This study investigated the risk of cancer in Finnish adults living within 500 meters of high voltage powerlines between 1970-1989. It identified 383,700 people exposed to calculated magnetic fields of at least 0.01 microtesla. Cancer cases were identified by linking nationwide health registers. A total of 8,415 cancer cases were observed, similar to expected numbers. Most cancer types showed no significant associations with magnetic field exposure. However, multiple myeloma was elevated in men and colon cancer was elevated in women exposed to higher magnetic fields. Overall, typical residential magnetic fields from powerlines did not appear related to overall cancer risk in adults.
This study investigated the risk of cancer in Finnish adults living within 500 meters of high voltage powerlines between 1970-1989. It identified 383,700 people exposed to calculated magnetic fields of at least 0.01 microtesla. Cancer cases were identified by linking nationwide health registers. A total of 8,415 cancer cases were observed, similar to expected numbers. Most cancer types showed no significant associations with magnetic field exposure. However, multiple myeloma was elevated in men and colon cancer was elevated in women exposed to higher magnetic fields. Overall, typical residential magnetic fields from powerlines did not appear related to overall cancer risk in adults.
Original Description:
Journal paper
Original Title
Magnetic fields of high voltage power lines and risk of cancer in
Finnish adults: nationwide cohort study
This study investigated the risk of cancer in Finnish adults living within 500 meters of high voltage powerlines between 1970-1989. It identified 383,700 people exposed to calculated magnetic fields of at least 0.01 microtesla. Cancer cases were identified by linking nationwide health registers. A total of 8,415 cancer cases were observed, similar to expected numbers. Most cancer types showed no significant associations with magnetic field exposure. However, multiple myeloma was elevated in men and colon cancer was elevated in women exposed to higher magnetic fields. Overall, typical residential magnetic fields from powerlines did not appear related to overall cancer risk in adults.
This study investigated the risk of cancer in Finnish adults living within 500 meters of high voltage powerlines between 1970-1989. It identified 383,700 people exposed to calculated magnetic fields of at least 0.01 microtesla. Cancer cases were identified by linking nationwide health registers. A total of 8,415 cancer cases were observed, similar to expected numbers. Most cancer types showed no significant associations with magnetic field exposure. However, multiple myeloma was elevated in men and colon cancer was elevated in women exposed to higher magnetic fields. Overall, typical residential magnetic fields from powerlines did not appear related to overall cancer risk in adults.
Magnetic f iel d s o f h igh v o l tage po werl ines and ris k o f cancerin
F innis h ad u l ts : natio nwid e co h o rt s tu d y
P ia KV erk as al o , E ero P u k k al a, J aak k o Kaprio , Kau k o V H eik k il a, Mark k u Ko s k env u o A b s tract O b j ectiv e- T o inv es tigate th e ris k o f cancer in as s o ciatio n with m agnetic f iel d s in F innis h ad u l ts l iv ing cl o s e to h igh v o l tage po werl ines . D es ign- N atio nwid e co h o rt s tu d y . S u b j ects - 3 8 3 7 0 0 peo pl e wh o l iv ed d u ring 1 9 7 0 - 8 9 with in 5 0 0 m etres o f o v erh ead po werl ines o f 1 1 0 - 4 0 0 k V in a m agnetic f iel d cal cu l ated to b e 0 . 0 1 F T . S tu d y s u b j ects were id entif ied b y reco rd l ink ages o f natio nwid e regis ters . Main o u tco m e m eas u res - N u m b ers o f o b - s erv ed and ex pected cas es o f cancer, s tand ard is ed incid ence ratio s , and incid ence rate ratio s ad j u s ted f o r s ex , age, cal end ar y ear, and s o cial cl as s - f o r ex am pl e, b y co ntinu o u s cu m u l ativ e ex - po s u re per 1 tT y ear- with 9 5 % co nf id ence inter- v al s f ro m m u l tipl icativ e m o d el s f o r al l cancers co m b ined and 2 1 s el ected ty pes . R es u l ts - A l to geth er 8 4 1 5 cas es o f cancer were o b s erv ed ( s tand ard is ed incid ence ratio 0 . 9 8 ; 9 5 % co nf id ence interv al 0 . 9 6 to 1 . 0 0 ) in ad u l ts . A l l inci- d ence rate ratio s f o r b o th s ex es co m b ined were no n- s ignif icant and b etween 0 . 9 1 and 1 . 1 1 . S ignif icant ex ces s es were o b s erv ed in m u l tipl e m y el o m a in m en ( incid ence rate ratio 1 . 2 2 ) and in co l o n cancer in wo m en ( 1 . 1 6 ) . C o ncl u s io ns - T y pical res id ential m agnetic f iel d s generated b y h igh v o l tage po werl ines d o no t s eem to b e rel ated to th e ris k o f o v eral l cancer in ad u l ts . T h e prev io u s l y s u gges ted as s o ciatio ns b etween ex trem el y l o w f req u ency m agnetic f iel d s and tu m o u rs o f th e nerv o u s s y s tem , l y m ph o m a, and l eu k aem ia in ad u l ts and b reas t cancer in wo m en were no t co nf irm ed . D epartm ent o f P u b l ic H eal th , P O Bo x 4 1 , F I N - 0 0 0 1 4 U niv ers ity o f H el s ink i, F inl and P ia KV erk as al o , res earch er J aak k o Kaprio , s enio r res earch f el l o w Kau k o V H eik k il d , s y s tem s anal y s t F innis h C ancer R egis try , I ns titu te f o r S tatis tical and E pid em io l o gical C ancer R es earch , L iis ank atu 2 1 B, F I N - 0 0 1 7 0 H el s ink i E ero P u k k al a, res earch er D epartm ent o f P u b l ic H eal th , U niv ers ity o f T u rk u , L em m ink f iis enk atu 1 , F I N - 2 0 5 2 0 T u rk u , F inl and Mark k u Ko s k env u o , pro f es s o r C o rres po nd ence to : D rV erk as al o . BMJ 1 9 9 6 ; 3 1 3 : 1 0 4 7 - 5 1 I ntro d u ctio n T h e epid em io l o gical ev id ence f o r a carcino genic ef f ect o f m agnetic f iel d s co m es f ro m th ree paral l el l ines o f res earch : f irs tl y , s tu d ies o n ch il d h o o d cancer in th e v icinity o f po wer l ines ' - 6 ; s eco nd l y , s tu d ies o n ad u l t can- cer in rel atio n to el ectrical o ccu patio ns o r m eas u red m agnetic f iel d s f ro m o ccu patio nal s o u rces ' 7 1 4 ; and , th ird l y , s tu d ies o n ad u l t cancer in rel atio n to m agnetic f iel d s f ro m no n- o ccu patio nal s o u rces incl u d ing po wer l ines . " 2 6 E l ev en epid em io l o gical s tu d ies h av e s o f ar b een pu b - l is h ed o n ad u l t cancer in rel atio n to res id ential m agnetic f iel d s . O f th em , s ev en s tu d ies with co nf l icting res u l ts and at m o s t m o d es t increas es in ris k h av e as s es s ed ex po s u res to po wer l ines " 2 2 and f o u r to el ectric b l ank ets . 2 3 2 6 T h ree s tu d ies h av e incl u d ed al l ty pes o f cancer' 5 - 1 7 2 1 wh ereas f o u r were res tricted to h aem ato - l o gical m al ignancies o r l eu k aem ia, " 8 1 2 0 2 3 o ne incl u d ed b o th l eu k aem ia and b rain tu m o u rs , 2 2 two were o n b reas t cancer in wo m en, 2 5 2 6 and o ne was l im ited to tes ticu l ar cancer. 2 4 We inv es tigated th e ef f ect o f 5 0 H z m agnetic f iel d s o n th e o ccu rrence o f s ev eral ty pes o f cancer. H igh v o l tage po werl ines were ch o s en as th e s o u rce o f m agnetic f iel d s m erel y b ecau s e th ere were ex is ting d o cu m ents f ro m po wer co m panies th at al l o wed th e retro s pectiv e es tim a- tio n o f ex po s u re to m agnetic f iel d s . A natio nwid e ex po - s u re regis ter was created o n pers o nal ex po s u res to 5 0 H z m agnetic f iel d s f ro m h igh v o l tage po wer l ines o v er a perio d o f 2 0 y ears ( 1 9 7 0 - 8 9 ) in F inl and , wh ich m ad e it po s s ib l e to inv es tigate s ev eral ty pes o f cancer in o ne po pu l atio n b as ed co h o rt s tu d y . We pres ent th e res u l ts o n th e rel atio n b etween m agnetic f iel d s o f h igh v o l tage po wer l ines and ris k o f cancer in ad u l ts . S u b j ects and m eth o d s A S S E S S ME N T O F E XP O S U R E T h e ex po s u re as s es s m ent o f th is s tu d y was b as ed o n cal cu l atio ns o f th e av erage annu al m agnetic f iel d s ( m ag- netic f iel d d ens ities ) s eparatel y f o r each o f th e y ears 1 9 7 0 - 8 9 at al l b u il d ings l o cated cl o s er th an 5 0 0 m etres f ro m 1 1 0 k V , 2 2 0 k V , and 4 0 0 k V o v erh ead po wer l ines in F inl and . T h e b as el ine ex po s u re d ata incl u d ed th e l o catio ns o f b u il d ings ( ex act co o rd inates incl u d ed in th e C entral P o pu l atio n R egis ter) and po wer l ines ( in b as ic m aps , s cal e 1 : 2 0 0 0 0 ) , as wel l as th e v o l tage, apparent po wer, and to wer ty pes o f th e 1 1 0 - 4 0 0 k V po wer l ines ( f ro m po werco m pany reco rd s ) . T h es e d ata co v er ab o u t 1 8 0 0 0 k m - th at is , 9 0 % o f th e to tal l ength o f res pectiv e po wer l ines in F inl and . I n th e cal cu l atio n o f m agnetic f iel d s , v ariab l es s u ch as cu rrent, ty pical s patial l o catio ns o f ph as e co nd u cto rs in po wer l ines , and th e s h o rtes t d is tance to th e central po int o f th e b u il d ing were tak en into acco u nt. C u rrents were cal cu l ated f ro m o peratio nal / no m inal v o l tages and apparent po wers o f al l po wer l ines . T h e ex po s u re as s es s m ent is m o re th o ro u gh l y d es crib ed in earl ier pu b l icatio ns . 4 2 7 R E C O R D L I N KA GE S F u rth er d ata co l l ectio n was b as ed ex cl u s iv el y o n th e l ink ages o f natio nwid e regis ters ( th at is , pers o n and b u il d ing d ata f il es o f th e C entral P o pu l atio n R egis ter, P o pu l atio n C ens u s 1 9 7 0 , and F innis h C ancer R egis - try ) . T h e s tu d y co h o rt co ns is ted o f al l peo pl e wh o h ad l iv ed in a b u il d ing with a cal cu l ated m agnetic f iel d o f 0 . 0 1 m icro tes l a ( gT ) f o r any perio d b etween th e y ears 1 9 7 0 - 8 9 . T h is m eans th at th e s tu d y po pu l atio n co m pris ed b o th peo pl e with cl earl y increas ed ex po s u res to res id ential m agnetic f iel d s o f po wer l ines and th o s e wh o h ad b een ex po s ed o nl y to ty pical res id ential b ack - gro u nd l ev el s . T h e gro u p with l eas t cu m u l ativ e ex po s u re ( < 0 . 2 , u T y ears ) s erv ed as th e internal ref erence gro u p in s o m e o f th e cancer anal y s es . A s th e res u l t o f s ev eral s u cces s iv e reco rd l ink ages , th e s u m m ary d ata f il es o f th e s tu d y end ed u p incl u d ing th e f o l l o wing inf o rm atio n: d ates o f b irth and d eath , s ex , res id ential h is to ry with d ates o f m o v ing in and o u t o f th e b u il d ing, h is to ry o f ex po s u re with cal cu l ated av erage annu al m agnetic f iel d s f o r each o f th e cal end ar y ears b etween 1 9 7 0 - 8 9 , s o cial cl as s , and cancer d ata incl u d ing prim ary s ites , h is to l o gical co d es , and d ates o f d iagno s is f o r al l incid ent prim ary tu m o u rs . T h e f o l l o wing prim ary s ites o f cancerwere s el ected to b e s tu d ied : o v eral l cancer, s to m ach ( co d e 1 5 1 in th e I nternatio nal C l as s if icatio n o f D is eas es , s ev enth rev is io n) , co l o n ( 1 5 3 ) , rectu m and recto s igm o id ( 1 5 4 ) , pancreas ( 1 5 7 ) , l ary nx and epigl o ttis ( 1 6 1 ) , l u ng and trach ea ( 1 6 2 . 0 - 1 ) , b reas t in wo m en ( 1 7 0 ) , u terine co rpu s ( 1 7 2 ) , o v ary ( 1 7 5 ) , pro s tate ( 1 7 7 ) , tes tis ( 1 7 8 ) , k id ney ( 1 8 0 ) , b l ad d er, u reter, and u reth ra ( 1 8 1 ) , m el ano m a o f th e s k in ( 1 9 0 ) , no n- m el ano m a o f th e s k in ( 1 9 1 ; ex cl u d es b as al cel l carcino m as ) , nerv o u s s y s tem ( 1 9 3 ) , th y ro id BMJ V O L U ME 3 1 3 2 6 O C T O BE R 1 9 9 6 1 0 4 7 gl and ( 1 9 4 ) , no n- H o d gk in' s l y m ph o m a ( 2 0 0 , 2 0 2 ) , H o d gk in' s d is eas e ( 2 0 1 ) , m u l tipl e m y el o m a ( 2 0 3 ) , and l eu k aem ia ( 2 0 4 - 2 0 8 ) . Gl io m as ( co d e 4 7 , A m erican C ancer S o ciety Manu al o f T u m o r N o m encl atu re C o d ing) and m eningio m as ( co d e 4 6 ) were anal y s ed s eparatel y f ro m o th er tu m o u rs o f th e nerv o u s s y s tem . Breas t can- cer in m en was no t incl u d ed b ecau s e o nl y f o u r cas es were ex pected to o ccu r in th e wh o l e co h o rt. S T A T I S T I C A L A N A L YS E S T h e pers o nal h is to ries ( b etween 1 J anu ary 1 9 7 0 and 3 1 D ecem b er 1 9 8 9 ) o f ex po s u re to res id ential m agnetic f iel d s were s u m m aris ed in th e m eas u re o f cu m u l ativ e ex po s u re, wh ich in th is s tu d y ref ers to th e s u m o f th e pro d u cts o f res id ential m agnetic f iel d ( in , u T ) and l ength o f s u ch ex po s u re ( in y ears ) . C u m u l ativ e ex po s u re was catego ris ed b y u s ing th e ex po s u re cu t o f f po ints o f 0 . 2 , 0 . 4 , 1 . 0 , and 2 . 0 th at were s el ected a prio ri o n th e b as is o f th e d is trib u tio n o f cu m u l ativ e ex po s u re in th e s tu d y po pu l atio n to s et apart s tu d y s u b j ects with d if f erent ex po s u res . T h e d ates o f entering each ex po s u re catego ry were d ef ined f o r each s tu d y s u b j ect and o nl y th e ex po s u res b ef o re th e d ate o f d iagno s is were co ns id - ered in th e cancer anal y s es . F o r d eterm inatio n o f incid ence rates , th e cal cu l atio n o f pers o n y ears b egan o n 1 J anu ary 1 9 7 4 at th e earl ies t o r, if l ater, at b irth o r wh en th e s u b j ect m o v ed into a b u il d ing with a cal cu l ated m agnetic f iel d o f > 0 . 0 1 pT . P ers o n y ears f ro m th o s e aged l es s th an 2 0 y ears were ex cl u d ed f ro m th e anal y s es . C al cu l atio n o f pers o n y ears end ed at d eath o r o n 3 1 D ecem b er 1 9 8 9 . T h e o b s erv ed nu m b ers o f cas es and pers o n y ears at ris k were cal cu l ated b y age ( f iv e y ear age gro u ps ) , s ex , cal end ar perio d ( 1 9 7 4 - 8 1 ; 1 9 8 2 - 9 0 ) , ex po s u re catego ry ( < 0 . 2 0 ; 0 . 2 0 - 0 . 3 9 ; 0 . 4 0 - 0 . 9 9 ; 1 . 0 0 - 1 . 9 9 ; 2 2 . 0 0 gT y ears ) , and s o cial cl as s ( I ( h igh es t) ; I I ; I I I ; I V ; u nk no wn) . E ach ex po s u re catego ry was d eterm ined b y cu m u l ativ e ex po - s u re u p to th e d ate o f d iagno s is , d eath , o r 3 1 D ecem b er 1 9 8 9 . T h e s o cial cl as s v ariab l e was u s ed o nl y in th e incid ence rate ratio anal y s es . T h e ex pected nu m b ers o f cas es were cal cu l ated b y m u l tipl y ing th e s pecif ic nu m b er o f pers o n y ears f o r th e catego ry b y th e co rres po nd ing incid ence o f cancer in F inl and . T h e s tand ard is ed incid ence ratio s were cal cu l ated b y d iv id ing th e o b s erv ed nu m b er o f cas es b y th e nu m b er ex pected . T h e 9 5 % co nf id ence interv al s were d ef ined u nd er th e as s u m ptio n th at th e o b s erv ed nu m b er o f cas es f o l l o wed th e P o is s o n d is trib u tio n. T h e pu rpo s e o f co nd u cting th e incid ence rate ratio anal y s es in ad d itio n to th e s tand ard is ed incid ence ratio anal y s es was , f irs tl y , to u s e th e l eas t ex po s ed gro u p as th e ref erence ( ins tead o f th e general F innis h po pu l a- tio n) and , s eco nd l y , to ad j u s t f o r po tential co nf o u nd ing b y s o cial cl as s . Mu l tipl icativ e P o is s o n regres s io n with s pecif ic nu m b ers o f o b s erv ed cas es and pers o n y ears at ris k f o r each catego ry was u s ed to tak e ex po s u re and co nf o u nd ing v ariab l es into acco u nt s im u l taneo u s l y ; th e incid ence rate ratio s were es tim ated f ro m th es e m o d el s . T h e s ignif icance o f a v ariab l e was as s es s ed b y a l ik el ih o o d ratio tes t; th e criterio n o f s ignif icance was s et at th e 5 % l ev el . A m o d el with s ex , age, cal end ar perio d , ex po s u re, and s o cial cl as s was f itted ( s ee prev io u s para- graph f o r catego ries ) . E x po s u re and age were treated as b o th nu m erical and catego rical v ariab l es ; ex po s u re was ad d itio nal l y treated as a d ich o to m o u s v ariab l e ( in th e res u l ts s ectio n, h o wev er, th e incid ence rate ratio s f o r catego rical and d ich o to m o u s ex po s u re are pres ented in o nl y a f ew cas es o f s pecial interes t) . T h e cal cu l ated m eans o f cu m u l ativ e ex po s u re with in each ex po s u re catego ry ( 0 . 0 5 , 0 . 2 8 , 0 . 6 2 , 1 . 3 8 , and 5 . 0 4 tT y ears ) were u s ed as nu m erical es tim ates o f ex po s u re. T h e h igh es t cu t o f f po int was s et at 1 . 0 , u T y ears in th e anal y s is o f H o d gk in' s d is eas e and 0 . 4 , u T y ears in th e anal y s is o f tes ticu l ar cancer, h o wev er, b ecau s e th e nu m b er o f ex pected cancer cas es in th e h igh es t ex po s u re catego ry wo u l d o th erwis e h av e b een l es s th an two . A d d itio n o f th e age s q u are term was al s o attem pted wh en we u s ed a nu m erical age v ariab l e. T h e po s s ib il ity o f ef f ect m o d if icatio n b y s ex , age, and s o cial cl as s was inv es tigated b y f itting th e co rres po nd ing interactio n term . Wh en appro priate, th e s pecif ic incid ence rate ratio s f o r each catego ry o f co nf o u nd er were al s o es tim ated f ro m m o d el s . R es u l ts T h e s tu d y co h o rt co ns is ted o f 3 8 3 7 0 0 peo pl e ( 1 8 9 3 0 0 m en) wh o co ntrib u ted a to tal o f 2 . 5 m il l io n pers o n y ears af ter th e age 2 0 . A l to geth er 2 7 5 0 0 0 ( 1 1 %) pers o n y ears were as s o ciated with a cu m u l ativ e ex po s u re o f 0 . 4 0 J T y ears ( tab l e 1 ) . T h ere were 8 4 1 5 o b s erv ed cas es o f cancer ( 4 0 8 2 in m en) . T ab l e 2 s h o ws th e nu m b ers b y prim ary s ite. R is k o f cancerin wh o l e co h o rt- T h ere was a d ecreas e o f 2 % in th e ris k o f o v eral l cancer in th e wh o l e s tu d y co h o rt co m pared with th e general F innis h po pu l atio n; th e d ecreas e was o f s im il ar m agnitu d e in m en and wo m en. A l l s tand ard is ed incid ence ratio s s pecif ic f o r a prim ary s ite were b etween 0 . 9 0 and 1 . 1 0 with th e ex ceptio n o f l ary ngeal cancer ( s tand ard is ed incid ence ratio 0 . 8 5 ; 9 5 % co nf id ence interv al 0 . 6 6 to 1 . 0 8 ) . T h e o nl y s ignif icant s tand ard is ed incid ence ratio was o b s erv ed f o r l u ng cancer wh ich was d ecreas ed b y 7 % co m pared with th e general F innis h po pu l atio n ( 0 . 9 3 ; 0 . 8 7 to 0 . 9 8 ) . R E S U L T S F O R S P E C I F I C C A N C E R S T h e f o l l o wing paragraph s s u m m aris e th e res u l ts ; th e o b s erv ed ris k s o f cancer with catego rical and co ntinu - o u s cu m u l ativ e ex po s u re. ( th at is , ch ange in ris k giv en per 1 , T y ear increas e in ex po s u re) are pres ented in m o re d etail in tab l es 2 and 3 . H aem ato l o gical m al ignancies - O v eral l , th e ris k s o f no n- H o d gk in' s l y m ph o m a, H o d gk in' s d is eas e, m u l tipl e m y el o m a, and l eu k aem ia d id no t v ary b y l ev el o f m agnetic f iel d . T h e incid ence rate ratio f o r no n- H o d gk in' s l y m ph o m a was 1 . 0 3 ( tab l e 3 ) . F o r m u l tipl e m y el o m a, a s ignif icant ex ces s ris k ( 1 . 2 2 ) was o b s erv ed in m en b u t a no n- s ignif icant d ecreas e in ris k ( 0 . 8 7 ) was o b s erv ed in wo m en ( tab l e 3 ) . T u m o u rs o f th e nerv o u s s y s tem - N o as s o ciatio n was o b s erv ed b etween th e ris k o f tu m o u rs o f th e nerv o u s T ab l e 1 - D is trib u tio ns o f pers o n y ears b y ex po s u re and co v ariates S tu d y v ariab l e P ers o n y ears ( %) C u m u l ativ e ex po s u re ( pT y ears ) * : < 0 . 2 0 2 0 4 0 7 0 0 ( 8 0 . 2 ) 0 . 2 0 - 0 . 3 9 2 2 2 8 0 0 ( 8 . 8 ) 0 . 4 0 - 0 . 9 9 1 7 0 2 0 0 ( 6 . 7 ) 1 . 0 0 - 1 . 9 9 6 1 7 0 0 ( 2 . 4 ) 2 . 0 0 4 7 8 0 0 ( 1 . 9 ) S ex : Men 1 2 2 2 8 0 0 ( 4 8 . 1 ) Wo m en 1 3 2 0 3 0 0 ( 5 1 . 9 ) A ge ( y ears ) : 2 0 - 3 5 9 7 7 5 0 0 ( 3 8 . 4 ) 3 5 - 4 9 8 1 1 7 0 0 ( 3 1 . 9 ) 5 0 - 6 4 4 6 6 7 0 0 ( 1 8 . 4 ) 6 5 2 8 7 2 0 0 ( 1 1 . 2 ) C al end ar pero d : 1 9 7 4 - 8 1 6 4 3 5 0 0 ( 2 5 . 3 ) 1 9 8 2 - 9 0 1 8 9 9 6 0 0 ( 7 4 . 7 ) S o cial cl as s : 1 7 7 5 0 0 ( 7 . 0 ) 1 1 6 3 9 6 0 0 ( 2 5 . 2 ) I I I 1 2 8 3 6 0 0 ( 5 0 . 5 ) I V 3 2 2 6 0 0 ( 1 2 . 7 ) U nk no wn 1 1 9 8 0 0 ( 4 . 7 ) T o tal 2 5 4 3 1 0 0 * T o tal is 2 5 4 3 2 0 0 b ecau s e o f ef f ect o f ro u nd ing. BMJ V O L U ME 3 1 3 2 6 O C T O BE R 1 9 9 6 1 0 4 8 T ab l e 2 - N u m b ers o f cas es o f cancer b y prim ary s ite acco rd ing to cu m u l ativ e ex po s u re to m agnetic f iel d s o f h igh v o l tage po wer l ines in ad u l ts . F igu res are nu m b ers o f o b s erv ed cas es ; s tand ard is ed incid ence ratio s ( 9 5 % co nf id ence interv al s ) E x po s u re ( pT ) P rim ary s ite < 0 . 2 0 0 . 2 0 - 0 . 3 9 0 . 4 0 - 0 . 9 9 1 . 0 0 - 1 . 9 9 > 2 . 0 0 A l l s ites ( n = 8 4 1 5 ) 6 3 9 3 ; 0 . 9 8 ( 0 . 9 6 to 1 . 0 1 ) 8 7 7 ; 0 . 9 8 ( 0 . 9 1 to 1 . 0 4 ) 6 9 5 ; 0 . 9 7 ( 0 . 9 0 to 1 . 0 4 ) 2 3 0 ; 0 . 8 9 ( 0 . 7 8 to 1 . 0 1 ) 2 2 0 ; 0 . 9 2 ( 0 . 8 1 to 1 . 0 5 ) S to m ach ( n = 6 3 3 ) 4 7 0 ; 0 . 9 9 ( 0 . 9 0 to 1 . 0 8 ) 7 3 ; 1 . 1 1 ( 0 . 8 7 to 1 . 3 9 ) 6 0 ; 1 . 1 3 ( 0 . 8 6 to 1 . 4 5 ) 1 1 ; 0 . 5 8 ( 0 . 2 9 to 1 . 0 3 ) 1 9 ; 1 . 0 6 ( 0 . 6 4 to 1 . 6 6 ) C o l o n ( n = 4 4 4 ) 3 2 4 ; 0 . 9 7 ( 0 . 8 6 to 1 . 0 8 ) 4 5 ; 0 . 9 6 ( 0 . 7 0 to 1 . 2 8 ) 4 5 ; 1 . 1 7 ( 0 . 8 6 to 1 . 5 7 ) 1 2 ; 0 . 8 7 ( 0 . 4 5 to 1 . 5 1 ) 1 8 ; 1 . 4 0 ( 0 . 8 3 to 2 . 2 2 ) R ectu m ( n = 3 1 3 ) 2 3 9 ; 0 . 9 9 ( 0 . 8 7 to 1 . 1 2 ) 3 7 ; 1 . 0 8 ( 0 . 7 6 to 1 . 1 8 ) 2 5 ; 0 . 9 0 ( 0 . 5 8 to 1 . 3 3 ) 3 ; 0 . 3 0 ( 0 . 0 6 to 0 . 8 8 ) 9 ; 0 . 9 6 ( 0 . 4 4 to 1 . 8 2 ) P ancreas ( n = 3 4 1 ) 2 5 4 ; 1 . 0 4 ( 0 . 9 2 to 1 . 1 8 ) 3 9 ; 1 . 1 2 ( 0 . 8 0 to 1 . 5 3 ) 2 7 ; 0 . 9 5 ( 0 . 6 3 to 1 . 3 9 ) 1 3 ; 1 . 2 7 ( 0 . 6 8 to 2 . 1 8 ) 8 ; 0 . 8 3 ( 0 . 3 6 to 1 . 6 4 ) L ary nx ( n = 6 6 ) 4 7 ; 0 . 8 0 ( 0 . 5 9 to 1 . 0 7 ) 1 0 ; 1 . 2 3 ( 0 . 5 9 to 2 . 7 7 ) 3 ; 0 . 4 7 ( 0 . 1 0 to 1 . 3 7 ) 3 ; 1 . 3 0 ( 0 . 2 7 to 3 . 8 1 ) 3 ; 1 . 3 9 ( 0 . 2 9 to 4 . 0 6 ) L u ng ( n = 1 1 2 2 ) 8 5 3 ; 0 . 9 4 ( 0 . 8 8 to 1 . 0 0 ) 1 1 9 ; 0 . 9 3 ( 0 . 7 7 to 1 . 1 1 ) 9 0 ; 0 . 8 8 ( 0 . 7 1 to 1 . 0 8 ) 2 8 ; 0 . 7 7 ( 0 . 5 1 to 1 . 1 1 ) 3 2 ; 0 . 9 1 ( 0 . 6 2 to 1 . 2 9 ) - Breas t ( n = 1 2 2 9 ) 9 4 5 ; 1 . 0 5 ( 0 . 9 8 to 1 . 1 2 ) 1 3 0 ; 1 . 0 6 ( 0 . 8 8 to 1 . 2 5 ) 8 7 ; 0 . 8 9 ( 0 . 7 1 to 1 . 1 0 ) 4 4 ; 1 . 2 2 ( 0 . 8 9 to 1 . 6 4 ) 2 3 ; 0 . 7 5 ( 0 . 4 8 to 1 . 1 3 ) C o rpu s u teri ( n = 2 5 1 ) 1 8 2 ; 0 . 9 4 ( 0 . 8 1 to 1 . 0 8 ) 2 4 ; 0 . 8 7 ( 0 . 5 6 to 1 . 3 0 ) 3 4 ; 1 . 5 4 ( 1 . 0 7 to 2 . 1 5 ) 4 ; 0 . 5 0 ( 0 . 1 4 to 1 . 2 7 ) 7 ; 0 . 9 9 ( 0 . 4 0 to 2 . 0 3 ) O v ary ( n = 2 6 3 ) 2 0 7 ; 1 . 0 3 ( 0 . 9 0 to 1 . 1 8 ) 1 9 ; 0 . 7 1 ( 0 . 4 3 to 1 . 1 1 ) 2 2 ; 1 . 0 4 ( 0 . 6 5 to 1 . 5 7 ) 7 ; 0 . 9 1 ( 0 . 3 6 to 1 . 8 7 ) 8 ; 1 . 2 1 ( 0 . 5 2 to 2 . 3 9 ) P ro s tate ( n = 5 5 7 ) 4 0 0 ; 0 . 9 7 ( 0 . 8 8 to 1 . 0 8 ) 6 6 ; 1 . 0 9 ( 0 . 8 4 to 1 . 3 9 ) 5 9 ; 1 . 2 2 ( 0 . 9 3 to 1 . 5 7 ) 1 0 ; 0 . 5 9 ( 0 . 7 8 to 1 . 0 9 ) 2 2 ; 1 . 2 2 ( 0 . 7 7 to 1 . 8 5 ) T es tis ( n = 3 8 ) 3 2 ; 1 . 0 0 ( 0 . 6 8 to 1 . 1 4 ) 1 ; . 0 3 1 ( 0 . 0 1 to 1 . 7 0 ) 4 ; 1 . 6 0 ( 0 . 4 4 to 4 . 1 0 ) 1 ; 1 . 1 2 ( 0 . 0 3 to 6 . 2 6 ) - ; 0 ( 0 . 0 0 to 5 . 5 1 ) Kid ney ( n = 2 5 8 ) 1 9 1 ; 0 . 9 1 ( 0 . 7 8 to 1 . 0 4 ) 2 9 ; 0 . 9 7 ( 0 . 6 5 to 1 . 4 0 ) 2 7 ; 1 . 1 2 ( 0 . 7 4 to 1 . 6 3 ) 5 ; 0 . 5 8 ( 0 . 1 9 to 1 . 3 4 ) 6 ; 0 . 7 5 ( 0 . 2 8 to 1 . 6 3 ) Bl ad d er ( n = 3 0 0 ) 2 3 7 ; 1 . 0 8 ( 0 . 9 5 to 1 . 2 3 ) 2 9 ; 0 . 9 2 ( 0 . 6 2 to 1 . 3 3 ) 1 7 ; 0 . 6 7 ( 0 . 3 9 to 1 . 0 8 ) 1 1 ; 1 . 2 2 ( 0 . 6 1 to 2 . 1 9 ) 6 ; 0 . 6 8 ( 0 . 2 5 to 1 . 4 9 ) S k in, m el ano m a ( n = 2 4 5 ) 1 7 9 ; 0 . 9 4 ( 0 . 8 1 to 1 . 0 9 ) 2 1 ; 0 . 8 7 ( 0 . 5 4 to 1 . 3 3 ) 2 8 ; 1 . 4 7 ( 0 . 9 8 to 2 . 1 2 ) 1 0 ; 1 . 4 5 ( 0 . 6 9 to 2 . 6 6 ) 7 ; 1 . 1 9 ( 0 . 4 8 to 2 . 4 6 ) S k in, no n- m el ano m a ( n = 2 3 4 ) 1 7 7 ; 1 . 0 1 ( 0 . 8 7 to 1 . 1 7 ) 2 0 ; 0 . 9 2 ( 0 . 5 6 to 1 . 4 2 ) 2 6 ; 1 . 4 4 ( 0 . 9 4 to 2 . 1 2 ) 6 ; 0 . 9 3 ( 0 . 3 4 to 2 . 0 2 ) 5 ; 0 . 8 0 ( 0 . 2 6 to 1 . 8 8 ) N erv o u s s y s tem ( n = 3 0 1 ) 2 3 8 ; 0 . 9 4 ( 0 . 8 2 to 1 . 0 6 ) 3 5 ; 1 . 1 0 ( 0 . 7 7 to 1 . 5 3 ) 1 6 ; 0 . 6 4 ( 0 . 3 7 to 1 . 0 4 ) 5 ; 0 . 5 5 ( 0 . 1 8 to 1 . 2 8 ) 7 ; 0 . 9 2 ( 0 . 3 7 to 1 . 8 9 ) T h y ro id ( n = 1 5 1 ) 1 2 2 ; 1 . 0 1 ( 0 . 8 4 to 1 . 2 1 ) 8 ; 0 . 5 5 ( 0 . 2 4 to 1 . 0 9 ) 1 0 ; 0 . 8 9 ( 0 . 4 3 to 1 . 6 4 ) 7 ; 1 . 7 1 ( 0 . 6 9 to 3 . 5 2 ) 4 ; 1 . 2 0 ( 0 . 3 3 to 3 . 0 8 ) N o n- H o d gk in' s l y m ph o m a ( n = 2 7 3 ) 2 0 8 ; 1 . 0 2 ( 0 . 8 9 to 1 . 1 7 ) 3 4 ; 1 . 3 3 ( 0 . 9 2 to 1 . 8 6 ) 1 2 ; 0 . 5 4 ( 0 . 2 8 to 0 . 9 5 ) 1 0 ; 1 . 2 5 ( 0 . 6 0 to 2 . 2 9 ) 9 ; 1 . 2 5 ( 0 . 5 7 to 2 . 3 7 ) H o d gk in' s d is eas e ( n = 7 3 ) 6 1 ; 1 . 0 1 ( 0 . 7 7 to 1 . 3 0 ) 5 ; 0 . 7 5 ( 0 . 2 4 to 1 . 7 5 ) 4 ; 0 . 7 8 ( 0 . 2 1 to 2 . 0 0 ) 3 ; 1 . 6 4 ( 0 . 3 4 to 4 . 7 6 ) - ; 0 ( 0 . 0 0 to 2 . 5 1 ) Mu l tipl e m y el o m a ( n = 1 1 2 ) 8 2 ; 0 . 9 4 ( 0 . 7 5 to 1 . 1 7 ) 1 2 ; 0 . 9 6 ( 0 . 5 0 to 1 . 6 8 ) 1 0 ; 0 . 9 8 ( 0 . 4 7 to 1 . 8 1 ) 3 ; 0 . 8 2 ( 0 . 1 7 to 2 . 4 0 ) 5 ; 1 . 4 6 ( 0 . 4 7 to 3 . 4 0 ) L eu k aem ia ( n = 2 0 3 ) 1 5 6 ; 0 . 9 6 ( 0 . 8 2 to 1 . 1 3 ) 2 3 ; 1 . 0 6 ( 0 . 6 8 to 1 . 6 0 ) 1 5 ; 0 . 8 7 ( 0 . 4 9 to 1 . 4 3 ) 5 ; 0 . 8 1 ( 0 . 2 6 to 1 . 8 8 ) 4 ; 0 . 7 1 ( 0 . 1 9 to 1 . 8 1 ) T ab l e 3 - I ncid ence rate ratio s ( 9 5 % co nf id ence interv al s ) per 1 P T y ear increas e in ex po s u re b y prim ary s ite and s ex P rim ary s ite Men Wo m en T o tal A l l ty pes 0 . 9 6 ( 0 . 9 2 to 1 . 0 0 ) 1 . 0 0 ( 0 . 9 7 to 1 . 0 4 ) 0 . 9 8 ( 0 . 9 6 to 1 . 0 1 ) S to m ach 0 . 9 3 ( 0 . 8 1 to 1 . 0 8 ) 1 . 0 5 ( 0 . 9 3 to 1 . 1 9 ) 1 . 0 0 ( 0 . 9 1 to 1 . 0 9 ) C o l o n 0 . 9 5 ( 0 . 7 9 to 1 . 1 3 ) 1 . 1 6 ( 1 . 0 3 to 1 . 2 9 ) 1 . 0 7 ( 0 . 9 8 to 1 . 1 8 ) R ectu m 0 . 8 4 ( 0 . 6 5 to 1 . 0 8 ) 1 . 0 5 ( 0 . 8 8 to 1 . 2 4 ) 0 . 9 6 ( 0 . 8 3 to 1 . 1 0 ) L ary nx 1 . 1 2 ( 0 . 8 9 to 1 . 4 1 ) 1 . 1 1 ( 0 . 8 8 to 1 . 3 9 ) P ancreas 0 . 6 6 ( 0 . 4 5 to 0 . 9 7 ) 1 . 1 2 ( 0 . 9 7 to 1 . 2 8 ) 0 . 9 6 ( 0 . 8 4 to 1 . 1 0 ) L u ng 0 . 9 9 ( 0 . 9 2 to 1 . 0 7 ) 0 . 9 4 ( 0 . 7 7 to 1 . 1 6 ) 0 . 9 8 ( 0 . 9 2 to 1 . 0 6 ) Breas t - 0 . 9 5 ( 0 . 8 8 to 1 . 0 2 ) - C o rpu s u teri - 1 . 0 1 ( 0 . 8 8 to 1 . 1 7 ) O v ary - 1 . 0 3 ( 0 . 8 9 to 1 . 1 9 ) - P ro s tate 1 . 0 3 ( 0 . 9 5 to 1 . 1 3 ) - T es tis 1 . 0 6 ( 0 . 5 5 to 2 . 0 5 ) - - Kid ney 0 . 9 7 ( 0 . 7 9 to 1 . 1 8 ) 0 . 9 4 ( 0 . 7 3 to 1 . 2 1 ) 0 . 9 6 ( 0 . 8 2 to 1 . 1 2 ) Bl ad d er 0 . 8 6 ( 0 . 7 0 to 1 . 0 5 ) 1 . 0 5 ( 0 . 8 2 to 1 . 3 6 ) 0 . 9 1 ( 0 . 7 8 to 1 . 0 7 ) Mel ano m a 1 . 0 5 ( 0 . 8 6 to 1 . 2 8 ) 1 . 1 0 ( 0 . 9 2 to 1 . 3 3 ) 1 . 0 8 ( 0 . 9 4 to 1 . 2 3 ) S k in, no n- m el ano m a 0 . 9 0 ( 0 . 6 8 to 1 . 1 8 ) 0 . 9 8 ( 0 . 8 0 to 1 . 2 1 ) 0 . 9 5 ( 0 . 8 0 to 1 . 1 2 ) N erv o u s s y s tem 0 . 9 4 ( 0 . 7 2 to 1 . 2 2 ) 0 . 9 8 ( 0 . 8 1 to 1 . 2 0 ) 0 . 9 7 ( 0 . 8 3 to 1 . 1 3 ) T h y ro id 1 . 0 2 ( 0 . 8 1 to 1 . 2 8 ) 1 . 0 5 ( 0 . 8 7 to 1 . 2 7 ) N o n- H o d gk in' s l y m ph o m a 0 . 9 2 ( 0 . 7 2 to 1 . 1 7 ) 1 . 1 2 ( 0 . 9 5 to 1 . 3 1 ) 1 . 0 3 ( 0 . 9 0 to 1 . 1 8 ) H o d gk in' s d is eas e 0 . 1 8 ( 0 . 0 0 9 9 to 3 . 2 4 ) 0 . 9 3 ( 0 . 6 2 to 1 . 4 0 ) Mu l tipl e m y el o m a 1 . 2 2 ( 1 . 0 0 to 1 . 4 9 ) 0 . 8 7 ( 0 . 5 7 to 1 . 3 2 ) 1 . 0 9 ( 0 . 9 1 to 1 . 3 1 ) L eu k aem ia 0 . 8 3 ( 0 . 5 9 to 1 . 6 7 ) 1 . 0 2 ( 0 . 8 1 to 1 . 2 8 ) 0 . 9 3 ( 0 . 7 7 to 1 . 1 3 ) * T h e ad j u s ted m o d el d o es no t co nv erge. s y s tem and m agnetic f iel d s o f po wer l ines . T h is res u l t was co ns is tent f o r gl io m as and m eningio m as in m en as wel l as wo m en. Mel ano m a o f th e s k in- Mel ano m a o f th e s k in was th e o nl y ty pe o f cancer f o r wh ich th e ris k was s o m e- wh at increas ed th ro u gh o u t th e th ree h igh es t catego ries o f ex po s u re and in b o th s ex es . T h e incid ence rate ratio was 1 . 0 5 f o r m en and 1 . 1 0 f o r wo m en ( tab l e 3 ) . T h e incid ence rate ratio s ( no t s h o wn in tab l es ) f o r th e f o u r co ns ecu tiv e l ev el s o f ex po s u re were 0 . 9 1 ( 0 . 5 8 to 1 . 4 3 ) , 1 . 5 4 ( 1 . 0 3 to 2 . 2 9 ) , 1 . 5 2 ( 0 . 8 0 to 2 . 8 7 ) , and 1 . 2 5 ( 0 . 5 9 to 2 . 6 6 ) in rel atio n to th e l o wes t ex po s u re l ev el . O es tro gen d epend ent cancers - T h e o v eral l ris k o f b reas t cancer in wo m en d ecreas ed no n- s ignif icantl y with l ev el o f m agnetic f iel d ( 0 . 9 5 ; tab l e 3 ) . R is k s o f cancers o f th e u terine co rpu s and o v ary d id no t d if f er b y ex po s u re. C o l o n cancer- T h e ris k o f co l o n cancer increas ed with ex po s u re in wo m en ( 1 . 1 6 ; 1 . 0 3 to 1 . 2 9 ) b u t no t in m en ( 0 . 9 5 ; 0 . 7 9 to 1 . 1 3 ) ( tab l e 3 ) . I n wo m en th e ris k was h igh es t in th e h igh es t ex po s u re catego ry with 2 . 0 , u T y ears ( 2 . 0 3 ; 1 . 1 5 to 3 . 5 8 ) . O th er ty pes o f cancer- We f o u nd no as s o ciatio n b etween l ev el o f m agnetic f iel d and cancers o f th e s to m - ach , rectu m , tes tis , pro s tate, k id ney , b l ad d er, and th y ro id gl and and no n- m el ano m a o f th e s k in. T h e co n- s ecu tiv e incid ence rate ratio s s pecif ic f o r catego ry o f ex po s u re f o r l ary ngeal cancerwere 1 . 0 , 1 . 5 1 , 0 . 5 7 , 1 . 5 9 , and 1 . 6 8 ( al l no n- s ignif icant) and th e incid ence rate ratio was 1 . 1 1 per 1 , u T y ear increas e in ex po s u re ( tab l e 3 ) . I n pancreatic cancer th ere was a s ignif icant d ecreas e in m en ( 0 . 6 6 ) b u t no t in wo m en ( 1 . 1 2 ) . D is cu s s io n ME T H O D O L O GI C A L I S S U E S T h e pres ent s tu d y is th e f irs t natio nwid e s tu d y o n th e rel atio n o f res id ential m agnetic f iel d s and cancer in ad u l ts , and its nu m b er o f ex po s ed cas es ex ceed s th e nu m b ers o f th e prev io u s res id ential co h o rt s tu d ies 1 7 2 1 b y f iv ef o l d to 2 5 - f o l d . T h e as s es s m ent o f ex po s u re to o k into acco u nt h is to rical ex po s u res o v er a perio d o f 2 0 y ears . Mo re th an 2 0 ty pes o f cancer were s creened f o r an as s o ciatio n with m agnetic f iel d s . T h e interpretatio n o f th e res u l ts is a ch al l enging tas k , partl y b ecau s e o ne ris k es tim ate o u t o f 2 0 is ex pected to d if f er f ro m nu l l ( at 9 5 % s ignif icance) b ecau s e o f ch ance al o ne. T h e m agni- tu d e o f th e o b s erv ed ris k s and pres ence o f a d o s e res po ns e, s ignif icance, co ns is tency , and b io l o gical pl au - s ib il ity h av e al l to b e co ns id ered in th e ev al u atio n o f th e res u l ts . I n th e anal y s es with co ntinu o u s cu m u l ativ e ex po s u re, th e incid ence rate ratio s were pres ented f o r a u nit ch ange o f o nl y 1 , u T y ear, wh ich , h o wev er, repres ents th e actu al ex po s u re d if f erences with in th e s tu d y co h o rt. T h e as s es s m ent o f ex po s u re in o u r s tu d y h as b een d is cu s s ed in greater d etail in s o m e prev io u s pu b l icatio ns . 4 2 7 2 8 T h e d ata co l l ectio n was entirel y o n th e b as is o f natio nwid e regis ters , wh ich s et l im its to th e inf o rm atio n av ail ab l e o n po tential co nf o u nd ers . U nl ik e th e earl ier res id ential co h o rt s tu d ies , o u r s tu d y ad j u s ted th e ris k es tim ates f o r s o cial cl as s in ad d itio n to age, s ex , and cal - end ar perio d . T h is d id no t m aterial l y ch ange th e ris k es tim ates o f cancer. S o cial cl as s in F inl and h as b een s h o wn to b e a go o d pro x y o f a v ariety o f l if es ty l e f acto rs as s o ciated with cau s es o f cancer, 2 9 b u t o b v io u s l y th ere is s o m e res id u al co nf o u nd ing in th e pres ent anal y s es . T h e accu racy o f th e po pu l atio n regis tratio n s y s tem in F inl and is h igh , th e co v erage o f cancer regis tratio n is v irtu al l y co m pl ete, 3 0 and th e reco rd l ink age pro ced u res BMJ V O L U ME 3 1 3 2 6 O C T O BE R 1 9 9 6 1 0 4 9 rel iab l e. 3 ' T h eref o re, tech nical reas o ns are no t l ik el y to b ias th e res u l ts . T h e co h o rt is al s o a part o f th e ref erence po pu l atio n. T h o s e ex po s ed > 0 . 2 j T y ears , h o wev er, m ak e u p o nl y s o m e 2 % o f th e to tal po pu l atio n- th at is , th e d il u ting ef f ect in th e s tand ard is ed incid ence ratio s is negl igib l e. F I N D I N GS O v eral l , we f o u nd no m aj o r increas es - o r no increas es at al l - in th e ris k s o f cancer as s o ciated with m agnetic f iel d s o f h igh v o l tage po wer l ines f o r th e 2 1 prim ary s ites inv es tigated . T h is o b s erv atio n is q u ite co ns is tent with th e f ind ings o f th e prev io u s epid em io - l o gical s tu d ies wh ere th e o b s erv ed rel ativ e ris k s h av e m o s t o f ten b een cl o s e to u nity . Magnetic f iel d s h av e prev io u s l y b een as s o ciated with s ev eral ty pes o f cancer s u ch as l eu k aem ia, l y m ph o m a, tu m o u rs o f th e nerv o u s s y s tem , b reas t cancer, m el ano m a, and pro s tate cancer, b u t th e o u tco m e ev ent, if any , h as rem ained m o re o r l es s o b s cu re. I n ad d itio n to th e v ario u s epid em io l o gical o b s erv atio ns , s o m e th eo - retical papers h av e attem pted to b rid ge th e gaps b etween l ab o rato ry o b s erv atio ns and epid em io l o gical cancer s tu d ies with regard to s u s ceptib l e ty pes o f cancer. F o r ins tance, th e m el ato nin h y po th es is s u gges ts a l ink b etween el ectro m agnetic f iel d s and h o rm o ne rel ated cancers o n th e b as is o f th e o b s erv atio ns o f el ec- tro m agnetic f iel d s red u cing m el ato nin. ' 2 We o b s erv ed no increas e in th e ris k s f o r cancers o f f em al e b reas t, o v ary , u terine co rpu s , pro s tate, and tes tis as s o ciated with m agnetic f iel d s o f po wer l ines . T h e ris k s o f h aem ato l o gical m al ignancies and tu m o u rs o f th e nerv o u s s y s tem were al s o eq u al to u nity . T h e prev io u s l y o b s erv ed o r s u gges ted increas es in th es e ty pes 7 were th u s no t co nf irm ed . T h e ris k s o f th e cancers o f th e s to m ach , pancreas , l u ng, k id ney , b l ad d er, and th y ro id gl and and o f no n- m el ano m a o f th e s k in d id no t d if f er b y m agnetic f iel d ex po s u re. T h e ris k o f m el ano m a was th e o nl y cancer f o r wh ich we o b s erv ed s o m e ind icatio n o f an increas e in ris k , wh ich h as in s o m e s tu d ies b een rel ated to m agnetic f iel d s . 8 3 3 3 7 T h e m agnitu d e o f increas e was , h o wev er, rath er l o w ( rel ativ e ris k increas e l es s th an 1 0 % per 1 , u T y ear) , and th e h igh es t rel ativ e ris k o ccu rred in interm e- d iate ex po s u re catego ries . T h e po s s ib l e co nf o u nd ing b y s o l ar and no n- s o l ar ris k d eterm inants ' 8 3 9 co u l d no t b e co ns id ered in th e s tatis tical anal y s es b ecau s e o f th e l ack o f inf o rm atio n o n rel ev ant co nf o u nd ers . I f th e o b s erv ed 8 % increas e in th e ris k o f m el ano m a was d u e to ex trem el y l o w f req u ency m agnetic f iel d s , th is wo u l d co rres po nd to l es s th an o ne ex tra cas e o u t o f s ev eral h u nd red cas es annu al l y in th e F innis h po pu l atio n o f f iv e m il l io n. T h e o b s erv atio n o f a s ignif icant increas e in th e ris k o f cancer o f th e co l o n in wo m en was no t an anticipated f ind ing and m ay wel l b e d u e to ch ance. T h ere is o nl y o ne prev io u s epid em io l o gical o b s erv atio n o f an increas e th e ris k o f co l o n cancer am o ng peo pl e ex po s ed to m ag- netic f iel d s , 8 wh ereas th ree o th er s tu d ies pu b l is h ed s o f ar h av e f o u nd no ex ces s ris k s o f co l o n cancer. ' 2 2 1 4 0 A s to th e po s s ib l e carcino genic m ech anis m in co l o n cancer it h as b een s u gges ted th at th e earl ier o b s erv atio n o f h igh s to res o f iro n in th e b o d y increas ing th e ris k m igh t b e ex pl ained b y iro n h av ing a ro l e in catal y s ing o x y gen rad ical s . 4 ' O n th e o th er h and , 6 0 H z m agnetic f iel d s h av e b een repo rted to pro d u ce ex pres s io n o f trans f errin recepto rs o n h u m an co l o n cancer cel l s in v it: ro 4 2 - th at is , b o d y iro n co u l d act as a co carcino gen with ex trem el y l o w f req u ency m agnetic f iel d s . C O N C L U S I O N S I n co ncl u s io n, th e ex trem el y l o w f req u ency m agnetic f iel d s o f h igh v o l tage po wer l ines at ty pical res id ential l ev el s d o no t s eem to b e as s o ciated with an increas e in Key m es s ages * A l th o u gh s m al l increas es in th e ris k o f cancer in ch il d ren wh o l iv e near po wer l ines and in ad u l ts ex po s ed to m agnetic f iel d s f ro m o ccu patio nal s o u rces h av e b een o b s erv ed , s tu d ies o n res id ential ex po s u res and cancer in ad u l ts h av e b een f ew and inco ncl u s iv e * T h e pres ent po pu l atio n b as ed co h o rt s tu d y am o ng F innis h ad u l ts with 8 5 0 0 o b s erv ed cancer cas es f o u nd no increas e in th e ris k o f o v eral l cancer in ad u l ts ex po s ed to m agnetic f iel d s o f h igh v o l tage po wer l ines * T h e prev io u s l y s u gges ted as s o ciatio ns b etween m agnetic f iel d s and tu m o u rs o f th e nerv o u s s y s tem , l y m ph o m a, l eu k aem ia, and b reas t cancer in wo m en were no t co nf irm ed * T h e res u l ts o f th e pres ent s tu d y s u gges t s tro ngl y th at ty pical res id ential m agnetic f iel d s generated b y h igh v o l tage po werl ines are no t rel ated to cancer in ad u l ts * T h e po s s ib il ity o f an increas e in ris k at h igh er m agnetic f iel d l ev el s , o r in m o re s pecif ic cancer s u b ty pes , canno t b e ex cl u d ed o n th e b as is o f th is s tu d y cancer am o ng ad u l ts ; th is h o l d s f o r th e h aem ato l o gical m al ignancies and f o r tu m o u rs o f th e nerv o u s s y s tem as wel l as f o r th e gro u p o f h o rm o ne rel ated cancers in m al e and f em al e genital s . A s to s pecif ic cancer ty pes , th e ro l e o f ex trem el y l o w f req u ency m agnetic f iel d s in th e path o genes is o f m y el o m a, m el ano m a, and co l o n cancer rem ains u ncertain. We gratef u l l y ack no wl ed ge D rJ o rm a V al j u s ( I V O Gro u p) f o r h is patient wo rk in getting th is s tu d y s tarted and Mr Mik k o H o ngis to ( I V O Gro u p) f o r h is co m preh ens iv e wo rk in th e co m - pu tatio n o f m agnetic f iel d s . We al s o th ank Mr P ek k a J arv inen ( I V O Gro u p) , MrH annu L am m i ( Karttak es k u s O y ) , and Mr Martti Ko rpiranta ( T T - V al tio npal v el u t Q y ) . F u nd ing: I V O Gro u p ( a po wer co m pany ) f u nd ed th e d ata co l l ectio n and as s es s m ent o f ex po s u re. T h e A cad em y o f F inl and , I V O Gro u p, and I nd u s trial P o werT rans m is s io n C o m - pany ( ano th er po wer co m pany ) f u nd ed th e cancer anal y s es and repo rting. C o nf l ict o f interes t: N o ne. 1 Werth eim er N , L eeper E . E l ectrical wiring co nf igu ratio ns and ch il d h o o d cancer. A m J E pid em io l 1 9 7 9 ; 1 0 9 : 2 7 3 - 8 4 . 2 F ey ch ting M, A h l b o m A . Magnetic f iel d s and cancer in ch il d ren l iv ing cl o s e to S wed is h h igh v o l tage po wer l ines . A m J 7 E pid em io l 1 9 9 3 ; 1 3 8 : 4 6 7 - 8 1 . 3 O l s en J H , N iel s en A , S ch u l gen G. R es id ence near h igh v o l tage f acil ities and th e ris k o f cancer in ch il d ren. BMJ 1 9 9 3 ; 3 0 7 : 8 9 1 - 5 . 4 V erk as al o P K, P u k k al a E , H o ngis to MY, V al j u s J , J arv inen P , H eik k il a K, et al . R is k o f cancer in F innis h ch il d ren l iv ing cl o s e to po wer l ines . BM_ 1 9 9 3 ; 3 0 7 : 8 9 5 - 9 . 5 P res to n- Martin S , N av id i W, T h o m as D , L ee P ey - J , Bo wm an J , P o go d a J . L o s A ngel es s tu d y o f res id ential m agnetic f iel d s and ch il d h o o d b rain tu m o u rs . A m J E pid em io l 1 9 9 6 ; 1 4 3 : 1 0 5 - 1 9 . 6 Gu rney J G, Mu el l er BA , D av is S , S ch wartz S M, S tev ens R G, Ko peck y KJ . C h il d h o o d b rain tu m o r o ccu rrence in rel atio n to res id ential po wer l ine co nf igu ratio ns , el ectric h eating s o u rces , and el ectric appl iance u s e. A m J E pid em io l 1 9 9 6 ; 1 4 3 : 1 2 0 - 8 . 7 S av itz D A , A h l b o m A . E pid em io l o gic ev id ence o n cancer in rel atio n to res i- d ential and o ccu patio nal ex po s u res . I n: C arpenterD O , A y rapety an S , ed s . Bio l o gical ef f ects o f el ectric and m agnetic f iel d s . Benef icial and h arm f id ef f ects . V o l 2 . S an D iego : A cad em ic P res s , 1 9 9 4 : 2 3 3 - 6 1 . 8 T y nes T , A nd ers en A , L angm ark F . I ncid ence o f cancer in N o rwegian wo rk ers po tential l y ex po s ed to el ectro m agnetic f iel d s . A m J E pid em io l 1 9 9 2 ; 1 3 6 : 8 1 - 8 . 9 Gu enel P , R as k m ark P , A nd ers en J B, L y nge E . I ncid ence o f cancer in per- s o ns with o ccu patio nal ex po s u re to el ectro m agnetic f iel d s in D enm ark . Br J I nd Med 1 9 9 3 ; 4 0 : 7 5 8 - 6 4 . 1 0 F l o d eru s B, P eras o n T , S tenl u nd C , Wennb erg A , O at A , Knav e B. O ccu pa- tio nal ex po s u re tO el ectro m agnetic f iel d s in rel atio n to l eu k aem ia and b rain tu m o ra: a cas e- co ntro l s tu d y in S wed en. C ancer C au s es C entral 1 9 9 3 ; 4 : 4 6 5 - 7 6 . 1 1 S ah l J D , Kel s h MA , Greenl and S . C o h o rt and nes ted cas e- co ntro l s tu d ies o f h em ato po ietic cancera and b rain cancer am o ng el ectric u til ity wo rk era. E pid em io l o gy 1 9 9 3 ; 4 : 1 0 4 - 1 4 . 1 2 T h eriau l t G, Go l d b erg M, Mil l er A B, A rm s tro ng B, Gu enel P , D ead m an J , et al . C ancer ris k s as s o ciated with o ccu patio nal ex po s u re to m agnetic f iel d s am o ng el ectric u til ity wo rk ers in O ntario and Q u eb ec, C anad a, and F rance: 1 9 7 0 - 1 9 8 9 . A m J aE pid em io l 1 9 9 4 ; 1 3 9 : 5 5 0 - 7 2 . 1 0 5 0 BMJ V O L U ME 3 1 3 2 6 O C T O BE R 1 9 9 6 1 3 S av itz D A , L o o m is D P . Magnetic f iel d ex po s u re in rel atio n to l eu k em ia and b rain cancer m o rtal ity am o ng el ectric u til ity wo rk ers . A m I E pid em io l 1 9 9 5 ; 1 4 1 : 1 2 3 - 3 4 . 1 4 L o o m is D P , S av itz D A , A nanth C V . Breas t cancer m o rtal ity am o ng f em al e el ectrical wo rk ers in th e U nited S tates . J N atl C ancer I ns t 1 9 9 4 ; 8 6 : 9 2 1 - 5 . 1 5 Werth eim er N , L eeper E . A d u l t cancer rel ated to el ectrical wires near th e h o m e. I ntJ E pid em io l 1 9 8 2 ; 1 1 : 3 4 5 - 5 5 . 1 6 Werth eim er N , L eeper E . Magnetic f iel d ex po s u re rel ated to cancer s u b ty pes . A nn N YA cad S ci 1 9 8 7 ; 5 0 2 : 4 3 - 5 4 . 1 7 McD o wal l ME . Mo rtal ity o f pers o ns res id ent in th e v icinity o f el ectricity trans m is s io n f acil ities . Bry C ancer 1 9 8 6 ; 5 3 : 2 7 1 - 9 . 1 8 S ev ers o n R K, S tev ens R G, Kau ne WT , T h o m as D B, H eu s er L , D av is S , et al . A cu te no nl y m ph o cy tic l eu k em ia and res id ential ex po s u re to po wer f req u ency m agnetic f iel d s . A m j E pid em io l 1 9 8 8 ; 1 2 8 : 1 0 - 2 0 . 1 9 C o l em an MP , Bel l C M, T ay l o r H L , P rim ic- Z ak el j M. L eu k aem ia and res i- d ence near el ectricity trans m is s io n eq u ipm ent: a cas e- co ntro l s tu d y . BrJ C ancer 1 9 8 9 ; 6 0 : 7 9 3 - 8 . 2 0 Yo u ngs o n J H A M, C l ay d en A D , My ers A , C artwrigh t R A . A cas e/ co ntro l s tu d y o f ad u l t h aem ato l o gical m al ignancies in rel atio n to o v erh ead po wer- l ines . BrJ C ancer 1 9 9 1 ; 6 3 : 9 7 7 - 8 5 . 2 1 S ch reib er GH , S waen GM, Meij ers J M, S l angen J J , S tu rm ans F . C ancer m o rtal ity and res id ence near el ectricity trans m is s io n eq u ipm ent: a retro - s pectiv e co h o rt s tu d y . I ntJ E pid em io l 1 9 9 3 ; 2 2 : 9 - 1 5 . 2 2 F ey ch ting M, A h l b o m A . Magnetic f iel d s , l eu k em ia, and central nerv o u s s y s tem tu m o rs in S wed is h ad u l ts res id ing near h igh - v o l tage po wer l ines . E pid em io l o gy 1 9 9 4 ; 5 : 5 0 1 - 9 . 2 3 P res to n- Martin S , P eters J M, Yu MC , Garab rant D H , Bo wm an J D . My el o geno u s l eu k em ia and el ectric b l ank et u s e. Bio el ecrro m agnetics 1 9 8 8 ; 9 : 2 0 7 - 1 3 . 2 4 V erreau l t R , Weis s N S , H o l l enb ach KA , S trad er C H , D al ing J R . U s e o f el ectric b l ank ets and ris k o f tes ticu l ar cancer. A m J E pid em io l 1 9 9 0 ; 1 3 1 : 7 5 9 - 6 2 . 2 5 V ena J E , Grah am S , H el l m ann R , S wans o n M, Bras u re J . U s e o f el ectric b l ank ets and ris k o f po s tm eno pau s al b reas t cancer. A m J E pid em io l 1 9 9 1 ; 1 3 4 : 1 8 0 - 5 . 2 6 V ena J E , F reu d enh eim J L , Mars h al l J R , L au gh l in R , S wans o n M, S ax o n G. R is k o f prem eno pau s al b reas t cancer and u s e o f el ectric b l ank ets . A m J E pid em io l 1 9 9 4 ; 1 4 0 : 9 7 4 - 9 . 2 7 V al j u s J , H o ngis to M, V erk as al o P , J arv inen P , H eik k il a K, Ko s k env u o M. R es id ential ex po s u re to m agnetic f iel d s generated b y 1 1 0 - 4 0 0 k V po wer l ines in F inl and . Bio el ectro m agnetics 1 9 9 5 ; 1 6 : 3 6 5 - 7 6 . 2 8 V erk as al o P , P u k k al a E , H o ngis to M, V al j u s J , J arv inen P , H eik k il a K, et al . P o wer- f req u ency m agnetic f iel d s and po s s ib l e ad v ers e h eal th ef f ects ( l et- ter) . BMJ 1 9 9 4 ; 3 0 8 : 1 1 6 2 - 3 . 2 9 P u k k al a E . C ancer ris k b y s o cial cl as s and o ccu patio n. A s u rv ey o f 1 0 9 , 0 0 0 cancer cas es am o ng F inns o f wo rk ing age. I n: Wah rend o rf J , ed . C o ntrib u - tio ns to epid em io l o gy and b io s tatis tics . V o l 7 . Bas el : Karger, 1 9 9 5 . 3 0 T eppo L , P u k k al a E , L eh to nen M. D ata q u al ity and q u al ity co ntro l o f a po pu l atio n- b as ed cancer regis try . A cta O nco l 1 9 9 4 ; 3 3 : 3 6 5 - 9 . 3 1 P u k k al a E . U s e o f reco rd l ink age in s m al l - area s tu d ies . I n: E l l io t P , Gu zick J , E ngl is h D , S tern R , ed s . Geo graph ical and env iro nm ental epid em io l o gy . O x f o rd : O x f o rd U niv ers ity P res s , 1 9 9 2 : 1 2 5 - 3 1 . 3 2 S tev ens R G. E l ectric po wer u s e and b reas t cancer: a h y po th es is . A m J f E pi- d em io l 1 9 8 7 ; 1 2 5 : 5 5 6 - 6 1 . 3 3 V A gero D , O l in R . I ncid ence o f cancer in th e el ectro nics ind u s try : u s ing th e new S wed is h C ancer E nv iro nm ent R egis try as a s creening ins tru m ent. Br J I nd Med 1 9 8 3 ; 4 0 : 1 8 8 - 9 2 . 3 4 V agero D , A h l b o m A , S ah l s ten S . C ancer m o rb id ity am o ng wo rk ers in th e tel eco m m u nicatio ns ind u s try . BrJ I nd Med 1 9 8 5 ; 4 2 : 1 9 1 - 5 . 3 5 O l in R , V A gero D , A h l b o m A . Mo rtal ity ex perience o f el ectrical engineers . BrJ I nd Med 1 9 8 5 ; 4 2 : 2 1 1 - 2 . 3 6 S o rah an T , Waterh o u s e J A H , McKiernan MJ , A s to n R H . C ancer incid ence and cancer m o rtal ity in a co h o rt o f s em ico nd u cto r wo rk ers . BrJ I nd Med 1 9 8 5 ; 4 2 : 5 4 6 - 5 0 . 3 7 D e Gu ire L , T h ' I riau l t G, I tu rra H , P ro v ench er S , C y r D , C as e BW I ncreas ed incid ence o f m al ignant m el ano m a o f th e s k in in a tel eco m m u - nicatio ns ind u s try . Br I nd Med 1 9 8 8 ; 4 5 : 8 2 4 . 3 8 Wh item an D , Green A . Mel ano m a and s u nb u rn. C ancer C au s es C o ntro l 1 9 9 4 ; S : 5 6 4 - 7 2 . 3 9 R o ck l ey P F , T rief f N , Wagner R F J , T y ring S K. N o ns u nl igh t ris k f acto rs f o r m al ignant m el ano m a. I . C h em ical agents , ph y s ical co nd itio ns , and o ccu - patio n. I nty D erm ato l 1 9 9 4 ; 3 3 : 3 9 8 - 4 0 6 . 4 0 T y nes T , R eitan J B, A nd ers en A . I ncid ence o f cancer am o ng wo rk ers in N o rwegian h y d ro el ectric po wer co m panies . S cand Wo rk E nv iro n H eal th 1 9 9 4 ; 2 0 : 3 3 9 - 4 4 . 4 1 S tev ens R G. Bio l o gical l y b as ed epid em io l o gical s tu d ies o f el ectric po wer and cancer. E nv iro n H eal th P ers p S u ppl 1 9 9 3 ; 1 0 1 ( s u ppl 4 ) : 9 3 - 1 0 0 . 4 2 P h il l ips J L , R u tl ed ge L , Winters WD . T rans f errin b ind ing to two h u m an co l o n carcino m a cel l l ines : ch aracterizatio n and ef f ect o f 6 0 - H z el ectro m agnetic f iel d s . C ancer R es earch 1 9 8 6 ; 4 6 : 2 3 9 - 4 4 . ( A ccepted 6 S eptem b er 1 9 9 6 ) A ccid ent and T rau m a R es earch C enter, U KK I ns titu te f o r H eal th P ro m o tio n R es earch , P O Bo x 3 0 , F I N - 3 3 5 0 1 T am pere, F inl and P ek k a Kannu s , ch ief ph y s ician Mik a P al v anen, res earch f el l o w S eppo N iem i, res earch as s is tant J ari P ark k ari, res earch f el l o w I l k k a V u o ri, pro f es s o r D epartm ent o f S u rgery , T am pere U niv ers ity H o s pital , T am pere, F inl and Mark k u J arv inen, pro f es s o r C o rres po nd ence to : D rKannu s . BM7 1 9 9 6 ; 3 1 3 : 1 0 5 1 - 2 I ncreas ing nu m b erand incid ence o f o s teo po ro tic f ractu res o f th e pro x im al h u m eru s in el d erl y peo pl e P ek k a Kannu s , Mik a P al v anen, S eppo N iem i, J ari P ark k ari, Mark k u J arv inen, I l k k a V u o ri T h e m o s t co m m o n f o rm s o f o s teo po ro tic f ractu res o ccu r at th e d is tal f o rearm , s pine, and h ip. T reating l es s co m m o n f ractu res ( s u ch as th o s e o f th e pro x im al h u m eru s , pel v is , k nee, and ank l e' ) is al s o d em and ing and ex pens iv e. Yet epid em io l o gical inf o rm atio n o n th es e is s carce; no natio nwid e s tu d y o f th e incid ence o f th e o s teo po ro tic f ractu res o f th e pro x im al h u m eru s h as b een repo rted . S u b j ects , m eth o d s , and res u l ts We d ef ined an o s teo po ro tic f ractu re o f th e pro x im al h u m eru s as a f ractu re o ccu rring in peo pl e aged 2 6 0 th at res u l ted f ro m m inim al trau m a o nl y - th at is , a f al l f ro m s tand ing h eigh t o r l es s . We s el ected al l s u ch patients ad m itted to F innis h h o s pital s in 1 9 7 0 - 2 , 1 9 7 4 - 5 , 1 9 7 8 - 8 0 , 1 9 8 3 - 5 , 1 9 8 8 - 9 , and 1 9 9 1 - 3 f o r prim ary treatm ent o f f irs t f ractu re o f th e pro x im al h u m eru s f ro m th e N atio nal H o s pital D is ch arge R egis ter. P rev io u s s tu d ies h av e f o u nd th e accu racy o f th e regis ter to b e 9 6 _ 9 7 %. 2 3 4 T h e s tu d y co v ered th e entire F innis h po pu l atio n. We ad j u s ted f o r th e s u b j ects ' age b y u s ing th e m ean po pu l atio n b etween 1 9 7 0 and 1 9 9 3 as th e s tand ard po pu l atio n and ex pres s ed f ractu re incid ences as th e nu m b er o f cas es / l 0 0 0 0 0 pers o ns / y ear b y s ex and b y age. T h e to tal nu m b er o f o s teo po ro tic f ractu res o f th e pro x im al h u m eru s increas ed s tead il y d u ring th e s tu d y perio d , f ro m 2 1 2 in 1 9 7 0 to 8 4 4 in 1 9 9 3 ( an av erage increas e o f 1 3 % a y ear) . T h e m ean age o f patients al s o increas ed , f ro m 7 2 . 1 y ears ( 1 9 7 0 ) to 7 6 . 2 y ears ( 1 9 9 3 ) . T h e age ad j u s ted incid ence ( per 1 0 0 0 0 0 peo pl e aged 1 2 0 - Wo m en 1 0 0 - 8 0 - 0 z 6 0 Men 4 0 - < 2 0 - O 9 7 0 1 9 7 5 1 9 8 0 1 9 8 5 1 9 9 0 Year F ig 1 A ge ad j u s ted incid ence o f o s teo po ro tic f ractu res o f th e pro x im al h u m eru s in inh ab itants o f F inl and aged > 6 0 d u ring 1 9 7 0 - 9 3 , 6 0 ) o f f ractu res increas ed d u ring 1 9 7 0 - 9 3 f ro m 5 0 to 1 0 6 in wo m en and f ro m 1 4 to 4 1 in m en ( f ig 1 ) . C o m m ent We f o u nd a s tead y increas e in th e age s pecif ic incid ences f o r f ractu res o f th e pro x im al h u m eru s in b o th s ex es and in al l ages o v er th e co u rs e o f th e s tu d y . I n F inl and no t o nl y is th e incid ence o f f ractu res increas ing b u t al s o th e po pu l atio n at ris k is increas ing- a d ev el o p- m ent l ik el y to accel erate in th e nearf u tu re. T h e increas - ing m ean age o f patients is l ik el y to m ean m o re pro b l em s in treating th es e f ractu res , incl u d ing l o nger tim es f o r h eal ing and reh ab il itatio n and m o re co m pl ica- tio ns o f f ractu res s u ch as no n- u nio ns . A d d itio nal l y , a h igh erm ean age o f patients is l ik el y to increas e m o rb id - ity and m o rtal ity . BMJ V O L U ME 3 1 3 2 6 O C T O BE R 1 9 9 6 1 0 5 1