Table 2.

Prospective studies of functional outcomes in obese and non-obese patients with total hip arthroplasty (THA)

Study	N Follow-up	Sample	Surgical Type & Components	Results
Aderinto et al [15] (2005)	140 3 years; Prospective Follow-up	61% were women; groups were <30 or ≥30 kg/m2	Cemented prostheses; approaches or components were not described	At year 3, HHS scores ↑ from 44 to 90 points (non-obese) and from 42.5 to 85 points (obese), with no difference between groups; lower scores for stairs, sitting and putting on shoes-socks and range of motion were lower in the obese patients (p < 0.05).
Andrew et al [16] (2008)	1,421 5 years; Multi- center Prospective	62% were women; BMI groups were <30, 30- < 40 and ≥40 kg/m2 Mean ages were 69.1, 65.5 and 60.6 years	Anterolateral or posterior approaches were used; cemented Stryker Exeter femoral components and several different acetabular components	By year 5, OHS scores were best in the non-obese group and worst in the obese group (19.6 vs 25.6 points; p = 0.005), but no differences in the 5 year change in OHS scores existed. No differences in rates of revision, dislocations or medical complications existed.
Busato et al [3] (2008)	18,968 15 years; Multi- center prospective	sexes NS; BMI groups were <25, 25- <30 and ≥ 30 kg/m2	Surgical components not described; surgical type not described; Data were obtained from the Total Hip Registry (Switzerland)	High preoperative BMI was related with a dose-effect response with shorter unsupported walking time less normal stair climb, and shoe tying during the 15 year follow-up, despitesimilar pain relief across BMI brackets.
Chan and Villar [17] (1996)	166 to 3 years; Prospective cohort	59% were women BMI groups were <25, 25–29.9, 30–39.9, >40 kg/m2 Mean ages were 71.4, 69.0 and 68.1 years	Surgical components not described; surgical type not described	HHS scores were superimposed onto the Rosser Index Matrix (which ranks disability status); there were no differences in Rosser scores for disability among the BMI groups by year 3.
Chee et al [6] (2010)	108 5 years; Prospective cohort	41% were women; BMI groups were <35 kg/m2 or >35 (1 comorbidity) and >40 kg/m2	Anterolateral approach was used on all patients; 25.5% Charnley prosthesis (dePuy, Int.), 74.5% Lubinus SPII prosthesis (Waldmar-Link GmbH); cemented	Five year HHS were higher in non-obese than morbidly obese patients (91.8 vs 85.4 points; p < 0.0001) despite similar pre- operative scores; SF-36 subscores for physical functioning were lower in morbidly obese patients at year 5.
Dowsey et al [29] (2010)	471 1 year; Prospective follow-up	60% were women; BMI groups were <30, 30–39 and ≥ 40.0 kg/m2	Surgical procedures not described; surgical components not described; cement used varied across groups	Morbidly obese patients had a lower change in HHS function scores than obese and non-obese patients, respectively by year 1(11.5 vs 15.6 and 16.2 points, respectively); HHS were lowest in morbidly obese patients by year 1 (70.5 vs 79.8 and 80.8 points p = 0.03).
Gandhi et al [26] (2010)	707 hips I year Prospective cohort	59-66% were women; waist circumference was assessed for metabolic syndrome BMI ranged from 22.0 to 36.6 kg/m2 64.8-66.2 years	Surgical components or procedures not described; patients were obtained from a registry	1 year WOMAC scores (pain, function) were highest in patients with 4 metabolic syndrome factors compared to those with fewer factors; regression B coefficients showed that obesity predicted 1 year WOMAC scores (B = 2.4 1.4-4.2; 95% CI).
Jackson et al [23] (2009)	2,026 mean of 5 years; Prospective cohort	77% were women; BMI groups were Mean ages were 68 and 63 years	Posterior surgical approach; ABG2 Stryker cementless femoral and acetabular components	HHS were lower in obese vs non-obese patients at follow-up (89.9 vs 93.2 points); HHS functional scores were also lower in the obese group (29.6 vs 31.0 points); hip flexion, adduction and internal rotation ranges were less in the obese vs non-obese patients. HHS pain scores were not different between groups.
Judge et al [18] (2010)	1,327 1 year EUROHIP Study of 20 orthopedic centers	56% were women; BMI groups were <30, 30–39 and ≥40 kg/m2 Ages <50 to ≥ 70 years	Surgical components not described; surgical type not described	Median WOMAC scores were highest in the morbidly obese group pre-THR, (68.1 vs 61.5 and 57.6 points). but the 1 year change in WOMAC score was highest in the morbidly obese group (median score change of 56.1 vs 33.3 and 37.5 points); morbidly obese patients showed a ↑ OR of “returning to normal” (functionality) than the other groups.
Lubbeke et al [28] (2007)	435 5-year; Prospective cohort	53-55% are women; BMI groups were  < or ≥ 30 kg/m2 Mean ages were 68 72 years	85% of patients had mixed components (1 cemented,1 non)	Obesity was related with worse outcomes after revision than primary THA by year 5 (lower HHS scores: 76.7 vs 88.1 points; lower WOMAC function scores 61.6 vs 70.0 points). BMI was related to the mean difference of HHS scores for primary and revision THA (R coefficient = −1.0 [−0.1 to −1.95% CI]).
Lubbeke et al [8] (2007)	2,495 5-year; Prospective cohort	48.7-57.5% women; 8.5-9% were revisions; BMI groups were < or ≥ 30  kg/m2 Mean age 69 years	95% were lateral THA approach, 86% used Morscher press-fit uncemented actetabular component and Muller straight stem cobalt chromium femoral component	BMI of ≥30 kg/m2 was related with a RR of 3.7-4.0 for revisions, 9.1-12.5 for dislocations and1.9-8.0 for infections in obese compared to non-obese men and women. By year 5, HHS were 87.8 and 79.6 points in non- obese and obese women and 90.5 vs 87.4 points in non-obese and obese men; WOMAC function scores were 14.7% and 8.0% lower in obese women and men thantheir non-obese counterparts.
Lubbeke et al [30] (2008)	204 5-year; Prospective cohort	50-7.9% were women; BMI groups were <  or ≥ 30 kg/m2 Age range <50 to ≥80 years	Cemented acetabular cups were used in 67% and 80% of obese and non-obese patients	HHS were 82.8 ± 14.7 and 71.4 ± 17.0 points in the non-obese and obese patients by year 5. Surgical revisions were performed at 92 and 125 months in obese and non-obese groups, respectively. The adjusted hazard ratio for occurrence of infection, dislocation or re-revision increased from 1.0 (BMI < 25) to 1.5 (BMI 25–29.9) to 4.5 (BMI 30–34.9) to 10.9 (BMI ≥35.0).
Lubbeke et al [31] (2010)	503 5 or 10 years; Prospective follow-up	58% were women; BMI groups were <25, 25–29.9 and ≥ 30 kg/m2	Hybrid prosthesis; Morscher press fit uncemented cup and cemented cobalt-chromium stem (Zimmer); alumnia ceramic head, and a ceramic- polyethylene surface	At year 5, HHS ↓ with each rogressively higher BMI group (91.4, 88.4 and 85.1 points; p = 0.019). At year 10, HHS tended to be lower in patients with BMI ≥30 kg/m2 compared with those with BMI <25 and 25–29.9 kg/m2 (83.6 vs 87.3 and 87.1 points; p = 0.08); more obese patients had low UCLA scores and more non-obese patients had higher UCLA scores.
Moran et al [19] (2005)	800 6–18 mo; Prospective follow-up	61% women; BMI groups were <25, 25–29.9, 30–39.9 ≥40 kg/m2 Mean age was 68 years	All were anterolateral approach surgeries; components were not described	For every 1 point increase in BMI, HHS scores dropped by 0.25 by month 6 and by 0.35 by month 18 post-surgery. No BMI effect on early failure of THA was found.
Naylor et al [22] (2008)	198 1 year; Prospective observational	56% were women; BMI groups were <30 or ≥30 kg/m2 Mean age 67 years	Surgical components not described; approaches not described	Obese patients had smaller increases in timed mobility than non-obese patients (0.23 m/s slower on 15 m walk time) and the timed up and go test (3.1 sec slower) at year 1; WOMAC scores for function and pain were worse in obese than non-obese patients by year 1.
Singh et al [12] (2009)	2,687 2–5 years; Prospective cohort of revision THA	53-54% were women; BMI brackets were <25, 25–29.9, 30–39.9 ≥40 kg/m2 Mean 5 year age was 65 years	Surgical components not described; approaches not described	At year 2, the OR for complete dependence onwalking/gait aids was 2.0 (vs 0.9 for BMI 25–29.9); moderate to severe activity limitation was predicted by high BMI. The OR of reporting difficulty in 3 of 7 mobility and functional tasks ↑ from 1.2 to 2.7 with increased BMI from 25–29.9 to 40 kg/m2, by year five, the OR increased to 1.3 and 3.0 in these same BMI brackets (all p < 0.01).
Søballe et al (1987) [20]	125 5 years; Prospective follow-up	A weight index was calculated as < or > 120% of pre-surgical weight; analyses were also performed using weight brackets of < or > 80 kg; Mean age at follow-up 70 (28–89) years	All were posterolateral approach surgeries; Lubinus prostheses were used and fixed with gentamicin impregnated radiopaque PMMA; One surgeon performed all hip replacements	Walking ability, defined using the Charnley scoring system was lower in patients with a weight index >120 pre-surgery, but similar to patients with indexes <120 by year 5 (4.9 vs 5.0 points; p = NSig).
Stickles et al [13] (2001)	5921 year; Prospective follow-up	56% were women; BMI brackets were <25, 25–29.9, 30–40 >40 kg/m2 Mean age 69 years	Surgical components or procedures not described; patients were obtained from a registry	By year 1, stair ascension and descension difficulty was reported in 86-88% of very obese patients compared with 46-55% of non-obese patients; BMI did not correlate with change in WOMAC scores (31.8 and 35.9 points, in non-obese and very obese patients, respectively; p > 0.05).

THA = total hip arthroplasty; BMI = body mass index; EUROHIP = European Collaborative Database of Cost and Practice Patterns of Total Hip Replacement.

OR = odds ratio; RR = relative risk; HHS = Harris Hip Score; UCLA = University of California and Los Angeles (UCLA) activity scale; Medical Outcomes SF-12 = Short Form 12; WOMAC = Western Ontario and McMaster Osteoarthritis Index; VAS = visual analogue scale.

NSig = non-significant.