How Early Should Obesity Prevention Start?

Matthew W. Gill­man, M.D., and David S. Lud­wig, M.D., Ph.D., N Engl J Med 2013; 369:2173–2175 Decem­ber 5, 2013

Obe­si­ty has per­vad­ed the Unit­ed States and is spread­ing through­out the world. Fol­low­ing in its wake is type 2 dia­betes, which will affect at least half a bil­lion peo­ple world­wide by 2030. A major­i­ty of U.S. women of child­bear­ing age are over­weight or obese (as defined by a body-mass index [BMI, the weight in kilo­grams divid­ed by the square of the height in meters] >25). These women are like­ly to gain exces­sive weight when they’re preg­nant, mak­ing it hard­er for them to return to their prepreg­nan­cy weight after deliv­ery. Post­par­tum weight reten­tion not only por­tends increased life­long risks for obe­si­ty-relat­ed com­pli­ca­tions but also an increased BMI at the incep­tion of future preg­nan­cies. Dur­ing preg­nan­cy, exces­sive weight gain, along with oth­er risk fac­tors such as ges­ta­tion­al dia­betes, can alter fetal growth and metab­o­lism, lead­ing to high­er adi­pos­i­ty in the off­spring. If the child is female, grows up obese, and becomes preg­nant, the cycle begins again. It is time to inter­rupt this vicious cycle to pre­vent obe­si­ty and chron­ic dis­eases in moth­ers and chil­dren.

Once obe­si­ty is present, it is chal­leng­ing to treat because of mul­ti­ple phys­i­o­log­i­cal, behav­ioral, and cul­tur­al feed­back loops. The good news is that the pre­na­tal peri­od and the first post­na­tal year hold crit­i­cal clues that may lead to inter­ven­tions to reduce obe­si­ty in women and pre­vent it in chil­dren. In a range of ani­mal mod­els (from rodents to non­hu­man pri­mates), dietary, hor­mon­al, mechan­i­cal, and oth­er per­tur­ba­tions that occur pre­na­tal­ly and dur­ing infan­cy induce life­long, often irre­versible derange­ments in the offspring’s adi­pos­i­ty and metab­o­lism. These changes involve the envi­ron­men­tal alter­ation of genet­ic expres­sion, in part through epi­ge­net­ic mech­a­nisms, rather than changes in the genome itself. Thus, time­ly inter­ven­tion dur­ing the ear­ly, plas­tic phas­es of devel­op­ment — unlike cor­rec­tive efforts made lat­er in life — may lead to improved life­long health tra­jec­to­ries.

Because of chal­lenges in mea­sur­ing fetal expo­sures and the long laten­cy between ini­tial deter­mi­nants and salient health out­comes, how­ev­er, it is dif­fi­cult to trans­late such proofs of prin­ci­ple in ani­mals to human pop­u­la­tions. The first gen­er­a­tion of devel­op­men­tal-ori­gins stud­ies in humans linked birth weight to adult obe­si­ty-relat­ed mor­bid­i­ty and mor­tal­i­ty. We now rec­og­nize that birth weight and each of its com­po­nents, ges­ta­tion­al dura­tion and fetal growth, are low-res­o­lu­tion, momen­tary mark­ers for myr­i­ad pre­na­tal and peri­na­tal influ­ences. In the past decade, many such influ­ences have been iden­ti­fied and quan­ti­fied in epi­demi­o­log­ic stud­ies that have involved the peri­od before birth, used mod­ern meth­ods to mit­i­gate con­found­ing, and incor­po­rat­ed bio­mark­ers. These stud­ies have iden­ti­fied pre­na­tal risk fac­tors for obe­si­ty rang­ing from lifestyle fac­tors such as the mother’s smok­ing sta­tus to psy­choso­cial fac­tors includ­ing antepar­tum depres­sion, med­ical con­di­tions such as ges­ta­tion­al dia­betes, phys­i­o­log­i­cal stress as reflect­ed by fetal expo­sure to glu­co­cor­ti­coids, and epi­ge­net­ic mark­ers such as gene-spe­cif­ic DNA methy­la­tion lev­els in umbil­i­cal-cord tis­sue.

After birth, rapid weight gain in the first 3 to 6 months of life is a potent pre­dic­tor of lat­er obe­si­ty and car­diometa­bol­ic risk. Lac­ta­tion can­not be the entire expla­na­tion, because breast-fed babies tend to gain more weight than for­mu­la-fed babies in the first few months of life. The peri­na­tal hor­mon­al milieu may very well be a con­tribut­ing fac­tor. In one study, high­er lep­tin lev­els in umbil­i­cal-cord blood, chiefly reflect­ing pla­cen­tal pro­duc­tion, were asso­ci­at­ed with slow­er gain in infant weight-for-length and low­er adi­pos­i­ty at the ages of 3 years and 7 years. In con­trast, high­er lep­tin lev­els at 3 years of age were asso­ci­at­ed with faster gains in BMI from 3 to 7 years, sug­gest­ing that lep­tin resis­tance devel­ops between birth and 3 years of age. These find­ings are con­sis­tent with stud­ies in ani­mals show­ing a crit­i­cal peri­od of peri­na­tal lep­tin expo­sure that allows nor­mal mat­u­ra­tion of appetite-reg­u­lat­ing neu­rons in the hypo­thal­a­mus. Fea­tures of infant feed­ing oth­er than breast ver­sus bot­tle may also play a role. Among for­mu­la-fed infants, the intro­duc­tion of solids before 4 months was asso­ci­at­ed with a six­fold increase in the odds of obe­si­ty 3 years lat­er.

Emerg­ing risk fac­tors for obe­si­ty include expo­sure to endocrine dis­rup­tors, which appear to do the most dam­age dur­ing times of max­i­mum devel­op­men­tal plas­tic­i­ty, and the gut micro­bio­ta. Our bod­ies con­tain about 1013 cells but as many as 1014microor­gan­isms. Cer­tain mod­i­fi­ca­tions in the num­ber and type of microor­gan­isms dur­ing infan­cy are asso­ci­at­ed with excess weight gain, at least in rodents. The infant gut is nor­mal­ly col­o­nized dur­ing tran­sit through the birth canal, which could be one rea­son why chil­dren deliv­ered by cesare­an sec­tion appear to be at ele­vat­ed risk for obe­si­ty.

Giv­en obesity’s numer­ous devel­op­men­tal deter­mi­nants, it is log­i­cal that effec­tive pre­ven­tion would tar­get mul­ti­ple mod­i­fi­able fac­tors. In com­bi­na­tion, two well-stud­ied pre­na­tal risk fac­tors, exces­sive ges­ta­tion­al weight gain and mater­nal smok­ing dur­ing preg­nan­cy, and two post­na­tal fac­tors, few­er months of breast-feed­ing and a short­er dura­tion of dai­ly sleep dur­ing infan­cy, are asso­ci­at­ed with wide vari­a­tion in child­hood obe­si­ty. In one study, preschool-age chil­dren whose moth­ers did not smoke or gain exces­sive weight dur­ing preg­nan­cy and who were breast-fed for at least 12 months and slept for at least 12 hours per day dur­ing infan­cy had a pre­dict­ed obe­si­ty preva­lence of 6%, as com­pared with 29% among chil­dren for whom the oppo­site was true for all four risk fac­tors; the rates were sim­i­lar (4% and 28%, respec­tive­ly) when the chil­dren reached 7 to 10 years of age Prob­a­bil­i­ty of Obe­si­ty at 7 to 10 Years of Age for 16 Com­bi­na­tions of Four Mod­i­fi­able Pre­na­tal and Post­na­tal Risk Fac­tors.). These obser­va­tion­al data raise the pos­si­bil­i­ty that avoid­ing some or all of these risk fac­tors could sub­stan­tial­ly reduce the pro­por­tion of child­hood obe­si­ty.

Pre­vent­ing racial and eth­nic dis­par­i­ties in obe­si­ty risk will also require a devel­op­men­tal approach. By school age, rates of obe­si­ty among black and His­pan­ic chil­dren in the Unit­ed States are high­er than the rates among white chil­dren, even after adjust­ment for socioe­co­nom­ic cir­cum­stances. Many of the risk fac­tors dur­ing preg­nan­cy and ear­ly child­hood are more preva­lent among non­white per­sons, and they explain a sub­stan­tial pro­por­tion of racial and eth­nic dif­fer­ences in obe­si­ty in mid-child­hood.

Sev­er­al fea­tures of preg­nan­cy and infan­cy make the pre­na­tal and post­na­tal peri­ods con­ducive to behav­ior change to reduce the risk of obe­si­ty and its com­pli­ca­tions. First, women appear espe­cial­ly will­ing to mod­i­fy their behav­ior dur­ing these peri­ods to ben­e­fit their chil­dren. Sec­ond, since preg­nant women and infants receive fre­quent rou­tine med­ical care, inter­ven­tions involv­ing improved health care deliv­ery have great poten­tial. Third, these peri­ods are rel­a­tive­ly brief, and we know that behav­ior-change inter­ven­tions are typ­i­cal­ly most suc­cess­ful in the short term. Fourth, if effec­tive inter­ven­tions begun dur­ing preg­nan­cy are main­tained after birth, they will reduce the risk of mater­nal obe­si­ty for future preg­nan­cies and thus help to inter­rupt the inter­gen­er­a­tional cycle.

Ongo­ing inter­ven­tion stud­ies promise to inform med­ical prac­tice and pub­lic health. Many cur­rent tri­als tar­get exces­sive ges­ta­tion­al weight gain, includ­ing sev­en ran­dom­ized, con­trolled tri­als fund­ed by the Nation­al Insti­tutes of Health that will togeth­er include more than 1000 over­weight or obese women and fol­low infants through at least 1 year of age. It remains to be proven, how­ev­er, that reduc­ing ges­ta­tion­al weight gain reduces the obe­si­ty risk in off­spring. An alter­na­tive approach focus­es on dietary qual­i­ty, inde­pen­dent of calo­rie con­tent, to ame­lio­rate mater­nal insulin resis­tance and exces­sive pla­cen­tal nutri­ent trans­fer. Pilot stud­ies have sug­gest­ed that a mul­ti­ple-risk-fac­tor approach dur­ing infan­cy, tar­get­ing moth­ers as con­duits for changes in their infants, can improve sleep dura­tion and delay the intro­duc­tion of sol­id foods.

But even as we await the results of obe­si­ty-pre­ven­tion tri­als, some rec­om­men­da­tions are war­rant­ed because of their ben­e­fi­cial effects on oth­er health out­comes. Preg­nant women should not smoke. Treat­ment of ges­ta­tion­al dia­betes reduces macro­so­mia at birth, although such treat­ment hasn’t been proven to pre­vent obe­si­ty. U.S. rates of elec­tive cesare­an sec­tions have appar­ent­ly lev­eled off, but reduc­ing these rates, espe­cial­ly of cesare­an sec­tions per­formed before 39 weeks of ges­ta­tion, is a pub­lic health goal. Sim­ple sleep-hygiene mea­sures are worth try­ing, even in ear­ly infan­cy. The ide­al age, in terms of aller­gy pre­ven­tion, for intro­duc­ing sol­id foods appears to be 4 to 6 months, and fur­ther research may show that the same is true in terms of obe­si­ty pre­ven­tion.