been linked to the developmentofpre­di­

abetes. Without any intervention pre­diabetesturnsintodiabetes

in 37­70% of cases within four years. The sugardipsthatoftenfol­

low the spikes were recently found to be problematic,too,because

they make people feel hungry. People whoare“bigdippers”con­

sume about 300 more calories a day than thosewhoarenot.

In 2015 researchers in Israel showed thatanai­basedalgorithm

they had developed could predict someone’sblood­sugarreaction

to  various  foods.  The  algorithm’s  inputsincludedbloodtests,

sleep, exercise, height and weight, which allaffectdailymetabolic

variations. They also included the compositionofthegutmicro­

biome, the trillions of bacteria residing in thegutwhosecollective

job is to process what we eat. Microbiome analysisisdonebyshot­

gun genomic sequencing of everything foundina stoolsample.

In the past five years startups in Ameri­

ca,  Europe  and  Asia  have  launched  ai­

based  personalised­nutrition  apps  that

build  on  these  discoveries.  One  of  them,

Zoe, sends customers a set of specially for­

mulated muffins. By knowing exactly what

is in the food being eaten, and measuring

the  changes  in  blood  sugar  and  fat  that

come about in response, the company can

create a predictive model of its customers’

metabolism. Its algorithm then whips up a

bespoke  catalogue  of  foods  and  meals,

with  predicted  blood­sugar  reactions  to

each.  Tushar  Vashisht,  co­founder  of

HealthifyMe,  an  Indian  startup  that  pro­

vides digital coaching for weight loss, says

the trove of data from customers who can

afford  various  connected  devices  and

blood  tests  as  inputs  for  their  bespoke

plansisusefulfarbeyondthosecustomers.Itcanhelptobuild

appsforpeoplewhodonothavesuchdevices,whichwouldrely

onai­derivedproxiesofweight,bloodsugarandsoon.

Forsuchsystemstokeepmetabolismsinbalance,theyhaveto

beadheredto.Knowingwhatishappeninginsideyourbodyisno

helpifyoudonothingto changethepatternofbehaviourin­

volved.Justbeingtoldthatitisinyourinterestsisnot,typically,

enough.Soappssoldaswaysofachievinghealthgainsontheba­

sisofmeasurementsmadebywearablestypicallyincorporatea

varietyofbehaviouralnudgestokeeptheuserfocused.

Theaiinnovationinsuchpersonaliseddietsmakesthemeasi­

ertomaintaininthelongtermbecauseit givespeopleoptionson

howtomakethefoodsthatthealgorithmsaysareparticularlybad

forthema littlelessbad.Thealgorithms

maysuggestsmalltweaks,suchassprin­

klingsomenutsonthaticecreamorgoing

fora longwalkaftereatingit.Januaryai,

another personalised­nutrition startup,

saysithasderivedthenutritioncontents

of16m grocery­storeitems, recipes and

menusoflocalrestaurants,whichmakesit

easierforuserstoplanandtrackmeals.

Itisstillearlydays,butresultsreported

byusersofsuchprecision­nutritionpro­

grammeslookencouraging.Userssaythey

arelosingweight,havehigherenergylev­

elsandaresleepingbetter.Somediabetic

usersnolongerneedmedication.Studies

ofseveralappsareunderwaytoconfirm

andquantifythesebenefits.

However,thoughtfulsuggestionsthat

makecomplianceeasieraretheexception.

Better measurement does little good unlesstheappsthat

interpret the data are well designed

Dealing with the data

App-trition

Health-care apps released by year*

Source:IQVIAInstitute *Available on Apple Store and Google Play

120

90

60

0

0

2010 2018161412

No longer available