Tuesday October 28 2008
‘HRT can increase the risk of needing a joint replacement’ reports The Daily Telegraph today, claiming that women using hormone replacement therapy are 1.5 times more likely to need a knee replacement due to arthritis than women who have never used HRT. The newspaper also reports an increased risk of hip replacement. The Sun reports the risk of both operations as being doubled.
This study has strengths, such as the very large number of women involved (1.3 million) and reliable sources of medical data. The report does suggest a relationship between HRT use and increased risk of joint replacement, particularly of the knee, plus relationships between surgery risk and other reproductive factors. However, there were some limitations to the methods of data collection used in this study, and its findings conflict with those of previous studies.
Based on this study alone it cannot be concluded that HRT use increases the risk of osteoarthritis, and the reasons behind the observed relationship between HRT and joint replacement are still unclear. The risk of a woman needing joint replacement is relatively small, and the results of this study are not strong enough to suggest that women should change their current use of HRT.
Where did the story come from?
Dr Bette Liu and colleagues from the Universities of Oxford and Southampton carried out this research. The study was funded by Cancer Research UK, the NHS Breast Screening Program and the Medical Research Council. The study was published in the peer-reviewed medical journal Annals of Rheumatic Disease.
What kind of scientific study was this?
This was a cohort study which aimed to examine the effect of reproductive history and use of hormonal treatment on the risk of knee or hip replacement due to osteoarthritis.
The research follows on from previous observational studies suggesting a link between female sex hormones and the development of osteoarthritis, a condition which is more prevalent in women and also known to increase around the time of menopause.
The research followed 1.3 million women as part of the Million Women Study. They were recruited between 1996 and 2001 through NHS breast screening centres, and had an average age of 56 years. Information was collected on reproductive history and use of hormonal therapies, in addition to other medical history and lifestyle. A repeat survey was conducted three years later, which additionally linked participants, via their NHS number, to death and cancer registry data. Data was also gathered from hospital admission databases, which categorise admissions using up to 14 codes for diagnoses and 12 codes for procedures.
Researchers looked for procedure codes for hip or knee replacement with corresponding diagnostic codes for osteoarthritis. Women who had these operations before recruitment date and those with cancer were excluded from the study. The researchers calculated ‘person-years’ for individual participants, meaning the number of years they were each observed by the study. Person-years were calculated from the date of each woman’s entry into the study until the date of a joint-replacement operation, death or the end of the study.
Risk of operation was calculated for a number of variables including number of children, age at which periods started, age at menopause, contraceptive pill and HRT use. Each risk calculation was adjusted for other variables in the study that could also affect risk.
What were the results of the study?
A total of 1,306,081 women were monitored in the study after exclusions. The women were followed for an average of 6.1 years, during which time 12,124 had hip replacements (1.5 cases per 1000 person-years), and 9,977 had knee replacements for osteoarthritis (1.2 cases per 1000 person-years).
Women who had the joint replacements differed from the rest of those observed on several variables: those with hip or knee replacements were generally older at recruitment, more overweight and had less prior use of oral contraceptives. Those with knee replacement were also more often from lower socio-economic groups and less likely to smoke or drink alcohol. There were also relationships between some of the other variables tested, such as increased BMI with increased number of children.
There was no significant relationship between use of contraceptive pill and risk of joint replacement. Compared to those who had never used HRT, there were significantly increased risks of both operations for those who had used HRT in the past (13% for hip and 39% for knee replacement), and those who used HRT at the time of the study (38% for hip and 58% for knee replacement). There was a significant trend for decreased risk of hip replacement with increasing duration of HRT use (49% risk with use <5 years; 26% risk with use ≥12 years). There was no such significant relationship for knee replacement and duration of use.
Examining the role of reproductive factors the risk of knee replacement was significantly related to the number of children the woman had with risk increasing with every child, when compared to women with no children. There was a similar, but less significant, trend for hip replacement and number of children.
Menstrual periods commencing at the age of 11 or younger, compared to 12 years also carried a slightly increased risk of either operation. There was no relationship between risk and starting menstruation at an age greater than 12.
What interpretations did the researchers draw from these results?
The researchers conclude that ‘hormonal and reproductive factors increase the risk of hip and knee replacement, more so for the knee than for the hip. The reasons for this are unclear.’
What does the NHS Knowledge Service make of this study?
This study has strengths, in that it has included a large number of women and has used reliable sources of data to obtain medical information. The report does suggest relationships between both HRT use and other reproductive factors and the need for a joint replacement. However, there are some limitations:
- Information about HRT use was gathered at only two time points, at study entry and at the second survey about 3 years later. Assumptions may have to have been made about participants’ HRT use at the time of the knee or joint replacement.
- Although women with joint replacements prior to the study were excluded, no information is available on the severity of arthritis of those included. For example, some women could have had severe arthritis prior to commencing HRT. Therefore the study cannot conclude that HRT increases the risk of osteoarthritis.
- As the researchers acknowledge, the observed relationship between joint replacement and HRT may be confounded by other related factors; for example, women taking HRT may have had greater access to medical services and therefore been more likely to receive treatment for other conditions, such as surgery.
- The study was not able to assess other treatments that may have been used as an alternative to surgery, so some subjects’ might not have been inluded if their condition was treated by other non-surgical means.
- There may have been some miscoding in the hospital admissions database, allowing cases of hip/knee replacement to have been missed. Additionally as only an NHS admissions database was used, private cases would have been missed.
- The authors note that there have been other studies on the effect of HRT on hip and knee results, and that their findings have been inconclusive. It is unclear why some of these studies did not identify a link, while the current study did, although it may be related to the size or methods of these other trials.
The reasons for the observed relationship between HRT and joint replacement, or why the risk seems to be greater for knee than hip, are unclear and will require further study. The authors suggest that, as there are oestrogen receptors on bone and cartilage (giving the benefit of maintaining bone density and preventing osteoporosis) of exposure to oestrogen from HRT may cause osteoarthritic changes in the bone.
The risk of a woman needing joint replacement is relatively small, and given the conflicting evidence from this research and previous studies, the evidence is not strong enough to suggest that women should make changes to their current HRT use.
Sir Muir Gray adds...
On its own, the evidence from this study is not strong enough to be the deciding factor on whether or not to have HRT. But this is another factor for women to consider when facing this difficult choice.