Severs skada – paradigmskifte gällande diagnostik och behandling?

Severs skada – paradigmskifte gällande 

diagnostik och behandling? 

Ett planerat avhandlingsarbete av dr Stefan Perhamre 

[ Lat. söka ] 

Rapport 1• 2013 

Centrum för klinisk forskning 

Landstinget i Värmland

Severs skada – paradigmskifte 

gällande diagnostik och behandling? 

Ett planerat avhandlingsarbete av dr Stefan Perhamre 

Maria Klässbo 1 , Rolf Norlin 2 

1 medicine doktor, sjukgymnast, forskningsledare, 


Landstinget i Värmland, maria.klassbo@liv.se 

2 professor i ortopedi, Institutionen för hälsovetenskap och medicin 

Örebro universitet, rolf.norlin@orebroll.se 

Vår forskargrupp: 

Stefan Perhamre, leg läk, doktorand, Karlstad†. 

Maria Klässbo, med dr, Karlstad. 

Dagmara Lazowska, leg läk, Karlstad. 

Sofia Papageorgiou, leg läk, Karlstad. 

Fredrik Lundin, tekn dr, Karlstad. 

Staffan Janson, professor, Karlstad. 

Anders Lundin, med dr, Örebro. 

Rolf Norlin, professor, Örebro. 

Nyckelord: 

Severs skada, apofysit, calcaneus, inlägg, hälkopp 

Karlstad 2013

Rapporten kan beställas från 


Gamla Spinneriet - Älvgatan 49 

651 85 Karlstad 

054-61 70 75 

pia.tekin@liv.se 

Tryckt hos: Arkitektkopia, Karlstad 2013 

ISSN 1652-9324

Till Ola Wessmark, mister Idrottsmedicin i Värmland 

”Teorier vi tror på kallar vi fakta, 

fakta vi misstror kallar vi teorier.” 

okänd 

5

Innehåll 

SAMMANFATTNING ....................................................................................7 

ARTIKLAR .......................................................................................................8 

Definitioner och förkortningar .......................................................................9 

INLEDNING .................................................................................................10 

BAKGRUND .................................................................................................12 

SYFTE ............................................................................................................14 

MATERIAL ....................................................................................................15 

Deltagare delarbete I, IV och V ....................................................................16 

Deltagare delarbete II ...................................................................................17 

Deltagare delarbete III .................................................................................17 

METOD OCH RESULTAT ...........................................................................18 

Metod och resultat delarbete I ......................................................................18 

Metod och resultat delarbete II ....................................................................19 

Metod och resultat delarbete III ...................................................................23 

Hård hälkopp är bättre än kilklack .............................................................24 

Metod och resultat delarbete IV ...................................................................24 

Metod och resultat delarbete V ....................................................................26 

SLUTSATS .....................................................................................................27 

TACK .............................................................................................................28 

STEFAN PERHAMRES EGNA TACK I HANS 

PLANERADE AVHANDLING .....................................................................29 

REFERENSER ...............................................................................................32 

BILAGOR ......................................................................................................33 

Bilaga 1 ........................................................................................................33 

Bilaga 2 ........................................................................................................34 

Bilaga 3 ........................................................................................................35 

Bilaga 4 ........................................................................................................36 

Bilaga 5 ........................................................................................................41 

Bilaga 6 ........................................................................................................47

Sammanfattning 

Stefan Perhamre, idrottsläkare i Karlstad samarbetade tidigare med dr 

Ola Wessmark, och ortopedtekniker Ingemar Hansson vid Idrottshälsan, 

Landstinget i Värmland. Till Idrottshälsan sökte sig många barn och ungdomar 

med hälsmärta och de tyckte att en hård hälkopp fungerade påfallande 

bra som smärtlindring. 

Stefan Perhamre startade ett forskningsprojekt, tillsammans med Maria 

Klässbo, för att utvärdera hur bra hälkoppen verkligen var. När det blev 

klart att Stefan ville skriva en doktorsavhandling bildades vår forskargrupp 

kring hans avhandlingsprojekt. 

Med Stefans arbeten som grund har vi kunnat visa att diagnosen Severs 

skada (hälsmärta hos unga) kan fastställas med vanliga kliniska test. 

Slätröntgen har inget diagnostiskt värde. Studierna har visat att den hälkopp 

Ola Wessmark tidigare designat, fungerar mycket bra som behandling 

vid Severs skada och att orsaken till besvären rimligen är en överbelastningsreaktion 

i hälbenet. Vi har vidare kunnat visa att hälkoppen bidrar till 

att hälkudden trycks ihop från sidorna och blir lite tjockare under hälbenet 

och att toppbelastningen på hälen minskar. 

Stefan avled 2010, efter en tids sjukdom, men forskargruppen har kunnat 

gå vidare med Stefans artiklar så att de har publicerats. 

7

Artiklar 

Denna rapport är baserad på följande artiklar: 

I. Perhamre S, Lazowska D, Papageorgiou S, Lundin F, Klässbo M, Norlin 

R. Sever’s injury; a clinical diagnosis. J Am Podiatr Med Assoc, accepterad 

II. Perhamre S, Janson S, Norlin R, Klässbo M. Sever's injury: treatment 

with insoles provides effective pain relief. Scand J Med Sci Sports. 

2011;21(6):819-23 

III. Perhamre S, Lundin F, Norlin R, Klässbo M. Sever's injury; treat it 

with a heel cup: a randomized, crossover study with two insole alternatives. 

Scand J Med Sci Sports. 2011;21(6):e42-7 

IV. Perhamre S, Lundin F, Klässbo M, Norlin R. A heel cup improves the 

function of the heel pad in Sever's injury: effects on heel pad thickness, 

peak pressure and pain. Scand J Med Sci Sports. 2012;22(4):516-22 

V. Norlin R, Perhamre S, Lazowska D, Papageorgiou S, Klässbo M. 

Lundin A. MRI-findings in Sever’s injury are reduced during treatment: 

A randomised case-control study. Manuskript. 

Norlin R. Severs skada - lättbehandlad hälsmärta hos ungdomar. Svensk 

Idrottsforskning 2011;4:43-6. 

och Stefan Perhamres eget utkast till sin planerade avhandling. 

8

Definitioner och förkortningar 

VAS Visuell analog skala 0-10. 

Borg CR-10 VAS 0-10 med tillagda verbala ankare. 

IQR Kvartilsavstånd (Interquartile range), skillnaden mellan den 

tredje och första kvartilen. Kvartilavståndet anger inom vil 

k-et avstånd de 50% mittersta observationerna ligger (Q3- 

Q1). 

MRI/MRT Magnet Resonance Image/Tomography och kallas på sven- 

s ka för magnetkameraundersökning. Radiologisk metod 

som ger en avbildning av både skelett och mjukdelar. 

Diafys Skaftet, den mellersta delen, av ett rörben. 

Metafys Den del av ett långt rörben där övergången sker mellan 

skaft et och den förtjockade änden av benet. 

Fysen Den zon av änden på ett ben där längdtillväxten sker och 

som ligger mellan metafys och epifys/apofys. 

Epifys De förtjockade ändarna av ett rörben, klädda med led- 

brosk, som ligger i anslutning till en led. 

Apofys De förtjockade ändarna av ett rörben som inte ligger i an- 

slutning till en led, exempelvis hälbenets bakersta del. 

9

Inledning 

Idrottshälsan startade år 1986 och var från 1989 organiserad i Friskvården i 

Värmland, men finansierad direkt av Landstinget i Värmland. År 2003 tog 

landstinget över ansvaret för Idrottshälsan. Idrottshälsan bemannades inledningsvis 

med idrottsläkare, sekreterare och assistent. År 2009 anställdes 

också en sjukgymnast. Till Idrottshälsan kan den som drabbats av skador i 

leder och muskler, som har aktiva motionsvanor eller är aktiv inom idrotten, 

vända sig utan remiss för att få hjälp med bedömning, behandling, 

träning och förslag på förebyggande åtgärder. Man kan också söka om man 

har allmänna medicinska frågeställningar, om det har relevans för den fysiska 

aktiviteten, t ex luftrörsbesvär, hjärt-, allergi- eller infektionsproblem. 

Barn och ungdomar som har hälsohinder som påverkar deras förmåga att 

vara fysiskt aktiva i skolan eller inom sina idrotter är en stor målgrupp för 

Idrottshälsan. 

Ola Wessmark, specialist i ortopedi, var den förste idrottsläkaren på 

Idrottshälsan. Många barn och ungdomar sökte Idrottshälsan för ömmande 

hälar. Ola beskriver själv att han ”av en slump” 1979 kom på att göra 

en avlastning av en ömmande häl i Ortoplastmaterial för att hjälpa en 

fotbollsspelande son, bilaga 1 1 . Resultatet var sensationellt skriver Ola vidare. 

Sonen kunde följande dag återuppta sin träning och behövde sen 

aldrig avstå träning eller match på grund av besvär från hälen. Wessmark 

gick sen vidare och gjorde, med hjälp av Ingemar Hansson som också 

var anställd på Idrottshälsan, många inlägg som kom att kallas hälkoppar. 

Något år senare gjorde Wessmark en telefonuppföljning av 46 barn, varav 

sju flickor. Medianåldern för flickorna var 11,6 år och för pojkarna 12,6 

år. Hela 39/46 hade kunnat återgå till träning/tävling utan fortsatta besvär 

då inläggen användes. Sex kunde delta i aktiviteter trots besvär, men med 

besvär i mindre grad. Endast en hade så stora besvär att han inte kunde 

delta i fysisk aktivitet. 

10

Idrottsläkarna vid Idrottshälsan, Landstinget i Värmland, Ola Wessmark och Stefan Perhamre. 

År 1993 började Stefan Perhamre arbeta på Idrottshälsan. Perhamre var 

specialist i allmänmedicin och hade ett mycket stort intresse för idrottsmedicin. 

Också Stefan slogs av den goda effekten av de hårda individgjutna 

hälkopparna. År 2000 gjorde Stefan en enklare utvärdering då han skickade 

en enkät med fem frågor och fick in svar från 83/115 barn (tabell 1). 

Tabell 1: 

Utvärdering av behandling med hälkopp hos barn med hälsmärta år 2000 (n=83). 

11

Bakgrund 

Severs skada, smärta i hälen som känns vid varje hälisättning och ökar med 

högre aktivitetsnivå, drabbar växande ungdomar, vanligen i åldern 8 – 15 

år. Severs skada är vanligast bland dem som är aktiva inom löpning och 

fotboll. 

Redan 1907 beskrev den svenske röntgenläkaren Patrik Haglund hälsmärta 

av liknande karaktär 2 . Numera är tillståndet namngivet efter amerikanen 

James W Sever 3 . 

Genom åren har många teorier framförts både kring smärtorsak och 

kring bästa behandling. Inflammation i hälbenets tillväxtzon (apophysitis 

calcanei) är en spridd förklaring 3 , men också ökat drag i hälsenan. Felställning 

i fot och underben har framförts som möjliga förklaringar. Ogden har 

lanserat synsättet att tillståndet är en överbelastningsskada i apofysen i det 

icke mogna hälen, därav benämningen Severs skada 4 . 

Smärta vid vissa aktiviteter är symptomet vid Severs skada. Ett tioårigt 

barn har oftast begränsad erfarenhet av smärta och att tänka abstrakt. 

Smärta hos barn har framför allt studerats när det gäller kroniska sjukdomar 

5,6,7 . Man har funnit att redan vid femårs ålder kan barn kvantifiera 

smärta och vid åtta års ålder värdera dessa kvaliteter. Eftersom det hos barn 

och ungdom kan finnas könsskillnader gällande smärta kan det vara enklare 

att välja att i studier inkludera enbart ett kön. Om man ska mäta smärta 

med hjälp av en enkät är det avgörande om mätmetoden har god validitet, 

mäter rätt sak. Visuell analog skala (VAS) anses vara ”gold standard” att 

använda för barn och ungdom. Borgs CR-10 skala är en vidareutveckling 

av VAS skalan med verbala ankare där noll representerar ingen smärta och 

10 extrem smärta (bilaga 2) 8 . Den kliniska erfarenheten är att barn och 

ungdomar med hälsmärta inte har några som helst svårigheter att använda 

VAS-skalan för att visa hur ont de har vid en viss aktivitet. 

Många olika behandlingar har förespråkats och använts genom åren. 

Då man misstänkt inflammation som grundorsak till besvären har allt från 

olika fysikaliska behandlingar till läkemedel och till och med gipsning använts. 

Man har även spekulerat i att felställning i foten kan utlösa besvären 

varför olika korrigerande inlägg har använts som behandling. Inte sällan 

behandlas Severs skada på samma sätt som man tidigare behandlat hälsenesmärta 

hos vuxna, till exempel med allt från töjningar till avlastande 

kilinlägg. 

12

Vila och olika sjukgymnastiska behandlingar har testats genom åren. För 

fysiskt aktiva barn och ungdomar kan rekommendationen vila och uppehåll 

med fysisk aktivitet upplevas mycket negativ. 

Åldern nio till 15 år anses vara den viktigaste åldern när det gäller att 

bygga upp skelett och neuromuskulär funktion med betydelse för resten 

av livet. Sportaktiviteter är också en källa till livsglädje och en kulturform 

med speciella traditioner och normer. Att förlora ett år eller mer på grund 

av hälsmärta kan ödelägga senare sportande. Lars-Magnus Engström har 

utvecklat ett aktivitetsindex för att kunna följa barn och ungdomars fysiska 

aktivitet (se bilaga 3) 9 . 

Om inte förr så försvinner problemet med Severs skada när hälbenets 

tillväxtzon växer samman med resten av hälbenet, alltså när längdtillväxten 

avstannar. Därför är ibland effekten av behandling svår att bedöma, särskilt 

hos de äldre barnen som är nära att avsluta sin längdtillväxt. 

Det finns ingen konsensus i litteraturen när det gäller diagnostik av 

Severs skada vid hälsmärta, inte hur man behandlar eller vad som orsakar 

smärtan. 

13

Syfte 

Syftet med studierna var att ge förslag på hur Severs skada kan diagnosticeras 

med kliniska test (delarbete I), behandlas med skoinlägg (delarbete 

II, III och bilagorna 4,5) och att ge tänkbara förklaringar till varför det ena 

alternativet, hälkoppen, hade överlägsna smärtlindrande effekter jämfört 

med hälkil (delarbete IV, V och bilaga 6). 

14

Material 

Delarbetena I, IV och V är baserade på studie 2 med datainsamling åren 

2008-2009 och delarbetena II och III är baserade på studie 1 med datainsamling 

2004-2005. 

Deltagare 

Tabell 2: Översikt över deltagarna i de olika delarbetena. 

15

Deltagare delarbete I, IV och V 

Deltagarna i studie 2, delarbetena I, IV och V, (n=45 sammanlagt) var dels 

barn med hälsmärta som randomiserades till behandlingar med två olika 

inlägg (n=30) och femton smärtfria matchade kontroller (tabell 2). De 

femton matchade kontrollerna hade samma kön, ålder, fysisk aktivitetsnivå 

och var aktiva i samma idrottsklubb som de med hälsmärta. Inga deltagare 

föll bort. 

Inklusionskriterier för alla 45 deltagare var: 

- pojkar eller flickor 9-15 år 

- nivå D eller E på Engströms aktivitetsindex (se bilaga 3) 9 

- kan göra sig förstådd på och förstår själv svenska utan problem 

- har förälders/föräldrars skriftliga samtycke till att delta 

För de 30 barn som skulle prova inlägg krävdes också att 

- de hade haft hälsmärta i distala delen av hälen i mer än två veckor 

- den mest smärtsamma aktiviteten måste vara bollsport i någon form (fotboll, 

innebandy, handboll eller liknande) med en smärtnivå på fyra eller 

mer, den senaste veckan, registrerat med Borgs CR-10 skala 8 

- de blivit medicinskt undersökta och konstaterats ha aktuellt problem 

Exklusionskriterier var: 

- smärta lokaliserad till Akillessenan eller hälbenet som helhet 

- smärta på natten 

- intermittent smärta 

- spridd smärta eller svullnad 

- annan specificerad sjukdom som kunde interferera med hälsmärtan 

- tidigare frakturer i området 

- svårdefinierad värk i nedre extremitet 

- medverkan i annan studie som innefattade någon typ av smärtbehandling 

Deltagarna rekryterades från olika idrottsklubbar dit information om studien 

skickades ut. 

I delarbete IV deltog, förutom de ovan nämnda 45 barnen, också en 

pilotgrupp av barn med hälsmärta (n=5), där man testade den röntgenologiska 

undersökningsmetodiken. 

16

Deltagare delarbete II 

Trettioåtta pojkar med hälsmärta uppfyllde inklusionskriternan, men tre 

ville inte delta i studien. I denna prospektiva randomiserade studie accepterade 

inledningsvis 35 pojkar (två fullföljde inte (n= 33)), med en medianålder 

på 12 år (range 9-14) med Severs skada, att delta (tabell 2). Arton 

randomiserades till att använda hälkil och 17 till hälkopp. Fem pojkar, 

medianålder 12 år (range 10-13 år), rekryterades konsekutivt till en kontrollgrupp 

som inte fick några inlägg. Alla deltagare var höggradigt fysiskt 

aktiva i olika idrotter. 

Inklusionskriterier var: 

- pojkar 9-15 år 

- hälsmärta i mer än två veckor och

Metod och resultat 

Alla barn som ingick i studierna fick inledningsvis ange ålder, längd, vikt 

och registrera sin fysiska aktivitetsnivå från A-E med Engströms aktivitetsindex 

(se bilaga 3) 9 . Den fysiska aktivitetsnivån registrerades sedan vid 

varje datainsamlingstillfälle. 

Studierna har godkänts av etikprövningsnämnden i Uppsala, dnr 2004 

M-377 och 2008/094. Alla deltagare gav sitt informerade samtyckte till 

att delta i studierna. I delarbete IV användes radiologi för att analysera 

hälkuddens tjocklek. Av etiska skäl undersöktes först en pilotgrupp (n=5) 

för att tillse att undersökningsmetodiken var tillfredställande. 

Metod och resultat delarbete I 

Syftet med delarbete I var att undersöka diagnostisk precision av olika 

kliniska test gällande hälsmärta vid Severs skada. Barnen i delarbete I, 30 

ungdomar med symtom och 15 symtomfria, fick svara på frågor om ålder, 

längd, vikt, vilken sport de utövade och fysisk aktivitetsnivå från A-E med 

Engströms aktivitetsindex (se bilaga 3) 9 . Barnen med hälsmärta fick ange 

hur länge de haft hälsmärta och smärta i höger/vänster/bilteralt. 

Tre kliniska test utfördes av Stefan Perhamre på alla barn (n=45). I de 

fall smärtan var bilateral fick barnet själva avgöra vilken häl som var mest 

smärtande och på den utfördes testerna. Test ett (A) var barfota hälstående, 

test två (B) tuberkelpalpation (tryck på hälbenets bakersta topp) och test 

tre (C) sidokompression (eng. squeeze test), figur 1. Upplevd smärta vid 

testerna registrerades på Borgs CR-10 skala 8 . Testaren (SP) tränade trycket 

vid sidokompression (test C) på en tryckkänslig gummiblåsa (vigorimeter) 

för att standardisera tryckets storlek. 

Figur 1: De tre kliniska testen som användes för att diagnosticera Severs skada. 

18

En slätröntgenbild med lateral projektion togs på alla barn (n=45). För 

barnen med hälsmärta avbildades den mest smärtande hälen. Två röntgenologer 

analyserade, oberoende av varandra och blindade för hälsmärta 

eller ej, graden av skleros (ökad täthet) på en fyrgradig-skala; ingen skleros, 

lätt, måttlig och höggradig skleros. De bedömde också graden av fragmentering 

av tillväxtzonen (apofysen) också det på en fyrgradig-skala; ingen 

fragmentering, två, tre eller minst fyra fragment. Förekomst av distala fragment 

eller inte noterades också. 

De matchade kontrollbarnen testades på samma sätt både med de kliniska 

testen och med slätröntgenundersökning. 

I delarbete I beräknades såväl sensitivitet (testets förmåga att skilja ut 

de sjuka individerna), specificitet (testets förmåga att skilja ut de friska 

individerna), negativt och positivt prediktivt värde för de tre kliniska testen 

och för de röntgenologiska (skleros, fragmentering och förekomst av 

distala fragment). För de röntgenologiska testerna beräknades sensitivitet 

och specificitet med olika cut-off värden där den kombination som gav 

bäst värden för båda valdes, där sensitivitet ansågs vara viktigare än specificitet. 

Inledningsvisa skillnader mellan grupperna analyserades med Fishers 

exakta test för data på nominal- och ordinalskalenivå och med Students 

t-test för övriga. 

Tre bra kliniska test vid Severs skada medan röntgenförändringar är 

svårtolkade 

Alla tre testen; barfota hälstående, tuberpalpation och sidokompression 

(eng. squeeze test) (figur 1) uppvisade 100 procents specificitet vilket 

innebär att alla tre testen kunde identifiera alla de friska ungdomarna som 

friska. När det gäller sensitivitet hade hälstående 100 procent, tuberpalpation 

80 procent och sidokompression 97 procents träffsäkerhet. 

Röntgen av hälen påvisade att skleros fanns hos alla ungdomar, både 

symtomatiska och asymtomatiska i kontrollgruppen. Även vid gradering 

av sklerosen noterades samma mönster i båda grupperna. Fragmentering 

var vanligare i gruppen med besvär (87 procent) jämfört med kontrollgruppen 

(53 procent). Denna skillnad var signifikant (p=0,01). Samstämmighet 

mellan de båda radiologernas bedömning (reliabilitet) var 100 procent 

för både skleros och fragmentering. 

Metod och resultat delarbete II 

Syftet med delarbete II var att undersöka om hälinlägg, hälkopp och 

hälkil, kunde minska smärtan hos pojkar med hälsmärta, Severs skada 

(bilaga 4) 12 . Trettioåtta pojkar uppfyllde inklusionskriterierna. Tre föll bort 

av olika skäl. Resterande 35 pojkar, mellan 9 och 14 år, fick slumpmässigt 

19

ehandling med antingen hälkopp (n=17) eller en avlastande hälkil (n=18) 

(figur 2). De slumpandes genom att ta en lapp i en låda försedd med lock. 

Deltagarna fick välja två fysiska aktiviteter; aktivitet A och B, där hälen 

smärtade och som de utövade minst 3 gånger per vecka. Aktivitet A skulle 

vara den som gav mest smärta. Under de två första veckorna efter randomisering 

gavs ingen behandling. Ungdomarna registrerade själva smärta 

vid aktivitet A och B (Borgs CR-10 skala) och aktivitetsnivå (Engströms 

aktivitetsindex), se bilaga 3 och figur 3. Därefter använde hälften av ungdomarna 

hälkopp och hälften hälkil under 4 veckor. Slutligen följde en ny 

två-veckorsperiod utan behandling. 

För att få en uppfattning om naturalförloppet rekryterades konsekutivt 

fem patienter med Severs skada som inte fick några inlägg. De fick registrera 

smärta och aktivitetsnivå vid två tillfällen vid start och efter två månader. 

Figur 2: Hälkopp eller hälkil som ska användas i en stadig och knuten idrottssko. 

20

Figur 3. Studiedesign för delarbete II. 

21

I delarbete II beräknades mediansmärta för varje pojke i alla tre faser av 

studien, före, under och efter behandling, och separat för aktivitet A och 

B. Mediansmärta och kvartilsavvikelse för hela gruppen (n=33) beräknades 

sen för varje fas och aktivitet. Wilcoxons signed-rank test användes för att 

räkna ut förändring över tid separat för aktivitet A och B, med jämförelser 

mellan faserna, se figur 5. Holms sekventiella Bonferroni test användes för 

att korrigera för massignifikans. 

Figur 5: Staplar som visar smärta (median) i faserna före, under och efter behandling för de två självvalda 

smärtande fysiska aktiviteterna i studiegruppen (n=33) där A är den mest smärtande, delarbete II. Jämförelser 

mellan mediansmärta i faserna presenteras med p-värden. 

Inlägg reducerar smärtan påtagligt 

Resultatet blev att, trots att alla ungdomar bibehöll en hög fysisk aktivitetsnivå 

(Engströms aktivitetsnivå D och E), besvären reducerades signifikant 

under den period de använde något av inläggen. Mätt med Borgs CR-10 

skala (visuell analog skala mellan 0 och 10) sjönk smärtan från cirka 4,5 till 

2. Efter behandlingsperioden återkom smärtan, om än på en något lägre 

nivå än före behandlingen, cirka 3,5. 

I den obehandlade kontrollgruppen skedde ingen påtaglig förändring 

av smärtnivån under de 8 veckornas observation. En eventuell skillnad i 

behandlingseffekt mellan inläggen kunde inte säkert bedömas. 

22

Metod och resultat delarbete III 

Syftet med delarbete III var att analysera vilket inlägg, hälkil eller hälkopp, 

som gav mest smärtlindring hos ungdomar med hälsmärta, Severs skada 

(bilaga 5) 13 . 

Fyrtiofyra ungdomar med typiska hälsmärtor randomiserades till två 

olika behandlingsgrupper. Ungdomarna följdes under totalt 26 veckor. 

Efter en två veckor lång basal bedömningsperiod utan behandling (B1) 

randomiserades de antingen att få hälkopp (n=20) eller hälkil (n=24) att 

använda i fyra veckor i sina sportskor (vecka 3-6). Därefter följde en två 

veckor lång basperiod utan behandling (B2), varefter de fick använda den 

andra typen av inlägget under fyra veckor (vecka 9-12) i en cross-over design. 

Slutligen fick de efter en tredje basperiod (B3) utan behandling, fritt 

välja det inlägg de tyckte var bäst eller välja att vara utan inlägg. Denna 

tredje, valfria behandlingsperiod var 12 veckor lång (figur 4). Efter ett år 

fick deltagarna svara på frågor om grad av smärtlindring under studiens 

26 veckor, typ av inlägg de använt och smärtlindringsgrad under året. Tre 

pojkar svarade inte (n=41). 

Figur 4: Design av jämförande studie mellan hälkopp och hälkil, delarbete III. 

En ordinal mixad regression, där den initiala smärtan modellerades med 

tid och behandling som förklarande variabler, utfördes. Effekten av varje 

förklarande variabel var en oddskvot (OR), som representerade den multiplicerade 

effekten av förändring i varje variabel. Självrapporterad smärta 

(Borgs CR-10, 0-10) kategoriserades till fem nivåer (0-2, 3-4, 5-6, 7-8 

och 9-10) och smärtnivån vid varje mättillfälle modellerades genom att 

använda initial smärta och behandling som fixed effekt, antal dagar sedan 

baslinje 1-3 som en fixed och random effekt och patient som en random 

intercept. Låga oddskvoter tranformerades till mindre smärta och höga 

oddskvoter till mer smärta. 

23

Som ett beskrivande komplement till regressionsanalyserna konstruerades 

ett diagram med smärta (median med kvartilsavstånd) för varje fas 

(figur 6). Smärtlindringen för hälkil och hälkopp jämfördes inkluderande 

båda de valda fysiska aktiviteterna. 

Figur 6: Separerade resultat, median och kvartilsavstånd, för grupperna som randomiserats till att använda 

hälkopp och hälkil i delarbete III. Figuren visar de tre baslinjefaserna före behandling, efter första randomiserade 

alternativet och efter det andra då inlägg byttes (cross-over). Grupp I (vita staplar) startade med hälkopp 

och bytte sen till hälkil. Grupp II (gråa staplar) startade med hälkil. Resultatet innefattar smärtskattningar 

från både den mest smärtande aktiviteten, A, och den näst mest smärtande aktiviteten, B. 

Hård hälkopp är bättre än kilklack 

Utvärdering visade att de som använde hälkopp upplevde signifikant mindre 

smärta än de som använde kil. Detta visades också i att 77 procent av 

ungdomarna valde att fortsätta med hälkoppen. Ingen valde, vid det fria 

valet, att avstå från behandling. 

Metod och resultat delarbete IV 

Syftet med delarbete IV var att studera två tänkbara orsaker till varför 

hälkoppen ger så bra smärtlindring omedelbart när ungdomarna börjar 

använda den 14 . Vi ville studera om hälkudden ökade i tjocklek när man använde 

hälkoppen, vilket borde reducera de direkta tryckkrafterna på hälbenet 

vid belastning med kroppsvikten. Dessutom ville vi mäta fördelningen 

av tryckkrafterna under foten vid löpning. Bland 30 ungdomar med Severs 

24

skada randomiserades 15 till behandling med hälkopp och 15 fick ingen 

behandling. Ytterligare 15 ungdomar utan hälsmärta var kontrollgrupp. 

Hälkuddens tjocklek mättes på en röntgenbild (sidobild tagen stående 

på ett ben, hälstående, på den mest smärtande hälen på de med hälsmärta 

(n=30) och på samma häl för de matchade kontrollerna de utan hälsmärta 

(n=15)). Tre bilder togs; barfota (alla 45 ungdomar), med idrottssko (alla 

45 ungdomar)och med hälkopp i idrottssko (enbart behandlingsgruppen, 

n=30). 

Med mätutrustning (F-Scan från Tek scan) för tryckmätning kunde av 

ekonomiska skäl endast 10 ungdomars belastningsmönster under foten 

analyseras. Dessa 10 valdes ur behandlingsgruppen. En tryckkänslig film 

registrerade trycket inom varje kvadratcentimeter under hela foten. Tre 

mätningar gjorde under hälstående. Dessutom mättes belastningsmönstret 

under foten under nio löpsteg med löphastighet 100 (metronom). Dessa 

mätningar gjordes både med enbart idrottssko och med hälkopp i idrottsskon. 

Maximal belastning noterades under häl, framtrampet och under 

stortån. 

De röntgenologiska mätningarna, tjocklek i millimeter (medelvärde), 

konfidensintervall och p-värden, analyserades med Students t-test. En linjär 

mixad regressionsmodell användes för att analysera toppbelastning som 

utfallsvariabel, där varje barn användes som random effekt och löpning 1, 

2, 3 och hälkopp eller ej som fixed effekt. I båda fallen användes de logaritmerade 

värdena av toppbelastning. 

Hälkoppen ger tjockare hälkudde och lägre maximal tryck under foten 

I stående skiljde inte hälkuddens tjocklek mellan symtomatiska ungdomar 

och de i den symtomfria kontrollgruppen. Hälkuddens tjocklek ligger i 

denna grupp ungdomar mellan cirka 6,5 och 8 millimeter. Med enbart 

idrottssko på foten ökade hälkuddens tjocklek med drygt 2 millimeter, 

vilket kan förklara att många ungdomar upplever mindre besvär när de 

använder bra skor. Med hälkopp i idrottsskon ökade hälkuddens tjocklek 

ytterligare drygt 1,5 millimeter vilket kan förklara varför ungdomarna 

upplever ännu mindre smärta när de också har hälkoppen på plats (Tabell 

3). 

Tabell 3: Hälkuddens tjocklek vid stående på ett ben, hälstående, delarbete IV. 

25

Dataanalys visade att maximal tryckbelastning under hälstående minskade 

med 25% när hälkopp användes, jämfört med tryckvärdet när enbart 

idrottssko användes. Under löpning minskade maximal belastning med 

21% om hälkopp användes. Även denna signifikanta minskning av det 

maximala trycket kan förklara den minskning av belastningssmärta som 

ungdomarna noterar när de använder hälkoppen. 

Metod och resultat delarbete V 

Syftet med delarbete V var att, med magnetkamera (MR), undersöka hälarna 

på ungdomar med och utan hälsmärta och även undersöka de med 

hälsmärta efter att de behandlats med hälkopp. Samma 45 ungdomar 

som i studie I och IV, där trettio av dem hade hälsmärta, ingick i studien. 

Femton av de 30 med besvär hade fått behandling med hälkopp och 15 

hade varit utan behandling. Alla genomgick undersökning med MR både 

före behandlingsstart och en månad senare. Även kontrollgruppen (n=15) 

genomgick MR. 

Icke parametriska metoder, Wilcoxons rank sum test och Kruskal-Wallis 

test användes för att jämföra MR fynden mellan patienter och kontroller 

och före och efter behandling med hälkopp. 

Magnetkameraundersökning (MR) påvisar tecken på skaderelaterad skelettpåverkan 

Preliminära data, från denna ännu ej publicerad studie, visar att alla 

ungdomar, alltså även symtomfria, hade viss grad av ödem i hälbenet. Dock 

har de med Severs skada signifikant mer ödem. Detta ödem reduceras signifikant 

efter en månads behandling (figur 7) då också belastningssmärtan 

minskat påtagligt. 

Figur 7: Magnetkamerabilder, delarbete V. 

26

Slutsats 

Diagnosen Severs skada kan fastställas vid ett vanligt besök hos läkare eller 

sjukgymnast där belastningssmärta strikt lokaliserad till hälen, tydlig smärta 

vid hälstående, smärta vid sidokompression av hälen och vanligen också 

tryckömhet över tuber (mitt bak på hälknölen) är typiska fynd. Dessa tre 

test kan mycket väl skilja sjuka från friska ungdomar när det gäller Severs 

skada. Trots skillnader i graden av fragmentering av hälbenet är det inte 

särskilt meningsfullt att använda vanlig röntgen för att säkerställa diagnosen 

Severs skada. 

Inläggsbehandling, hälkil eller hälkopp, reducerar hälsmärta påtagligt 

så att en hög fysisk aktivitetsnivå kan bibehållas. Behandling med styv hälkopp, 

som formas direkt på hälen, ger snabb symtomlindring och bättre 

smärtlindring för de flesta jämfört med hälkil. 

Hälkoppen ökar hälkuddens tjocklek och minskar det maximala trycket 

på hälen. Besvären försvinner alltid när hälbenets tillväxt avstannar, men 

med behandling med hälkopp minskar eller försvinner besvären mycket 

snabbare utan att begränsa barnens fysiska aktivitet. 

Severs skada kan mycket väl vara en ren belastningsutlöst skada på hälbenet 

hos växande ungdomar som idrottar. Magnetkameraundersökning 

visar ett klart ökat ödem i hälbenet och detta ödem minskar under behandling, 

parallellt med att smärtan minskar. 

27

Tack 

Tack för finansiellt stöd till: 

Landstinget i Värmland; Primärvårdens FoU-enhet, Division hälsa, habilitering 

och rehabilitering, Centrum för klinisk forskning och Fonden för 

ortopedtekniska hjälpmedel. Tack också till Karlstad universitet, Örebro 

universitet och Centrum för idrottsforskning. 

28

Stefan Perhamres egna tack i 

hans planerade avhandling 

I would like to express my sincere and warm gratitude to: 

Rolf Norlin, my tutor and co-author who accepted me as a doctoral student 

when he started as professor at Örebro University. Thank you for 

guiding me through the project with lots of wisdom, support and patience. 

Maria Klässbo, my co-tutor and co-author who alone had to carry the 

burden guiding a scientific blueberry in the beginning of my scientific 

journey. Your work capacity, patience and accuracy will take you far. 

Staffan Janson, my co-author in the first paper, professor in two disciplines, 

and the visionary humanist who started up the local scientific and progressive 

group in Värmland County Council, where I later was included. 

You showed me the way. 

Fredrik Lundin, the appreciated statistician and co-author, who has been 

the guarantee for correct data analyses. 

Ingemar Hansson, who has produced all the heel cups in the studies. We 

have had a close relationship through all these years. Your loyalty has been 

unbeatable. 

Christer Carlmark, art responsible at the Centre for Public Health Research, 

Värmland County Council and Klara Perhamre, my daughter, for 

help with the lay-out of some figures. 

Ola Wessmark, the former chief doctor at our clinic, who was an intelligent 

clinical model for me during our first years together at the clinic. You 

made the first moulded Orfit heel cup. Without this prototype there would 

have been no scientific work by me on this item. 

Dagmara Makowska (Lazowska) and Sofia Papageorgiou, radiologists 

at the Central Hospital Karlstad, who did a professional job examining the 

radiograph pictures in Paper IV. 

29

Lars Hynsjö, M. Sc at the department of Chemistry, Sahlgrenska University 

Hospital for his stimulating input with multi-variat analyses of the 

topics. 

Freddy Gustavsen, former chief of the company Adapt AB, who has been 

an important enthusiast and discussion partner. Your great network and 

generosity has been much worth. 

Carola Dimitrov¸ product manager CA Mätsystem AB for initiated support 

in the peak pressure measurements and analyses. 

Mikael Hasselgren, medical chief of the Centre for Clinical Science, for 

his faith in this project and decisions of financial support from Värmland 

County Council. 

Eva Stjernström, divisional chief of the Värmland County Council for 

her faith in this project and financial support despite a difficult economic 

situation in the Division of Health, Habilitation and Rehabilitation, 

Värmland County Council . 

Göran Ejlertsson, professor in Public Human Science, who helped me in 

my academic statistical studies. 

Ulf Tidefelt and Gunilla Ahlsén, prefect and study headmistress of Örebro 

University, for solid support all the way through. 

Kerstin Granström and Margaretha Warnqvist, two excellent secretaries 

at the Sports Clinic and the Scientific Unit for friendship, help and support. 

We have been “tight-knit families” at work. 

Gordon Riemersma, GP in Karlstad and the examiner of our English. 

Håkan Alfredson, professor in Sports Medicine Umeå University for a 

solid friendship and active support in this project. 

Peter Nilsson, chief of the Orthopaedic device Company in association 

with Värmland County Council for your interest in developing science in 

this field. 

30

Lars Magnus Engström, former professor Dep. of Social and Cultural 

Studies in Education, Stockholm Institute of Education for support with 

the questionnaires. 

Volunteers and patients who participated in this study. 

My family, Anneli, Klara, Axel and Maja, for everything. 

31

Referenser 

1 Wessmark O. Apofysitis calcanei, Haglund-Severs sjukdom. Idrottsme- 

dicin 1987;3:10 

2 Haglund P. Archiv für klinische Chirurgie 1907;82:922-30 

3 Sever JW. Apophysitis of the os calcis. New York Medical Journal 

1912;95:1025-9 

4 Ogden JA, Ganey TM, Hill JD, Jaakkola JI. Sever’s injury: A stress fracture 

of the immature calcaneal metaphysis. J Pediatr Orthop 2004;24:488–92 

5 McGrath P, Finley GA. Measurement of Pain in Infants and Children. 

Seattle: IASP Press 1998: 210 

6 McGrath P, Finley GA. Chronic and Recurrent Pain in Children and 

Adolescents. Seattle: IASP Press 1999: 288 

7 O´Rourke D. The measurement of pain in infants, children, and adolescents: 

From policy to practice. Phys Ther. 2004:84:560-70 

8 Neely G, Ljunggren G, Sylén C, Borg G. Comparison between the 

Visual Analogue Scale (VAS) and the Category Ratio Scale (CR-10) for 

evaluation of leg exertion. Int J Sports Med 1992;13:133-6 

9 Engström LM. Barns och ungdomars idrottsvanor i förändring. Svensk 

Idrottsforskning 2004;4:1-6 

10 Perhamre S, Lazowska D, Papageorgiou S, Lundin F, Klässbo M, Norlin 

R. Sever’s injury; a clinical diagnosis. Accepted J Am Podiatr Med Assoc 

11 Perhamre S, Janson S, Norlin R, Klässbo M. Sever's injury: treatment 

with insoles provides effective pain relief. Scand J Med Sci Sports 

2011;21(6):819-23 

12 Perhamre S, Lundin F, Norlin R, Klässbo M. Sever's injury; treat it with a 

heel cup: a randomized, crossover study with two insole alternatives. Scand 

J Med Sci Sports 2011;21(6):e42-7 

13 Perhamre S, Lundin F, Klässbo M, Norlin R. A heel cup improves the 

function of the heel pad in Sever's injury: effects on heel pad thickness, 

peak pressure and pain. Scand J Med Sci Sports 2012;22(4):516-22 

14 Norlin R, Perhamre S, Lazowska D, Papageorgiou S, Klässbo M, Lundin 

A. MRI-findings in Sever’s injury are reduced during treatment: A randomised 

case-control study. Submitted 

32

Bilagor 

Att göra en hälkopp – 

Bilaga 1 

bästa behandlingen vid Severs skada (apofysit i calcaneus). 

Plastmaterialet ORFIT Classic Soft 3,2 mm tjock (ej perforerat) används. En skiva (60 x 90 

cm) brukar räcka till 10 inlägg och köps i Sverige via Össur (art.nr. 671 48 88). 

Barnet får ställa foten på plattan och lämpligt område ritas ut. Därefter klipps detta ut. Plattan 

läggs i varmt vattenbad (45 grader) under 1 minut (t.ex. Hydrokollator, Össur), varefter den 

direkt läggs på den obelastade foten. Plasten formas manuellt med särskild tillklämning kring 

hälen för att hålla hälkudden på plats. När plasten är modellerad fixeras den med elastisk linda 

mot foten och barnet får sitta stilla med obelastad fot i 6 minuter. Därefter klipps inlägget till 

så att det passar foten. 

Efter detta är inlägget färdigt att användas. Barnet får gå med inlägget för att kontrollera att 

det inte skaver. Inlägg görs alltid till båda fötter, även vid ensidiga besvär. 

Total tillverkningstid cirka 30 min/par. 

Särskilda synpunkter: 

Undvik det perforerade materialet eftersom dessa ofta spricker. 

För tunga personer kan detta material vara för tunt och spricka. 

Inlägget kan även användas vid hälsporre. 

Ingemar Hansson lär gärna ut hur detta går till i praktiken. 

Hans kontaktuppgifter är: 

Idrottshalsan.karlstad@liv.se 

070-340 94 07 

Ingemar Hansson/Rolf Norlin 

28 

33

Bilaga 2 

34 

Namn: Testnummer: v 1 

Dessa frågor skall du besvara för de aktiviteter du valt varje vecka under hela undersökningen. Du tar med det 

ifyllda protokollet när du kommer till Idrottshälsan och behåller det sedan för nästa period. Två siffror för smärta 

och två siffror för hur foten fungerar skall fyllas i varje vecka. ( Fyll i dina två valda aktiviteter på de streckade 

raderna). 

Hur ont gör det? 

För att kunna beskriva smärtan för aktivitet 1 För att kunna beskriva smärtan för aktivitet 2 

......................................................skall du ......................................................skall du 

ange en siffra från 0 till 10 eller om det är ange en siffra från 0 till 10 eller om det är 

maximalt anges det med ett +. Ringa in en maximalt anges det med ett +. Ringa in en 

siffra som passar bäst för aktiviteten denna vecka. siffra som passar bäst för aktiviteten denna 

vecka. 

0 Ingen alls 0 Ingen alls 

0.5 Extremt svag (knappt kännbar) 0.5 Extremt svag (knappt kännbar) 

1 Mycket svag 1 Mycket svag 

2 Svag (lätt) 2 Svag (lätt) 

3 Måttlig 3 Måttlig 

4 4 

5 Stark (kraftig) 5 Stark (kraftig) 

6 6 

7 Mycket stark 7 Mycket stark 

8 8 

9 9 

10 Extremt stark ( nästan maximal) 10 Extremt stark ( nästan maximal) 

+ Maximal + Maximal 

Hur stora problem har du? 

Om foten / fötterna fungerar utan problem Om foten / fötterna fungerar utan problem 

betyder det 100 poäng. Hur många poäng betyder det 100 poäng. Hur många poäng 

får din fot / dina fötter på en skala från får din fot / dina fötter på en skala från 

0 till 100 för aktivitet 1............................... 0 till 100 för aktivitet 2............................... 

denna vecka ? denna vecka ? 

...................................poäng ...................................poäng 

29

Namn: Testnummer: 

Hur fysiskt aktiv är du? (v 0 ) 

Bilaga 3 

Vilken person stämmer bäst in på dig? 

Person A: Rör sig ganska lite. 

Person B: Rör sig en hel del men aldrig så att han/hon blir andfådd och 

svettig. 

Person C: Rör sig en hel del och blir svettig och andfådd någon gång ibland. 

Person D: Rör sig så att han/hon blir svettig och andfådd flera gånger i 

veckan. 

Person E: Rör sig så att han/hon blir svettig och andfådd varje dag eller 

nästan varje dag. 

Hur länge har besvären funnits? 

Hur lång tid har du haft ont i hälen/hälarna i sådan omfattning att det begränsat 

dig fysiskt? 

I ...................veckor 

30 

35

Bilaga 4 

Scand J Med Sci Sports 2011: 21: 819–823 

doi: 10.1111/j.1600-0838.2010.01051.x 

Sever’s injury: treatment with insoles provides effective pain relief 

S. Perhamre 1 , S. Janson 2,3 , R. Norlin 4 , M. Kla¨ ssbo 5 

1 Centre of Sports Medicine in Va¨rmland, Va¨rmland County Council, Karlstad, Sweden, 2 Department of Paediatrics, Va¨rmland 

County Council, Karlstad, Sweden, 3 Department of Public Health, Karlstad University, Karlstad, Sweden, 4 Department of 

Orthopaedics, University Hospital, O¨rebro University, O¨rebro, Sweden, 5 Centre for Clinical Research, Va¨rmland County Council, 

Karlstad, Sweden 

Corresponding author: Stefan Perhamre, Idrottsha¨lsan i Va¨rmland, Kolvgatan 1 B, S-653 41 Karlstad, Sweden. Tel: 146 54 15 34 12, 

Fax: 146 54 15 58 98, E-mail: stefan.perhamre@liv.se 

Accepted for publication 7 October 2009 

Sever’s injury (apophysitis calcanei) is considered to be the 

dominant cause of heel pain among children between 8 and 

15 years. The traditional advice is to reduce and modify the 

level of physical activity. Recommended treatment in general 

is the same as for adults with Achilles tendon pain. The 

purpose of the study was to find out if insoles, of two 

different types, were effective in relieving heel pain in a 

group of boys (n 5 38) attending a Sports Medicine Clinic 

for heel pain diagnosed as Sever’s injury. The type of insole 

was randomized, and self-assessed pain during physical 

activity in the treatment phase with insoles was compared 

Sever’s injury (apophysitis calcanei, Sever’s disease) 

is considered to be the dominant cause of heel pain 

among growing children between 8 and 15 years of 

age, but epidemiologic data on its incidence are still 

lacking (McKenzie et al., 1981; Orava & Virtanen, 

1982; Staheli, 1998; Ogden et al., 2004; Hendrix, 

2005). The complaint is benign, and disappears without 

exception after puberty when the apophyseal 

growth plate is closed. The natural course is poorly 

studied, but the condition appears to settle within 6– 

12 months. Occasionally, symptoms will persist for 2 

years (Cyriax, 1982; Brukner & Khan, 2007). 

The condition includes a spectrum of pain from 

recurrent light pain to strong pain. Boys constitute 

two-thirds of the patients, probably due to different 

physical activity preferences compared with girls of 

the same age (Lutter, 1992). The physical activity 

that appears to produce the highest levels of pain 

includes frequent running and jumping, and the sport 

that tends to dominate is soccer (Micheli et al., 1987; 

Madden & Mellion, 1996). Winter sports are only 

represented in low frequencies (Orava & Virtanen, 

1982; Micheli et al., 1987). The average age at onset is 

11–12 years (Micheli et al., 1987; Lutter, 1992). 

Modern literature describes Sever’s injury in terms 

of mechanical overuse with repetitive microtrauma 

36 

& 2010 John Wiley & Sons A/S 

with pain in the corresponding pre- and post-treatment 

phases without insoles. There were no other treatments 

added and the recommendations were to stay on the same 

activity level. All patients maintained their high level of 

physical activity throughout the study period. Significant 

pain reduction during physical activity when using insoles 

was found. Application of two different types of insoles 

without any immobilization, other treatment, or modification 

of sport activities results in significant pain relief in boys 

with Sever’s injury. 

to the calcaneal apophysis and its growth plate. Some 

authors relate the injury to forces from the Achilles 

tendon and the calf muscle complex, causing shear 

stress that compromises the apophysis (Adams & 

Hamblen, 1990; Volpon & de Carvalho Filho, 2002; 

Hendrix, 2005). Others suggest the repetitive impact 

force at heel strike as an alternative or complementary 

explanation (McKenzie et al., 1981; Orava & 

Virtanen, 1982; Ogden et al., 2004). 

The clinical findings are tenderness over the calcaneal 

tubercle and a positive calcaneus compression test 

(Micheli et al., 1987; Adams & Hamblen, 1990; Madden 

& Mellion, 1996; Kasser, 2006). In 40% of the 

cases, only one heel is affected (Micheli et al., 1987). 

The traditional advice is to diminish or eliminate 

provoking activities for a period of time. This is 

similar to the recommendations for treating Mb 

Osgood Schlatter (Hendrix, 2005). In addition, patients 

are recommended treatment similar to that 

for adults with Achilles tendon pain, most frequently 

including stretching, strengthening exercises for 

calf and extensor muscles and also NSAID treatment 

(Micheli et al., 1987; Lutter, 1992; Ishikawa, 

2005; Kasser, 2006). The importance of correcting 

malalignment, dominated by pronation, is stressed 

by some authors (Szames et al., 1990; Madden & 

819

Perhamre et al. 

Mellion, 1996). Insoles of a molded plastizote orthosis, 

a heel cup to align pronation, or a heel wedge to 

reduce shear stress to the apophysis, are included in 

most recommendations (Micheli et al., 1987; Lutter, 

1992; Peck, 1995; Staheli, 1998; Kasser, 2006). Nonweightbearing 

with the lower limb immobilized in a 

short leg cast or a removable ankle orthosis worn 

during daytime is sometimes recommended (Lutter, 

1992; Ogden et al., 2004; Kasser, 2006). All studies 

are, without exception, observational studies. 

The purpose of this study was to find out if insoles, 

of two different types, without any other treatments 

or recommendations were effective in relieving heel 

pain in a group of 35 boys, 9–15 years of age, 

attending a Sports Medicine Clinic. 

Materials and methods 

The design in this experimental study was an intervention phase 

(4 weeks) with two randomized insoles, heel cup or heel wedge, 

and two non-treatment phases without insoles, one before and 

one after treatment, of 2 weeks, respectively (Fig. 1). Thirty-five 

boys were consecutively included with a median age of 12 

Intervention group, n=33 

Inclusion 

Nonintervention group, n=5 

Inclusion 

2 w 

4 w 

Heel cup 

Heel wedge 

2 w 

Randomisation End of 

study 

7-8 w 

Fig. 1. An experimental study in a group of boys with heel 

pain (n 5 33), with randomized types of insoles (heel wedge 

n 5 18, heel cup n 5 17). The cohort was followed for 2 

weeks in a pre-treatment phase, for 4 weeks in an intervention 

phase with insoles and for 2 weeks in a post-treatment 

phase, with no insoles. A separate consecutive series of 

patients in a ‘‘natural course group’’ was followed in 7–8 

weeks without insoles (n 5 5). 

820 

years (range 9–14 years). The patients were examined at the 

Sports Medicine Clinic for heel pain and were recruited during 

10 months. All the examinations were performed by the first 

author. 

Inclusion criteria were male gender, age 9–15 years, and a 

diagnosis of Sever’s injury. Findings confirming the diagnosis 

were tenderness over the calcaneal tuberosity and a positive 

compression test. Additional negative test results from the 

Achilles tendon and other parts of the hind foot were 

compulsory. Heel pain had to be present for more than 2 

weeks, but o26 weeks when examined. All participants were 

high-level athletes. Each patient had to specify two physical 

activities giving heel pain (here, named activity A and activity 

B), where the most painful activity was A. These two activities 

must be performed at least three times a week during the study 

period. Soccer was the dominating sports activity (n 5 28) in 

which the boys registered their pain, followed by running 

(n 5 22). In all cases where soccer was one of the choices, it 

was reported as the most painful activity. The median duration 

of pain on examination was 10 weeks (range 4–20 weeks). 

Eighteen of the boys were diagnosed with unilateral Sever’s 

injury, and 17 with bilateral. No previous treatment was given. 

Three boys fulfilling the inclusion criteria were not willing 

to participate in the study. Two boys dropped out of the study, 

one of them because he missed filling in the questionnaire, and 

the other because of a calf muscle strain (n 5 38). 

Borg’s CR-10 scale was used for measuring pain during 

activity A and B (Neely et al., 1992). Borg’s CR-10 scale is a 

self-assessment visual analog scale (VAS) with anchors of 

verbal explanations added to the numbers of the pain scale. 

Zero represents absence of pain and 10 represent maximal 

pain. 

Activity level was measured with Engstrom’s activity index 

(Engstrom, 2004) with five levels of activity, A–E. When 

included in the study, eight boys reported the second highest 

level D (strained to shortness of breath and sweating many 

times a week) and 27 boys reported the highest level of 

activity, E, (strained to shortness of breath and sweating every 

day or most days of the week) (Fig. 4). 

To study the natural course, we added a second inclusion 

period of 2 months with a consecutive series of patients with 

Sever’s injury, not using insoles, with the same inclusion and 

exclusion criteria (Fig. 1). The patients were examined at the 

Clinic, by the same person, in the same way as those in the 

cohort. Self-assessed pain recordings and activity levels were 

collected twice, at the start and at the end of the period. This 

group included five boys (median age 12 years, range 10–13 

years). One boy fulfilling the inclusion criteria was not willing 

to participate. Four of the boys were diagnosed with unilateral, 

and one with bilateral pain. The median time with 

symptoms was 14 weeks (range 6–26 weeks). 

In the experimental study, two alternative insoles, a heel 

wedge and heel cup, were used (Fig. 2). The heel wedge was a 

5-mm-cork wedge covered with a thin elastic surface. The 

purpose of the heel wedge was to lift the heel in the loading 

phase and thereby limit the shear stresses of the apophysis. 

The heel cup was a rigid plastic cup applied directly on the 

bare heel, with a moulded long arch support. It encloses the 

heel with a 2–3 cm brim. The purpose of the heel cup was to fix 

the heel pad directly below the calcaneal tuberosity during heel 

strike. Bilateral insoles were always used. 

After inclusion in the study, the boys were randomized by 

picking one ticket from a concealed box. Randomization 

resulted in a heel wedge group (n 5 18) and a heel cup group 

(n 5 17). During 2 weeks of pre-treatment, the pain questionnaire 

for pain related to the chosen sport activity was filled in 

six times for the two chosen activities A and B. No insoles 

were used. In the treatment phase of 4 weeks, when the insoles 

37

Fig. 2. The heel cup, the foot (in the study always supported 

by a sports shoe) and the heel wedge. 

were used, the questionnaire was filled in eight times per 

person for each activity. A post-treatment phase of 2 weeks 

finally followed, without any insoles, when again the pain 

questionnaire for pain related to the chosen sport activity was 

filled in six times each for the two chosen activities A and B 

(Fig. 1). Activity levels were assessed at the start of the study 

and in the two phase shifts, three times in total (Fig. 4). 

Measurement for the non-treatment group (n 5 5) was 

activity level, median and range for the self-assessed pain at 

the start and end. 

The study was approved by the Regional Ethics Board of 

Uppsala (dnr 2004 M-377). 

Statistics 

Median pain for each boy was calculated for each of the three 

phases, pre-treatment, treatment and post-treatment, separately 

for activities A and B, and termed the pain variable. 

Median pain for the whole study group (n 5 33) was thereafter 

calculated from these pain variables, for each phase and 

activity. Interquartile range (IQR) (i.e. the numerical difference 

between the 25th and 75th centiles) was also calculated. 

To analyze the change over time between the different 

phases, for activity A and activity B separately, Wilcoxon’s 

signed-rank test was used. The pain variables for the group for 

activity A and activity B in the pre-treatment phase were 

compared with the pain variables in the treatment phase. The 

pain variables in the treatment phase were also compared with 

the pain variables in the post-treatment phase, and finally the 

pain variables in the pre-treatment phase were compared with 

the variables in the post-treatment phase (Fig. 3). 

To correct for mass significance, Holm’s sequential Bonferroni’ssss 

test was used. 

Results 

There was a significant reduction in pain level during 

activity with insoles, when comparing pain levels in 

the pre-treatment and treatment phase, for both 

activities A and B (Po0.001) (Fig. 3). Median pain 

level in the pre-treatment phase was 4.5 in activity A 

(IQR 2.5) and 4.0 in activity B (IQR 2.5). During 

treatment, pain was reduced to 2.0 (IQR 3.50) in 

both activities A and B. After discontinuation of the 

38 

Median pain values (VAS) 

10.00 

9.00 

8.00 

7.00 

6.00 

5.00 

4.00 

3.00 

2.00 

1.00 

Sever’s injury: insole treatment 

p


Number of patients 

25 

20 

15 

10 

5 

Activity level 

D 

E 

Pre-treatment Treatment 

Phase 

Fig. 4. Level of activity measured with Engstrom’s activity 

index (n 5 33) in the pre-treatment and treatment phase. One 

person changed from level D (strained to shortness of breath 

and sweating many times a week) to E (strained to shortness 

of breath and sweating every day or most days of the week) 

during treatment, and the others were on the same activity 

level in both phases. 

beneficial in the CNS forever (Hanrahan & Carlson, 

2006). This period is also crucial to build up the 

skeleton (Schwinghandl et al., 1999; Linde´n et al., 

2006, 2007). Sport activities in general, before and 

during puberty, are also very important from mental 

aspects. The joy of being an athlete is of great value 

(Engstrom, 1999). To lose 1 year or more because of 

heel pain can be devastating for participation in 

sports later on. 

Ogden et al. (2004) challenges the view on this 

disease as a concept of apophysitis. Instead, he found 

a traumatic stress fracture of the immature calcaneal 

trabecular bone, visualized with MRI. Ogden therefore 

singled out the repeated and intense impact forces 

as being responsible for the damage of the trabecular 

bone. Our findings with excellent pain relief during the 

use of insoles support this concept. The wedge lifting 

the heel 5 mm probably reduces shear forces on the 

apophysis and the cup probably reduces heel strike 

impact forces on the calcaneus due to its effect of 

fixing the heel pad under the calcaneal tuberosity. 

The results in this study are valid for several 

reasons. It is a consecutive study, which includes 

more than 90% of all boys seeking help for Sever’s 

injury at our department during the study period. 

The study design focuses on a single treatment and 

uses an internal control with a pre- and post-treatment 

phase. None of the boys used any other type of 

treatment during the study period, neither drugs, nor 

physiotherapy, or other orthoses. Pain was possible 

to quantify from time to time during defined, often 

822 

performed activities, which excluded recall bias. The 

self-assessed questionnaire was filled in at home, with 

no influence from the investigators. 

There were two dropouts during the study period 

because of two specific reasons, with no relation to 

the study design. Because of this, we have analyzed 

those who completed the study, instead of using the 

intention-to-treat analysis. 

From pre-treatment to treatment phase, there was 

significant pain relief in both activities. After removing 

the insole, pain returned. This proves that pain 

reduction is an effect of the treatment and not due to 

spontaneous healing, as part of the natural course. 

The results in the non-treatment group also indicate 

this. The natural course in Sever’s injury, among 

physically active boys in general, lasts for longer than 

8 weeks (Cyriax, 1982; Brukner & Khan, 2007). 

Recurrence rate is estimated to be 20% (Kvist et 

al., 1984). Our non-treatment group is small and is 

not a strict control group because the boys were 

included separately during a stipulated period. On a 

group level, they had some characteristics that were 

different compared with the studied cohorts. The pain 

at start was worse than in the treatment group (VAS 7 

vs VAS 4–4.5), the duration of pain was longer (14 vs 

10 weeks) and there were unilateral symptoms in 4/5 

of the cases, compared with less than half in the 

cohort groups, where bilateral pain dominated. 

It might be considered a weakness that the study 

has no ‘‘proper’’ control group. From an ethical 

point of view, we found it questionable to leave a 

group without any treatment, as a third study group, 

and have them continue at the same high activity 

level, with considerable pain. It had also been more 

troublesome to recruit study patients with this design. 

Another weakness in the study is that we do not 

compare the benefit of the two insole alternatives. 

The number of patients was too small to obtain 

enough statistical power to answer this question. 

Treatments suggested previously in Sever’s injury 

often include a reduction of activity level and physiotherapy 

(Micheli et al., 1987; Lutter, 1992; Madden 

& Mellion, 1996; Ishikawa, 2005; Kasser, 2006). 

Limitation in activity is the dominating recommendation. 

However, the scientific support for these 

treatment regimens is lacking. Our study is designed 

as a single treatment study without any other therapies 

included. In literature, in general (except in those 

studies that treat patients with a cast for a stipulated 

time), there is a number of treatments offered as a 

mix. This study has therefore no opportunity to 

discuss whether, for instance, physical therapy provides 

additional benefit in Sever’s injury or not. 

The present study shows that two insoles provided 

significant pain relief in highly active boys with 

Sever’s injury. This treatment is effective without 

reducing their level of activity. Further studies are 

39

equired to determine the etiology of the pain and 

which insole is the better one. 

Perspectives 

Treating Sever’s injury with insoles makes it possible 

for the boys studied, 9–15 years old, to be active in 

sports at a high level, instead of being passive, 

waiting for the symptoms to fade away. Sever’s 

injury is common in this age, and for many reasons 

it is important to develop adequate treatments. 

References 

Adams JC, Hamblen DL. The leg, ankle 

and foot. Outline of orthopaedics, 11th 

edn. Edinburgh: Churchill Livingstone, 

1990, 433pp. 

Brukner P, Khan K. Clinical sports 

medicine, 3rd edn. North Ryde NWT: 

McGraw-Hill, 2007, xxxix, 737pp. 

Cyriax J. The foot. In: Cyriax J, ed. 

Textbook of orthopaedic medicine vol 

1, diagnosis of soft tissue lesions, 8th 

edn. London: Baillie´ re Tindall, 1982: 

435pp. 

Engstrom L-M. Sports as a social marker 

[in Swedish]. Stockholm: HLS, 1999, 

143pp. 

Engstrom L-M. School – sports – health: 

studies of the subject sports and health, 

and young people’s physical activity, 

physical capacity and health condition: 

basis, purposes and methods [in 

Swedish]. Stockholm: 

Idrottshogskolan, 2004. 

Falk B, Tenenbaum G. The effectiveness 

of resistance training in children. A 

meta-analysis. Sports Med 1996: 22: 

176–186. 

Hanrahan SJ, Carlson TB. Game skills: a 

fun approach to learning sport skills. 

Champaign 2006: 175. 

Hendrix CL. Calcaneal apophysitis (Sever 

disease). Clin Podiatr Med Surg 2005: 

22: 55–62. 

Ishikawa SN. Conditions of the calcaneus 

in skeletally immature patients. Foot 

Ankle Clin 2005: 10: 503–513. 

Kasser JR. The foot – acquired 

conditions. In: Weinstein SL, Lovell 

WW, Winter RB, Morrissy RT, eds. 

Lovell and winter’s pediatric 

orthopaedics 2, 6th edn. Philadelphia: 

40 

Lippincott Williams & Wilkins, 2006: 


Kvist M, Kujala U, Heinonen O, Kolu T. 

Osgood Schlatter and Sever’s disease in 

young athletes. Duodecim. 1984: 100: 

142–150. [in Finnish]. 

Linde´ n C, Ahlborg H, Besjakov J, 

Gardsell P, Karlsson M. A school 

curriculum-based exercise program 

increases bone mineral accrual and 

bone size in prepubertal girls: two-year 

prospective data from the Pediatric 

Osteoporosis Prevention (POP) study. 

J Bone Miner Res 2006: 21: 829–835. 

Linde´ n C, Alwis G, Ahlborg H, Gardsell 

P, Valdimarsson Ö, Stenevi-Lundgren 

S, Besjakov J, Karlsson M. Exercise, 

bone mass and bone size in prepubertal 

boys: one-year data from the pediatric 

osteoporosis prevention study. Scand J 

Med Sci Sports 2007: 17: 340–347. 

Lutter LD. Sports related injuries. In: 

Drennan JC, ed. The child’s foot and 

ankle. New York: Raven Press, 1992: 


Madden CC, Mellion MB. Sever’s disease 

and other causes of heel pain in 

adolescents. Am Fam Physician 1996: 

54: 1995–2000. 

McKenzie DC, Taunton JE, Clement DB, 

Smart GW, McNicol KL. Calcaneal 

epiphysitis in adolescent athletes. Can J 

Appl Sport Sci 1981: 6: 123–125. 

Micheli LJ, Ireland ML, Micheli LJ, 

Ireland ML. Prevention and 

management of calcaneal apophysitis 

in children: an overuse syndrome. J 

Pediatr Orthop 1987: 7: 34–38. 

Neely G, Ljunggren G, Sylén C, Borg G. 

Comparison between the Visual 

Sever’s injury: insole treatment 

Key words: calcaneus, sever’s injury, apophysitis calcanei, 

insoles, heel cup, heel wedge. 

Acknowledgements 

Thanks to Mikael Hasselgren Ph.D. from the Primary Care 

Research Unit and statistician Katarina Wiberg Hedman and 

Fredrik Lundin, Centre for Clinical Research, Va¨ rmland 

County Council, for support. Financial support was provided 

by the Primary Care Research Unit and by the foundation for 

Orthopaedic Devices, Va¨ rmland County Council. 

Analogue Scale (VAS) and the 

Category Ratio Scale (CR-10) for 

evaluation of leg exertion. Int J Sports 

Med 1992: 13: 133–136. 

Ogden JA, Ganey TM, Hill JD, Jaakkola 

JI. Sever’s injury: a stress fracture of 

the immature calcaneal metaphysis. J 


Orava S, Virtanen K. Osteochondroses in 

athletes. Br J Sports Med 1982: 16: 


Ozmun JC, Mikesky AE, Surburg PR. 

Neuromuscular adaptations following 

prepubescent strength training. Med 

Sci Sports Exerc 1994: 26: 510–514. 

Peck DM. Apophyseal injuries in the 

young athlete. Am Fam Physician 

1995: 51: 1891–1895. 

Schwinghandl JK, Sudi B, Eibl B, 

Wallner S, Borkenstein M. Effects of 

an individualised training program 

during weight reduction on body 

composition. A randomised trial. Arch 

Dis Child 1999: 81: 426–428. 

Staheli LT. Sports/foot and ankle. 

Fundamentals of pediatric orthopedics, 

2nd edn. Philadelphia: Lippincott- 

Raven, 1998: 111–128. 

Szames SE, Forman WM, Oster J, Eleff 

JC, Woodward P. Sever’s disease and 

its relationship to equinus: a statistical 

analysis. Clin Podiatr Med Surg 1990: 


Volpon JB, de Carvalho Filho G. 

Calcaneal apophysitis: a quantitative 

radiographic evaluation of the 

secondary ossification center. Arch 

Orthop Trauma Surg 2002: 122: 338– 

341. 

823

Scand J Med Sci Sports 2011: 21: e42–e47 

doi: 10.1111/j.1600-0838.2010.01140.x 

Sever’s injury; treat it with a heel cup: a randomized, crossover study 

with two insole alternatives 

S. Perhamre 1{ , F. Lundin 2 , R. Norlin 3 , M. Kla¨ ssbo 2 

1 2 3 

Centre of Sports Medicine in Va¨rmland, Karlstad, Sweden, Centre for Clinical Research, Karlstad, Sweden, Department of 

Orthopaedics, University Hospital, O¨rebro University, O¨rebro, Sweden 

Corresponding author: Rolf Norlin MD, Department of Orthopaedic Surgery, University Hospital, 70185 O¨rebro, Sweden. 

Tel: 146 19 60 25 366, Fax: 146 19 60 25 337 E-mail: rolf.norlin@orebroll.se 

{ 

Deceased. 

Accepted for publication 24 March 2010 


dominant cause of heel pain among children. Common 

advice is to reduce physical activity. However, our previous 

study showed that application of insoles reduced pain in 

Sever’s injury without having to reduce physical activity. 

The purpose of this study was to test which of the two 

insoles, the heel wedge or the heel cup, provided best pain 

relief during sport activity in boys with Sever’s injury 

(n 5 51). There was a crossover design in the first rando- 



among growing children between 8 and 15 years of 

age (Staheli, 1998; Ogden et al., 2004; Hendrix, 

2005). The heel complaint is benign, but includes a 

wide spectrum of pain intensity, and duration. It 

disappears without exception after puberty (Cyriax, 

1982; Brukner & Khan, 2007). 

Boys constitute two-thirds of the patients, probably 

due to different physical activity preferences 

compared with girls at the same age (Lutter, 1992). 

The physical activity that produces the highest levels 

of pain includes frequent running and jumping, and 

the sport that tends to dominate is soccer (Madden & 

Mellion, 1996). Outdoor winter sports are represented 

at a low frequency (Orava & Virtanen, 1982; 

Micheli & Ireland, 1987). The average age at onset is 

11–12 years (Lutter, 1992). In 35–40% of the cases, 

only one heel is affected (Micheli & Ireland, 1987; 

Perhamre et al., 2009) 

Sever’s injury is in modern literature described as 

induced by mechanical overuse with repetitive microtrauma 

to the calcaneal apophysis and its growth 

plate – the physis. The injury is reported to be related 

to forces from the Achilles tendon and calf complex 

due to shear stresses that compromise the apophysis 

(Peck, 1995; Volpon & de Carvalho Filho, 2002). 

Some authors suggest the repetitive impact forces at 

e42 

Bilaga 5 


mized part of the study. In the second part, the boys, 9–14 

years, chose which insole they preferred. There was a 

reduction in odds score for pain to a fifth (a reduction of 

80%) for the cup compared with the wedge (Po0.001). 

When an active choice was made, the heel cup was preferred 

by 475% of the boys. All boys maintained their high level 

of physical activity throughout. At 1-year follow-up, 22 boys 

still used an insole and 19 of them reported its effect on pain 

as excellent or good (n 5 41). 

heel strike as an alternative or a complementary 

explanation (McKenzie et al., 1981; Orava & Virtanen, 

1982), including a microtraumatically induced 

calcaneal stress fracture of trabecular bone in refractory 

cases (Ogden et al., 2004). 

The clinical findings include tenderness over the 

lower part of the tubercle of calcanei and a positive 

calcaneal compression test, the squeeze test (Adams 

& Hamblen, 1990; Madden & Mellion, 1996; Kasser, 

2006). 

The traditional advice is to diminish or eliminate 

provoking activities for a prescribed period of time. 

This is similar to the recommendations for Mb 

Osgood Schlatter (Hendrix, 2005). The proposed 

treatment consists of a traditional concept for adults 

with Achilles tendon pain, most frequently including 

stretching, strengthening exercises for calf and extensor 

muscles, ice and NSAID treatment (Ishikawa, 

2005; Kasser, 2006). The importance of correcting 

malalignment, dominated by pronation, is stressed 

(Madden & Mellion, 1996) as well as the relation 

between pain and a tight heel cord (Szames et al., 

1990). Insoles, either a molded plastizote orthosis to 

align pronation, or a heel wedge to directly reduce 

shear stresses to the apophysis, are the major insole 

proposals (Staheli, 1998; Kasser, 2006). Non-weightbearing 

with the lower limb immobilized in a short 

leg cast or a removable ankle orthosis worn during 

41

daytime is sometimes recommended (Ogden et al., 

2004; Kasser, 2006). Our previous study showed that 

two specific types of insoles, with different approaches 

to microtraumatic stress, were effective in 

relieving pain among physically active boys (Perhamre 

et al., 2009). By lifting the heel 5 mm with 

the wedge, the tension forces are expected to be 

reduced in mid stance and in push off. The heel cup 

is thought to act as an internal tight box in the shoe, 

fixating the heel pad in order to reduce the direct 

impact force in heel strike. 

The purpose of this study was to test which of the 

two insoles, the heel wedge or the heel cup provided 

best pain relief in boys with Sever’s injury, without 

reducing their level of physical activity. 


The study was a prospective crossover designed, with alternation 

between baseline phases and intervention phases with the 

two different insole alternatives (Fig. 1). The first randomized 

part of the study period included 12 weeks. In the second nonrandomized 

part of the study (14 weeks), each boy chose one 

treatment alternative (Fig. 2). 

Inclusion criteria were male between 9 and 15 years with a 

history of Sever’s injury, and heel pain present for 42 weeks, 

but o26 weeks when examined. Findings confirming the 

diagnosis were tenderness over the calcaneal tubercle, and a 

positive compression test (Kasser, 2006). Additional negative 

test results from the Achilles tendon were compulsory. The 

level of physical activity must be high, at least the second 

highest level of sport activity (D 5 strained to shortness of 

breath and sweating many times a week or E 5 strained to 

Fig. 1. The heel wedge used in the study, lifting the heel 

5 mm and the heel cup (of 3 mm. thickness) fixating the heel 

pad. The cup offers no wedge effect, due to equal lift of the 

rear and fore foot. 

42 

shortness of breath and sweating every day or most days of the 

week) according to Engstro¨ m’s activity index (Engstro¨ m, 

2004). 

Two alternative insoles, the heel wedge and the heel cup, 

were used (Fig. 1). Bilateral treatment was always performed. 

The heel wedge was a 5 mm cork wedge covered with a thin 

elastic surface. It lifted the heel 5 mm in mid stance and push 

off. The heel cup was a rigid thermoplastic cup (of 3 mm height 

but with no wedge effect) molded directly on the bare heel with 

a three-fourth length arch support. It enclosed the heel with a 

2–3 cm high brim in order to stop the heel pad from excessive 

deformation. The high brim made it different from earlier 

described molded heel insoles with their purpose to align the 

foot in running. There are some similar rigid alternatives, like 

the prefabricated Whitman–Roberts and the semi-molded 

Lange insoles in hard plaster. 

All boys were examined by the first author at the Sports 

Medicine Clinic for heel pain and recruited during February– 

June during 3 following years, totally 15 months. A total of 51 

boys were invited of which four did not accept. During the 

study, three boys dropped out, one due to social complications, 

one due to difficulties filling in the questionnaire and one 

due to a calf muscle strain injury, leaving 44 boys. The 

analyses below were performed on data supplied by these 44 

boys who completed the study. Nineteen (43%) of the boys 

were diagnosed with unilateral Sever’s injury (10 right foot, 

nine left foot) and 25 with bilateral. 

After inclusion, the boys were randomized into two groups 

with tickets concealed in a box. No stratification was made. 

Thereafter, the pain questionnaire was filled in for the chosen 

activities during the baseline phase (weeks 1–2) preceding the 

interventions, including six registrations for each of the two 

activities. During the first intervention phase (weeks 3–6), the 

boys were using either heel cup (n 5 20, group I) or heel wedge 

(n 5 24, group II) including eight pain registrations for each of 

the two activities A and B. A second baseline phase (wash-out) 

followed (weeks 7–8) without any treatment. During the 

second intervention phase (weeks 9–12), there was a shift 

between insoles (crossover). In the second and final part of the 

study (after 12 weeks), the boys started with another baseline 

phase (wash-out) of 2 weeks without insoles followed by a last 

intervention period of 12 weeks. Here every boy, instead of 

being randomized, chose a treatment alternative with respect 

to his overall preferences, and was asked to focus on pain 

relief. He could also stay active without the insole. The 

intervention phases were concluded after 26 weeks (Fig. 2). 

There was a follow-up questionnaire added 1 year later 

concerning pain relief during the study period. The continuing 

need of insoles and pain relief with their used insoles the 

Following 

Randomisation Cross-over 

Choice 

choice: 

Group 1 

n=20 

Cup Cup 

19 

Cup 

n=34 

Group 2 

n=24 

Wedge Wedge Wedge 

5 

Time line: 0w 2w 6w 8w 12w 14w 

Period: B1 I1 B2 I2 B3 

Sever’s injury; treat it with heel cup 

Fig. 2. Study design with alternating periods of baseline ( 5 no treatment, B1, B2 and B3) and intervention phases 

( 5 treatment with insoles, I1, I2 and I3). The first part, including intervention phases (I1 and I2) were randomized with a 

crossover design, and the second part (I3) represented each boy’s personal choice (n 5 44). 

I3 

26w 

n=10 

e43


following year were also asked about. Three boys did not 

answer, leaving n 5 41. 

Pain was assessed with Borg’s CR-10 scale (Neely et al., 

1992). It is a VAS scale with anchors of verbal explanations 

added to the numbers in the pain scale, and was used for 

measuring pain associated with the two sports activities 

chosen by each boy. Zero was absence of pain and 10 was 

maximal pain. These self-assessed pain recordings were 

made in direct relation to training and sport events, and 

were answered with no influence from the investigators. 

Activity level was measured with Engstro¨ m’s activity index 

(Engstro¨ m, 2004), which includes five levels of physical 

activity, A–E and assessed at start, at phase shifts and at the 

end of week 26. 

When soccer was one of the chosen activities (n 5 33), it was 

always chosen as the first activity A, producing the highest 

levels of heel pain. Running was the dominating activity 

chosen as activity B (n 5 29). A majority of the boys (n 5 26) 

had soccer as activity A, and running as activity B. 

The study has been approved by the Regional Ethics Board 

of Uppsala, Sweden (dnr 2004 M-377). 

Neither the Public Sports Medicine Clinic nor the authors 

have any relation to, profit or other benefits from the 

companies that manufacture the products in this study. 

Table 1. Baseline characteristics for the primary heel wedge group and 

the primary heel cup group at study start 

Primary heel Primary heel 

wedge cohort cup cohort 

Patients, numbers 24 20 

Median age, years (range) 12 (10–14) 12 (9–14) P 5 0.844* 

Median pain level (IQR) 4 (2.75) 5 (3.75) P 5 0.149* 

Pain history, weeks (range) 8 (4–25) 12 (4–24) P 5 0.102* 

*P-values computed using the two-tailed asymptotic Mann–Whitney 

U-test. 

IQR, interquartile range. 

Table 2. The results of the first randomized 12 weeks of the study, using 

a mixed ordinal logistic regression model to accommodate the longitudinal 

design of the study (n 5 44) 

Variable Odds ratio P-value 95% confidence interval 

Activity A 

Time 0.93 0.001 [0.89, 0.97] 

Wedge 1 – – 

No treatment 2.32 o0.001 [1.69, 3.19] 

Cup 0.22 o0.001 [0.15, 0.34] 

Initial pain 1.75 o0.001 [1.37, 2.23] 

Activity B 

Time 0.93 0.002 [0.89, 0.97] 

Wedge 1 – – 

No treatment 2.29 o0.001 [1.67, 3.16] 

Cup 0.18 o0.001 [0.12, 0.27] 

Initial pain 1.81 o0.001 [1.51, 2.17] 

The wedge was chosen as base category (OR 1) for the treatment 

variable, because our primary target was the comparison between the 

wedge and the cup. The hypothesis was that the heel cup provided 

significant better pain relief than the heel wedge. This regression analysis 

included all measured data, and was corrected for time. 

e44 

Pain value (median), IQR 

10.00 

Borg CR-10 

7.50 

5.00 

2.50 

0.00 

Cup 

Wedge 

Wedge 

B1 I1 B2 

Phase 

I2 B3 

Fig. 3. Separated cup and wedge group showing median 

pain values and interquartile range (IQR) for the randomized 

first part of the study and the corresponding two 

baseline phases. Group I (white boxes) started with heel cup 

(with crossover to heel wedge in the second part 9 group II 

(gray boxes) started with heel wedge. The statistics includes 

both activity A and B. 

Statistics 

In an ordinal mixed regression, the estimated pain was 

modeled using initial pain, time and treatment as explanatory 

variables. The effect of each explanatory variable was an odds 

ratio (OR) representing the multiplicative effect of a change in 

this variable. We categorized the self-reported pain (VAS, 

score 0–10) into five levels (0–2, 3–4, 5–6, 7–8 and 9–10) and 

modeled the pain level at each occasion using initial pain and 

treatment as fixed effects, number of days since baseline 1–3 as 

a fixed and random effect and patient as a random intercept. 

Low odds scores translated into less pain and large odds into 

more pain. 

Modeling time since baseline 1–3 as both random and fixed 

effect takes into account a possible healing effect over time 

because this is a known property of the etiology of Sever’s 

injury. The regression analyses were based on baseline phase 1, 

intervention 1, baseline phase 2 and intervention 2 as these 

were the phases included in the randomized part of the study. 

The treatment variable had three levels where the wedge was 

chosen as the base category as our primary target was the 

comparison between the wedge and the cup. We analyzed the 

two activities A and B separately. All regression analyses were 

made using the statistical package gllamm (Rabe-Hesketh et 

al., 2002; Stata, 2007) with adaptive quadrature in STATA/IC 

10.0 (Stata Corp. LP, College Station, Texas, USA), (Table 2). 

As a descriptive supplement to these regression analyses, a 

diagram was added with pain medians and interquartile 

ranges (IQR) for each phase. The pain relief for the wedge 

and the cup was compared including both activities (Fig. 3). 

Results 

A total of 44 boys completed the study. There were 

no significant differences in age, median pain level 

Cup 

43

and pain history in weeks before the randomization 

between the two groups, starting with either heel 

wedges or heel cups (Table 1). All boys continued 

their high levels of physical activity during the study 

(Engstro¨ m’s activity index level E n 5 38, and D 

n 5 6). 

Pain was significantly lower in the heel cup group 

compared with the wedge group. For activity A and 

B, the effect of the cup was OR 5 0.22 (Po0.001 

[0.15, 0.34]) and OR 5 0.18 (Po0.001 [0.12, 0.27]), 

respectively, thus reducing the odds score for pain to 

a fifth compared with the wedge (ordinal data). There 

was also a significant effect of time for activity A 

OR 5 0.93 (P 5 0.001) and for activity B OR 5 0.93 

(P 5 0.002), which means that pain was reduced with 

time as an independent variable (Table 2). We did 

check duration of pain at baseline and age as 

explanatory variables, but none of them had any 

significant effect in the model. 

When able to freely choose in the final second part 

(intervention 3), 34 boys (77%) chose the heel cup 

and 10 (23%) chose the heel wedge (Fig. 2). No one 

preferred to stop using insoles. 

There was a continuous reduction of pain between 

the three intervention phases for both groups. After 

each intervention phase, the pain reoccurred in the 

next baseline phase (Fig. 3). During the prolonged 

phase with the free choice of insole (8 weeks compared 

with 4 weeks), the median pain was 0.5. After 

26 weeks when the study was concluded, the median 

value for pain was 0 (IQR 1). 

The 1-year follow-up questionnaire showed that 

pain relief during the study period was good or 

excellent in 40 answers (n 5 41) for the chosen insole, 

and in 14 (n 5 40) for the non-chosen alternative. 

During the following year, 14 were pain free throughout, 

12 still had pain when being active on their level 

and 14 had pain that disappeared in 3 months 

(median), where six boys had relapses thereafter 

(n 5 40). The number of boys that used insoles after 

1 year was 22. Nineteen of the boys assessed its effect 

on pain as excellent or good. Five boys found the 

insoles (all cups) inconvenient in the post-study 

phase. 

Discussion 

This study shows that treatment with heel cup was 

superior to treatment with heel wedge for relieving 

pain during sport activities in Sever’s injury without 

reducing physical activity, indicating that the reduction 

of repetitive impact forces in heel strike is the 

most important factor for pain relief. 

The result is valid for several reasons. Our randomized 

study was focused on a single treatment regime 

for heel pain: insole treatment (Figs 1 and 2). There 

44 


was no advertisement or active recruitment made and 

the boys were included in the same season during 3 

following years. There were two parallel groups with 

the same overall experiences during the randomized 

first half of the study with baseline phases and 

crossover treatment (Fig. 2). All pain recordings 

were self–assessed with no influence from the investigators. 

All boys continued being physically active at 

the same high level throughout the study period. The 

studied group appears to be typical for physically 

active boys, 9–15 years old, seeking Health Care for 

heel pain (Table 1). We put the upper inclusion limit 

in pain history before start to 26 weeks to avoid the 

most refractory cases, where 14% had a pain history 

between 4 and 6 months at start. No one had a pain 

history shorter than 4 weeks (mean 11 weeks). At the 

end of phase B2, 39 boys still had pain (n 5 44) (Fig. 

3). Less than 10% (n 5 3) of all the boys fulfilling the 

inclusion criteria at our clinic did not want to 

participate. Hence, missing cases were few and the 

reasons for dropping out were not related to the 

study design. 

We chose to include only boys with Sever’s injury, 

because boys totally have dominated in frequency, 

probably due to different gender preferences and 

training patterns (Lutter, 1992). 

Previous guidelines for treating Sever’s injury are 

built on principles for adults suffering from heel pain 

(Micheli & Ireland, 1987; Ishikawa, 2005; Kasser, 

2006). Scientific evidence for the effect of these 

treatment regimens are lacking. Insole treatment 

together with limitations in activity are the most 

frequent advice, and the wedge is the most frequent 

insole advice in the literature (Peck, 1995; Staheli, 

1998; Kasser, 2006). The approach is to limit the 

stress to the apophysis – Achilles tendon complex by 

placing the foot in slight equinus. It points out the 

apophysis and its growth plate, the physis, as the 

morphologic origin of pain. The second proposed 

alternative in the literature is to correct malalignment 

particularly pronation or flat-footed running, by a 

molded plastizote orthosis or a heel cup (Micheli & 

Ireland, 1987; Madden & Mellion, 1996). 

Ogden et al. (2004) questioned the concept of 

apophysitis. Instead, he found a traumatic stress 

fracture of the immature calcaneal trabecular bone, 

visualized with MRI. Ogden therefore singled out the 

repeated and intense impact forces as responsible for 

the damage of the trabecular bone close to the physis, 

the metaphysis and not to the apophysis. His arguments 

are supported by the fact that sports like ice 

hockey and cross country skiing, with very little 

impact forces compared with soccer and running, 

but with great shear stresses to the apophysis in pull 

off, are very uncommon as exacerbating sports for 

heel pain (Orava & Virtanen, 1982; Micheli & Ireland, 

1987). The thesis by Jo¨ rgensen (1989) (includ- 

e45


ing adults), found that a tight heel box in the sports 

shoe in order to minimize the deformation of the heel 

pad, gave less overuse injuries from foot, lower leg 

and knee compared with optimal external shockabsorbing 

material that was put under the heel 

without the box support. The rigid heel cup with its 

brim, produced unloaded, directly on the bare heel 

that was used in our study, appears to be optimal to 

minimize the deformation of the heel pad. This cup is 

different from the other insoles that are described as 

heel cups, where the ‘‘cups’’ are made like spoons 

without the brim, often in flexible material. 

The present study showed that the heel cup fixating 

the heel pad provided better pain relief in young boys 

with Sever’s injury compared with the traditional 

heel wedge. This indicates that the reduction of 

impact forces, rather than reduction of shear stresses, 

is crucial for optimal pain relief. We agree with the 

demand to change the terminology from Sever’s 

disease and apophysitis calcanei to Sever’s injury, 

as proposed by Ogden et al. (2004). This is supported 

by a histological analysis of the region that corroborates 

a primary compression, rather than a tension 

cytoarchitecture (Ogden, 2000). Cortical bone (80% 

of the total bone mass) is primarily found in the 

diaphyses of the long bones. Trabecular bone is 

mainly found in the metaphyseal regions, the vertebral 

bones and pelvic bones (Alfredson, 1997). In 

adulthood, 95% of calcaneus is built up with trabecular 

bone, the biggest part among all body bones 

(Apurva et al., 1997). 

There was an 80% reduction of the odds score for 

pain when using a heel cup compared with a wedge. 

The cup was also preferred by 475% in the second 

part of the study. It although appears obvious that 

both insoles offered pain relief, supported by the pain 

relapses when the insoles were removed in every 

following baseline phase for both activities (Table 

2, Fig. 3). After 26 weeks, when the intervention was 

concluded the pain level in general was very low 

(median 5 0), but at 1-year follow-up half of the boys 

still used insoles in their physical activities where half 

of the group used it for pain relief (n 5 12) and half of 

the group for prevention (n 5 10). The chosen insole 

still offered adequate pain relief, and appeared to 

protect the heel from repetitive traumatic stress, but 

did not in all cases appear to improve healing of the 

injury. The inconvenience with the cup (n 5 5) that 

first occurred after the study, can possibly be ex- 

References 

Adams JC, Hamblen DL, eds. The leg, 

ankle and foot. In: Outline of 

orthopaedics, 11th edn. Edinburgh: 

Churchill Livingstone, 1990: 359–402. 

e46 

Alfredson H. Exercise and bone 

in females and in athletes with 

chronic Achilles tendinosis. Umea˚ : 

Umea˚ University, 1997. 

plained by the fact that they can be grown out within 

a period of 18 months. The small group (n 5 10) who 

preferred the wedge had as good effect on pain with 

their choice, during the study period and at followup, 

as the cup group. A hypothesis is that there can 

be more than one morphologic origin in Sever’s 

injury. 

Another interesting finding was that 43% of the 

boys in the study had only unilateral pain, and were 

symptom free in the other foot. This big number of 

boys with unilateral pain is also earlier presented 

(Micheli & Ireland, 1987; Perhamre et al., 2009). This 

indicates that there may be internal unknown factors 

involved. 

Perspectives 

Heel pain is a common condition in an important 

period of development, between 8 and 15 years of 

age, and should be named Sever’s injury, not Sever’s 

disease. An individually made molded rigid heel cup 

with a fixating brim should primarily be offered in 

Sever’s injury, in order to protect the heel from a 

continuing overload with repetitive stress. If this does 

not work the next offer, with respect to the minor 

group in this study, can be a pair of 5 mm wedges. 

Sport activities in general, before and during puberty, 

are very important especially from physiological, 

neuromuscular and mental aspects. To lose a 

long period of time because of heel pain can be 

devastating for participation in sports later on, and 

is according to the results in this study, not necessary. 

Key words: calcaneus, Sever’s injury, apophysitis calcanei, 

insoles, heel cup, heel wedge. 


Thanks to Ingemar Hansson, who made the insoles, Mikael 

Hasselgren PhD medical chief Centre for Clinical Science, Eva 

Stjernstro¨ m RPT, division chief Va¨ rmland County Council 

and Staffan Janson professor, Department of Paediatrics, 

University of O¨ rebro, and professor Department of Public 

Health, Karlstad University, for support. Financial support 

was provided by the Primary Care Research Unit, by the 

division Health, Habilitation and Rehabilitation and by the 

foundation for Orthopaedic Devices, Va¨ rmland County 

Council. 

Apurva S, Vliet E, Lewiecki M, Maricic M, 

Abdelmalek A, Gluck O, Baylink D. 

Bone density measurement – a systematic 

review. JIM 1997: 241: 12–22. 

45


medicine, 3rd edn. North Ryde: 

McGraw-Hill, 2007, p. 737. 

Cyriax J. The foot. In: Cyriax J, ed. 

Textbook of orthopaedic medicine Vol 

1, Diagnosis of soft tissue lesions, 8th 

edn. London: Baillie´ re Tindall, 1982: 

435. 

Engstro¨ m L-M. School – sports – health: 

studies of the subject Sports and 

Health, and young people’s physical 

activity, physical capacity and health 

condition: basis, purposes and methods 

[in Swedish]. Stockholm: 

Idrottsho¨ gskolan, 2004: 142. 




Ishikawa SN. Conditions of the calcaneus 

in skeletally immature patients. Foot 

Ankle Clin 2005: 10: 503–513. 

Jo¨ rgensen U. Implication of heel strike: 

an anatomical, biomechanical, 

physiological and clinical study with 

focus on the heel pad. Link o¨ ping: 

Linko¨ ping University, 1989. 

Kasser JR. The Foot – Aquired 

conditions. In: Weinstein SL, Lovell 

WW, Winter RB, Morrissy RT, eds. 

Lovell and winter’s pediatric 

orthopaedics 2, 6th edn. Philadelphia: 

46 

Lippincott Williams & Wilkins, 2006: 


Lutter LD. Sports related injuries. In: 

Drennan JC, ed. The child’s foot and 

ankle. New York: Raven Press, 1992: 





54: 1995–2000. 

McKenzie DC, Taunton JE, Clement DB, 

Smart GW, McNicol KL. Calcaneal 

epiphysitis in adolescent athletes. Can 

J Appl Sport Sci 1981: 6: 123–125. 

Micheli LJ, Ireland ML. Prevention and 


in children: an overuse syndrome. 

J Pediatr Orthop 1987: 7: 34–38. 

Neely GL, Sylén G, Borg CG. 

Comparison between the Visual 

Analogue Scale (VAS) and the 

Category Ratio Scale (CR-10) for 

evaluation of leg exertion. Int J Sports 

Med 1992: 13: 133–136. 

Ogden JA. Foot. In: Ogden JA, ed. 

Skeletal injury in the child, 3rd edn. 

New York: Springer, 2000: 1091–1159. 



the immature calcaneal metaphysis. 

J Pediatr Orthop 2004: 24: 488–492. 


Orava S, Virtanen K. Osteochondroses in 

athletes. Br J Sports Med 1982: 16: 


Peck DM. Apophyseal injuries in the 

young athlete. Am Fam Physician 

1995: 51: 1891–1895. 

Perhamre S, Janson S, Norlin R, Klässbo 

M. Sever’s injury: treatment with 

insoles provides effective pain relief. 

Scand J Med Sci Sports 2009, in press. 

Rabe-Hesketh S, Skrondal A, Pickles A. 

Reliable estimation of generalized linear 

mixed models using adaptive quadrature. 

The Stata Journal 2002: 2: 1–21. 

Staheli LT, ed. Sports/Foot and ankle. In: 

Fundamentals of pediatric orthopedics, 

2nd edn. Philadelphia: Lippincott- 

Raven, 1998: 111–128. 

Stata P. STATA/IC release 10 user 

manual [XT] – longitudinal/panel data. 

College Station: Stata Press, 2007. 






Volpon JB, de Carvalho Filho G. Calcaneal 

apophysitis: a quantitative radiographic 

evaluation of the secondary ossification 

center. Arch Orthop Trauma Surg 

2002: 122: 338–341. 

e47

Scand J Med Sci Sports 2011 

doi: 10.1111/j.1600-0838.2010.01266.x 

A heel cup improves the function of the heel pad in Sever’s injury: 

effects on heel pad thickness, peak pressure and pain 

S. Perhamre 1 , F. Lundin 2 , M. Kla¨ ssbo 2 , R. Norlin 3 

1 2 

Centre of Sports Medicine in Va¨rmland, Va¨rmland County Council, Karlstad, Sweden, Centre for Clinical Research, Va¨rmland 

County Council, Karlstad, Sweden, 3 Department of Orthopaedics, University Hospital, O¨rebro University, O¨rebro, Sweden 

Corresponding author: Rolf Norlin, MD, Department of Orthopaedics, University Hospital, 70185 O¨rebro, Sweden. E-mail: 

rolf.norlin@orebroll.se 

Accepted for publication 9 November 2010 


dominant cause of heel pain among children between 8 and 

15 years. Treating Sever’s injury with insoles is often 

proposed as a part of a traditional mix of recommendations. 

Using a custom-molded rigid heel cup with a brim enclosing 

the heel pad resulted in effective pain relief without reducing 

the physical activity level in our previous two studies. The 

purpose of this study was to assess the effect of the heel 

cup on heel pad thickness and heel peak pressure (n 5 50). 

The difference in heel pad thickness and in heel peak 



among children between 8 and 15 years of age 

(Staheli, 1998; Ogden et al., 2004; Hendrix, 2005). 

The condition was described by Haglund 1907 and 

named by Sever (1912) as a disease, later described as 

an osteochondrosis. Focus has thereafter been concentrated 

on the apophysis. The dominating theory 

at present suggests that the cause of heel pain in 

Sever’s injury is a mechanical overuse due to repetitive 

impact pressure and shear stresses, most noticeable 

in activities including much running and 

jumping (Micheli & Ireland, 1987; Madden & Mellion, 

1996). Ogden et al.(2004) therefore proposed a 

change in terminology from disease to injury. 

The heel pad with its sum of cooperating compartments 

provides cushioning, and thereby reduces the 

peak force at the instance of heel strike. Weight 

bearing activities strongly stimulate its development 

and function. The heel pad develops from birth to 

adulthood and is composed of defined fat vesicles 

rich in elastic fibers and encapsulated with a thick 

fibrous outer coat, with a capability to resist the 

compressive and shear forces of ambulation. A 

healthy heel pad is highly resilient, returning about 

70% of the energy used for deformation in its elastic 

recoil. Atrophy due to inactivity results in fragmen- 

Bilaga 6 


pressure using a sports shoe without and with a heel cup 

was compared. With the heel cup the heel pad thickness 

improved significantly and the heel peak pressure was 

significantly reduced. These effects correlated with a significant 

reduction in pain when using the heel cup in a sports 

shoe, compared with using a sports shoe without the heel 

cup. A heel cup, providing an effective heel pad support in the 

sports shoe, improved the heel pad thickness and reduced 

heel peak pressure in Sever’s injury with corresponding pain 

relief. 

tation of the elastic components, affecting the cellular 

biomechanics (Ogden, 2000). About 60% of the 

weight-bearing load is carried in the rear foot in 

static standing. This pressure is approximately 30% 

of the peak pressure produced in walking and 16% of 

that in running (Cavanagh et al., 1987; Donatelli, 

1990). 

Treating Sever’s injury with insoles is often a part 

of a traditional mix of recommendations. Different 

etiological theories result in two different main insole 

strategies. The dominating advice in literature is to 

choose a heel wedge, in order to unload the Achilles 

tendon and thereby reduce shear stresses to the 

apophysis/physis (Staheli, 1998; Brukner & Khan, 

2007). The importance of correcting malalignment, 

dominated by pronation, and a short heel cord, is 

stressed by other authors as the other main strategy 

(Szames et al., 1990; Madden & Mellion, 1996). 

In a recent study by our group we compared 

treatment with insoles of two different types (a cup 

with a brim and a heel wedge of 5 mm), with pre- and 

posttreatment phases without insoles. Both insoles 

effectively relieved heel pain in Sever’s injury (Perhamre 

et al., 2010a). The pain returned swiftly when 

the insoles were removed. In a following study, we 

found that the heel cup was significantly better than 

the heel wedge in relieving pain (Perhamre et al., 

1 

47


2010b). More than 75% of the boys preferred the 

heel cup (Fig. 1). 

The purpose of this study was to analyze if thickness 

and function of the heel pad was affected by the 

use of a heel cup. Our two main research questions: 

Does the thickness of the heel pad differ between 

standing bare feet, in the sport shoe, and in the same 

sport shoe but with the heel cup? 

Is there a difference in local peak pressure in standing 

and running with vs without the heel cup in the shoe? 


Design 

The study included a pilot study (n 5 5) and a prospective 

intervention study (n 5 30). Children attending our clinic and 

diagnosed with Sever’s injury were randomized to either group 

I, heel cup treatment or to group II, without heel cup or other 

insole treatment during the study period of 4 weeks (Fig. 2). A 

third matched group of painless children (n 5 15) were included 

as a control group. 

Sample 

A total of 53 children fulfilling the inclusion criteria were 

offered to participate in the study. Three children (all girls) did 

not accept. No one dropped out of the study, leaving 50 who 

finished the study. Forty were boys and 10 were girls. Thirtyfive 

children with Sever’s injury were consecutively included, 

seeking for heel pain at a Sports Medicine Clinic. The median 

Fig. 1. The individually made molded rigid ‘‘Wessmark’’ 

cup, which was used in the study. 

48 

Pilot group 

n = 5 

Untreated patients 

n = 15 

Randomised group 

n = 30 

Heel cup treated patients 

n = 15 

age was 11 years (range 9–14 years). Twenty-four children had 

bilateral pain and 11 children had unilateral heel pain (eight 

boys, three girls, five with pain in the right and six in the left 

foot). The painless children (n 5 15) were individually matched 

(same gender, age, activity level and playing in the same sports 

team) as the children in group I (Fig. 2). The children were 

recruited during a 6-month period. 

Inclusion and exclusion criteria 

Inclusion criteria were both gender, age 9–15 years, and the 

diagnosis of Sever’s injury, except for the children without 

symptoms who did not suffer from Sever’s injury. Findings 

confirming the diagnosis were tenderness over the lower onethird 

of the posterior calcaneus and a positive calcaneus 

compression test, the squeeze test, where all participants 

were examined by the first author. Heel pain had to be present 

for more than 2 weeks when examined. The most painful 

activity had to be a ball sport of some kind (soccer, floor ball, 

handball, etc.) with a pain level of 4 or more on the Borg CR- 

10 scale during the last week activities. The Borg scale is a self 

assessment VAS scale with anchors of verbal explanations 

added to the numbers in the pain scale (Neely et al., 1992). 

Zero represents absence of pain and 10 represents maximal 

pain. All participants had to be high-level athletes, measured 

with Engstrom’s five level activity score index (Engstro¨ m, 

2004). They had to master Swedish, and together with their 

parents they had to agree to participate. 

Exclusion criteria were intermittent pain, specified disease 

that could interfere with the heel pain, dominant pain from the 

Achilles tendon, earlier fracture in the area, a poorly defined 

ache in the lower extremities and participation in another 

study including any kind of treatment for pain. 

Radiographic examination – heel pad thickness 

The pilot group was only recruited to test the radiographic 

examination. In the true radiographic examination the thickness 

of the heel pad was measured (n 5 45) not including the 

pilot group, and the examinations were analyzed by the last 

author. The lower vertex of calcaneus was the bone reference 

mark, and a steel band fixated on the skin under the center 

portion of the heel was the skin reference mark. A 5 mm steel 

bullet marker was fixed on the medial side of the ankle as a 

reference, in order to calibrate each radiograph for each child. 

The heel pad thickness is defined as the distance between the 

lower vertex of the calcaneus and the steel band fixed under 

the heel. The measured error within the computer analysis was 

estimated to 0.2 mm. The radiographic investigator was partly 

blinded. The reliability test for heel pad thickness measure- 

Matched control group 

n = 15 

Fig. 2. Study design with two randomized groups with Sever’s injury. There was a 4 weeks intervention with heel cup in the 

treated group (group I). The untreated randomized control group (group II) is supposed to represent the natural course. 

2

ments on radiographs (n 5 30) had a difference between the 

two series of 1.8%, (mean error 0.15 mm, SD 0.31). In the 

reliability test the investigator was blinded, analyzing single 

radiographs. 

The thickness of the heel pad in the child’s most painful heel 

was estimated on a lateral view, with the matched painless 

child’s same heel as a control. 

In all three groups two radiographs were taken as follows: 

Barefoot standing on one heel. 

In the sports shoe standing on one heel. 

In the treatment group one additional radiograph was taken 

as follows: 

In the sports shoe standing on one heel with the heel cup in 

the shoe. 

Peak pressure measurements 

Ten study subjects were included, randomized from the 

intervention group with insoles, group I. Each participant 

used a pair of indoor sports shoes. The test included three 

measurements standing with shoes alone and three with heel 

cups in the sports shoes, nine-step measurements of running 

with shoes alone and finally nine step measurements with heel 

cups in the sports shoes. 

A plantar pressure measuring system was used (F-Scan; 

Tekscan, Boston, Massachusetts, USA), which is a system 

tested for validity and reliability, recommended for relative 

comparisons of plantar pressure distribution (Ahroni et al., 

1998; Randolph et al., 2000; Tsung et al., 2004). The F-Scan 

sensor consisted of two thin, flexible polyester sheets that had 

electrically conductive electrodes deposited in stripes perpendicular 

to the stripes on the other sheet. By measuring the 

changes in current flow at each intersection point (sampling 

frequency 100 Hz), the applied pressure distribution could be 

measured by a computer and displayed as an image with a 

resolution of 10 10 mm (see Fig. 4). The peak pressure is 

defined as the maximum pressure measurement within a 

defined area under the heel (the area surrounded by a green 

line in Fig. 4). The area and location of maximum pressure 

and the measured value is automatically determined by the 

software. The F-Scan sensor was calibrated according to the 

procedure proposed by the manufacturer. 

The first part of the test was standing on their heels, in the 

same way as in the radiographic examination. Three snap shot 

measurements were made in each trial. The second part was 

in-door running at a modest tempo. The children did three 

short trials with shoe and three with shoe with heel cup each. 

Peak pressure for the three steps in the middle of each trial 

were measured. 

To estimate the reliability there was a test made for the peak 

pressure tests using the mixed regression analysis Cronbach’s 

a. The coefficient varies over the interval [0,1] where 0 means 

that the included variables have no reliability at all and 1 

means perfect reliability. All four coefficients showed significant 

reliability ( 0.97). 

Questionnaire 

All children answered questionnaires at the study start 

(n 5 45). Everyone answered specific questions about heel 

pain in their chosen ball sport (Borg CR-10) during the last 

week before study start (week 0). The insole groups filled in the 

same questions immediately after treatment. All randomized 

A heel cup improves the function of the heel pad 

children (n 5 30) answered the same questions at the end of the 

study. 

Statistics 

Radiographic measurements were analyzed with Student’s ttest 

with registration of millimeter thickness (mean), confidence 

interval and P-value. From the questionnaire measurement 

of pain level at start and follow-up with Borg CR-10 was 

performed. 

A linear mixed regression model was used for the analysis 

of peak pressure as response variable, using the child as a 

random effect and step 1, 2, 3 in running and treatment (heel 

cup yes/no) as fixed effects. In both cases the logarithm of the 

peak pressure was used resulting in a multiplicative model. 

The effect of the heel cup was given as the peak pressure ratio 

with values o1 meaning decrement and values 41 meaning 

increment. All analyses were made using the xtreg procedure 

(Stata, 2007). The normal distribution approximation fitted 

well for the logarithm of the running peak pressure. The 

approximation for the logarithm of the standing peak pressure 

showed over-dispersion. To remedy this we tried a model 

based fitted with procedure gllamm (Skrondal & Rabe-Hesketh, 

2004) using adaptive quadrature (Rabe-Hesketh & 

Skrondal, 2008) on pressure levels (six levels each 200 mmHg 

wide) of peak pressure yielding approximately the same fit for 

the predicted levels as we did using the previously described 

normal error model. Using this extra analysis we addressed the 

potential problem of the poor normal approximation. 

All data were calculated in commercially available software 

(SPSS, Stata Corp.). 

The study has been approved by the Regional Ethics Board 

of Uppsala, Sweden (dnr 2008/094). 

Neither the Public Sports Medicine Clinic nor the authors 

have any relation to, profit or other benefits from the 

companies that manufacture the products in this study. 

Results 

Activity level 

All children in the two symptomatic groups continued 

their high levels of physical activity during the 

study period. Three children, two from the intervention 

group and one from the control group, increased 

their activity level from D to E from start of the study 

to follow-up. There was no decrease to a lower 

activity level. 

Heel pad thickness 

Heel pad thickness was significantly increased by 

2.19 mm (mean) when using sports shoe without 

heel cup compared with standing barefoot 

(Po0.001). Adding the heel cup in the shoe again 

significantly increased the heel pad thickness by a 

further 1.43 mm (P 5 0.002). Hence, heel pad thickness 

was increased by 3.57 mm (mean) when using a 

sports shoe with heel cup compared with barefoot 

standing (Po0.001), Table 1. An example of the 

assessment is shown in Fig. 3. One radiograph in the 

treatment group with heel cup in the shoe was 

missing, because of a mistake by the staff. Age and 

3 

49


Table 1. Analysis of heel pad thickness during standing bare foot vs in shoe, and without heel cup in shoe vs with heel cup 

body weight did not seem to affect the thickness of 

the heel pad standing barefoot or in shoes with or 

without a heel cup, even though there was a total age 

range of 4 years of age and a weight range of 32 kg 

(Table 2). There was no difference in heel pad 

thickness when comparing studied soccer shoes 

with indoor sports shoes. Comparing the heel cup 

group with pain with their matched controls without 

pain, the heel cup group showed a non-significant 

tendency toward thinner heel pads both bare foot 

and in shoe, Table 2. 

Peak pressure 

Maximal peak pressure comparing sports shoe with 

and without heel cup was reduced during running 

and standing in favor of the sports shoe with heel 

cup. In the statistical analysis during running the 

effect of heel cup treatment was significant. The 

multiplicative effect of heel cup was 0.79 [0.75, 

0.83], resulting in a reduction of heel pad peak 

pressure with 21% compared with the shoe without 

insole. Data from one child was missing, because of 

damage to the F-Scan sensor (n 5 9), and the last 

trial running including heel cup treatment was deleted 

due to sensor errors for one child. In the 

corresponding analysis of standing peak pressure, 

the effect of the heel cup was significant, with a 

multiplicative effect of 0.75 [0.72, 0.78], resulting in 

a reduction with 25%. There were no data missing. 

50 

n Increase in heel pad 

thickness, mean (mm) 

95 % confidence interval P-value 

Bare foot vs Shoe only 45 2.19 1.69–2.69 o0.001 

Bare foot vs Shoe with cup* 15 3.57 2.56–4.29 o0.001 

Shoe only vs Shoe with cup* 15 1.53 0.63–2.23 5 0.002 

*Treatment group only. 

Fig. 3. Radiographic assessment of heel pad thickness when standing on heel in shoe without cup (left) and with heel cup 

(right). The most painful heel was investigated and the distance from vertex calcanei and the skin reference mark was measured. 

Note how the cup restores soft tissue thickness. The difference in heel pad thickness was in this case 1.95 mm. 

4 

Table 2. Heel pad thickness in different subgroups 

Heel pad thickness during standing, 

mean (mm) 

Bare 

foot 

Sports 

shoe 

Shoe with 

cup* 

Age 

9–10 years (n 5 10) 6.86 8.55 10.66 



Body weight 

30–37 kg (n 5 15) 6.30 8.31 9.43 



Type of sports shoe 

Soccer shoes (n 5 23) 6.66 8.93 10.12 

Indoor shoes (n 5 22) 6.79 8.90 10.18 

Cup treated (group I) 6.51 8.83 10.37 

No treatment (group II) 7.13 9.22 – 

Controls (group III) 7.00 9.42 – 

*Treatment group only. 

Table 3 shows the mean values in mmHg in running 

and standing. Figure 4 shows one typical example of 

maximal peak pressure measurement. 

Pain relief 

Table 4 shows a significant pain reduction in sport 

activities in the treated group during the study 

period. There was an immediate reduction of pain

Table 3. Peak pressure during standing and running in tested subgroup 

Peak pressure, mean (mmHg) 

n Shoe 

only 

Shoe with 

cup 

Difference 

(%) 

Standing on heel 10 1407 1056 25 

Running 9 1412 1109 21 

Fig. 4. Maximal peak pressure during running (three steps) 

without cup (upper row) and with cup (lower row) for one 

child. Square show area of maximal pressure under forefoot 

and hindfoot, respectively. Red color represents the highest 

pressure (41385 mmHg). 

when using heel cup, with further improvement at 

follow-up. The persisting pain with the cup was close 

to 0 ( 5 0.5). The reduction in pain level in the non- 


treatment group is thought to reflect the natural 

course. 

Discussion 

This study showed that the use of a heel cup resulted 

in increased heel pad thickness and decreased heel 

pad peak pressure with corresponding pain relief. 

The findings indicate improved heel pad function, 

that probably protects the injured heel from further 

micro trauma. The very low level of persisting pain 

after 4-week treatment with the cup was close to 0 

( 5 0.5). These results confirm our results in previous 

studies. 

The importance of a functional heel box to avoid 

overuse injuries in adults was presented by Jo¨ rgensen(1989). 

Inspired by the thesis by Jo¨ rgensen, the 

former chief at our clinic, Dr. Wessmark, designed 

and produced a thermoplastic rigid heel cup with a 

2–3 cm brim, instead of the traditional cup with its 

spoon shape. The ‘‘Wessmark cup’’ has thereafter 

been used on this indication. There is no need to 

change shoe size because of the added heel cup. 

Earlier studies proposing insoles have focused on 

the relationship between Sever’s injury, pes equinus 

and malalignment. Szames pointed out that equinus 

had a definite impact on predisposing a child to 

Sever’s disease (Szames et al., 1990). In a retrospective 

analysis from his clinic, 482% with Sever’s 

injury also had equinus. The primary ambition has 

been to normalize the gait/running pattern (Madden 

& Mellion, 1996). An excessive pronating foot type is 

thought to be associated with poor shock absorption 

and therefore contributes to micro trauma of the 

heel. This malalignment also may contribute to overactivity 

of the calf muscle complex, because of the 

necessity to stabilize the talocrural joint and/or to 

substitute the eversion of calcaneus. A heel wedge, a 

molded plastizote orthosis or a spoon-shaped heel 

cup to align the running cycle, have been recommended 

from these points of view (Micheli & Ireland 

1987; Staheli, 1998; Kasser, 2006). All our children 

have been able to walk and stand on their heels with 

dorsiflected feet and do not seem to have tight 

Achilles tendons, defined as having o101 of dorsiflexion 

in the ankle joint, with the knee fully extended. 

In an often cited retrospective study by Micheli 

and Ireland (1987), the frequency of pronation (without 

definition) in a group of 85 patients was 27%. 

Almost everyone (98%) was treated with insoles 

despite malalignment or not. The insoles used were 

soft plastizote orthotic 75%, viscoelastic heel cup 

18%, or heel wedge 5%. Approximately 30% in our 

studies had unilateral pain. The fact that unilateral 

pronation and unilateral short calf muscle seem to be 

5 

51


Table 4. Pain level measured with Borgs CR-10 scale 

unusual, are two reasons to question these treatment 

strategies in Sever’s injury. The golden age of tissue 

development peaks at 11–12 years. The physiological 

remodelling to develop strong and lasting tissues 

with enough capacity to resist injuries (Wolff’s law) 

requires physical activity with repeated impact force 

to the hind foot. This relation between stimulating 

load and developing capacity seems to be disturbed 

in Sever’s injury. Histological analysis of the region 

shows a primary compression, rather than a tension 

cytoarchitecture (Ogden, 2000). The heel pad with its 

sum of cooperating compartments provides cushioning, 

and thereby reduces the impact force on calcaneus 

in heel strike (Ogden, 2000). An effective 

support to the heel pad therefore seems to be crucial, 

and this is supported by the findings in our previous 

study were the heel cup was shown to be superior to 

the wedge in giving pain relief (Perhamre et al., 

2010a). 

Soccer is the predominating sport among our 

patients. Walter & Ng(2002) showed that the studded 

soccer shoe placed the foot in a dorsiflexed position, 

increasing the amount of pull from the Achilles 

tendon, exceeding the pressure placed upon the 

calcaneal apophysis. The conclusions from our previous 

studies are that pain in Sever’s injury primarily 

is an effect of impact, not shear forces. The reason 

why soccer dominates is probably simply because it is 

the dominating sport in this age, including both 

genders. It includes much running and jumping and 

the shoes offer little heel support. 

Our two previous studies have shown immediate 

and lasting pain relief with the heel cup, confirmed in 

the present study (Perhamre et al., 2010a, b). The 

purpose with this study was to assess the possible 

mechanisms behind the rapid and impressive pain 

relief. There was a significant increase of heel cup 

thickness from barefoot to shoe and from shoe to 

heel cup (Table 1, Fig. 3). There was also a significant 

reduction of the heel peak pressure using heel cups 

(Fig. 4). We believe that both these effects strongly 

support Ogden’s theory that Sever’s injury is a 

traumatic condition due to repetitive impact forces 

during heel strike with corresponding pathology in 

52 

Pain at start (median values) Pain at 4 weeks (median values) 

Shoes only With cup Shoes only With cup 

Group I (treatment) 7.0 2.0 3.0 0.5 

Group II (no treatment) 7.0 – 5.0 – 

6 

the rear foot (Ogden et al., 2004). An effective 

treatment seems to reduce these forces. The cup 

seems to protect the heel, significantly reducing 

pain, compared with our non-treatment group (Table 

3). The sample in the peak pressure test was small 

(n 5 10) with one child missing in the running trials. 

This reduces the power of our conclusions. An 

interesting finding was that children with Sever’s 

injury show a tendency toward thinner heel pads 

compared with the painless control group (Table 2). 

This could possibly be a predisposing factor to 

Sever’s injury. 

Perspectives 

Sever’s injury seems to be a result of high impact 

forces during an intense puberty period with an 

immature heel. The traditional recommendation to 

stop or diminish the sports activity for an unspecified 

period of time seems inappropriate and we suggest 

that this should be replaced by the general advice to 

use heel cups and continue with the activities. The 

effect of using the rigid heel cup can be explained 

both by less deformation, preserving the heel pad 

thickness when loaded, and by the reduction of heel 

pad peak pressure during heel strike. 

Key words: Sever’s injury, apophysitis calcanei, insoles, 

heel cup, heel pad thickness, heel peak pressure. 


Thanks to Ingemar Hansson, who made the heel cups, Klara 

Perhamre for help with the figure layout, Mikael Hasselgren, 

PhD, medical chief Centre for Clinical Science, Eva Stjernstro¨ 

m RPT, division chief Va¨ rmland County Council for 

support and Ola Wessmark, former chief doctor at our Clinic, 

who made the prototype of this heel cup. 

Financial support was provided by the Primary Care 

Research Unit, by the division Health, Habilitation and 

Rehabilitation, the foundation for Orthopaedic Devices, 

Va¨ rmland County Council, O¨ rebro University Hospital Research 

Committee and Swedish National Centre for Research 

in Sports.

References 

Ahroni JH, Boyko EJ, Forsberg R. 

Reliability of F-scan in-shoe 

measurements of plantar pressure. 

Foot Ankle Int 1998: 19: 668–673. 


medicine, 3rd edn. New York: 

McGraw-Hill, 2007: 737pp. 

Cavanagh P, Rogers MM, Iiboshi A. 

Pressure distribution under symptomfree 

feet during barefoot standing. Foot 

Ankle 1987: 7: 262–276. 

Donatelli R. The biomehanics of the foot 

and ankle. Philadelphia: F.A. Davis 

Company, 1990: 284pp. 

Engstro¨ m LM. School–sports–health: 

studies of the subject Sports and 

Health, and young people’s physical 

activity, physical capacity and health 

condition: basis, purposes and methods 

[in Swedish]. Stockholm: 

Idrottsho¨ gskolan, 2004: 142pp. 




Jo¨ rgensen U. Implication of heel strike: 

an anatomical, biomechanical, 

physiological and clinical study with 

focus on the heel pad. Link o¨ ping: 

Linko¨ ping University, 1989. 

Kasser JR The foot-aquired conditions. 

In: Weinstein SL, Lovell WW, Winter 

RB, Morrissy RT, eds. Lovell and 

Winter’s pediatric orthopaedics 2, 6th 

edn. Philadelphia: Lippincott Williams 

& Wilkins, 2006: 1257–1329. 




54: 1995–2000. 

Micheli LJ, Ireland ML. Prevention and 


in children: an overuse syndrome. J 


Neely GLG., Syle´ n C, Borg G. 

Comparison between the visual 

analogue scale (VAS) and the category 

ratio scale (CR-10) for evaluation of leg 

exertion. Int J Sports Med 1992: 13: 


Ogden JA. Foot. In: Ogden JA, ed. 

Skeletal injury in the child, 3rd edn. 

New York: Springer, 2000: 1091–1159. 



the immature calcaneal metaphysis. J 


Perhamre S, Janson S, Norlin R, Kla¨ ssbo 

M. Sever’s injury: treatment with 

insoles provides effective pain relief. 

Scand J Med Sci Sports 2010a May 

(Epub ahead of print). 

Perhamre S, Lundin F, Norlin R, Kla¨ ssbo 

M. Sever’s injury; treat it with heel cup – 

a randomised, crossover study with two 

insole alternatives. Scand J Med Sci 

Sports 2010b , (Epub ahead of print). 

Rabe-Hesketh S, Skrondal A. Multilevel 

and longitudinal modeling using stata, 

2nd edn. College Station: Stata Press, 

2008. 


Randolph AL, Nelson M, Akkapeddi S, 

Levin A, Alexandrescu R. Reliability 

of measurements of pressures applied 

on the foot during walking by a 

computerized insole sensor system. 

Arch Phys Med Rehabil 2000: 81: 573– 

578. 

Sever JW. Apophysitis of the os calcis. 

NY State J Med 1912: 95: 1025–1029. 

Skrondal A, Rabe-Hesketh S. 

Generalized latent variable modeling, 

1st edn. Boca Raton: Chapman and 

Hall/CRC, 2004: 508pp. 

Staheli LT. Sports/foot and ankle. In: 

Staheli LT, ed. Fundamentals of 

pediatric orthopedics, 2nd edn. 

Philadelphia: Lippincott-Raven, 1998: 


Stata P. STATA/IC release 10 user 

manual [XT] – longitudinal/panel 

data. College Station: Stata Press, 

2007. 






Tsung BYS, Zhang M, Mak AFT, Wong 

MWN. Effectiveness of insoles on 

plantar pressure redistribution. JRRD 

2004: 41: 767–774. 

Walter JH, Ng GK. The evaluation of 

cleated shoes with the adolescent 

athlete in soccer. The Foot 2002: 12: 


7 

53

Stefan Perhamre, specialist i allmänmedicin arbetade sen 1993 på Idrottshälsan i Karlstad. 

Severs skada – paradigmskifte gällande 

diagnostik och behandling? 

Stefan Perhamre, idrottsläkare i Karlstad samarbetade tidigare med dr Ola Wessmark, 

och ortopedtekniker Ingemar Hansson vid Idrottshälsan, Landstinget i Värmland. Till 

Idrottshälsan sökte sig många barn och ungdomar med hälsmärta och de tyckte att en 

hård hälkopp fungerade påfallande bra som smärtlindring. 

Stefan Perhamre startade ett forskningsprojekt, tillsammans med Maria Klässbo, för att 

utvärdera hur bra hälkoppen verkligen var. När det blev klart att Stefan ville skriva en 

doktorsavhandling bildades en forskargrupp kring hans avhandlingsprojekt. 

Med Stefans arbeten som grund har vi kunnat visa att diagnosen Severs skada (hälsmärta 

hos unga) kan fastställas med vanliga kliniska test. Slätröntgen har inget diagnostiskt värde. 

Studierna har visat att den hälkopp Ola Wessmark tidigare designat, fungerar mycket 

bra som behandling vid Severs skada och att orsaken till besvären rimligen är en överbelastningsreaktion 

i hälbenet. Vi har vidare kunnat visa att hälkoppen bidrar till att hälkudden 

trycks ihop från sidorna och blir lite tjockare under hälbenet och att toppbelastningen 

på hälen minskar.

Severs skada – paradigmskifte gällande diagnostik och behandling?

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?