Abstract
This review covers the evidence relating to lifestye modification in the big three musculoskeletal conditions: osteoarthritis, osteoporosis and rheumatoid arthritis. Lifestyle is of considerable importance in the first two and there is emerging evidence for rheumatoid arthritis despite it not traditionally being considered a lifestyle disease.
Introduction
Osteoarthritis (OA), osteoporosis and rheumatoid arthritis (RA) are three musculoskeletal diseases with substantial impact on both the affected individual and society as a whole. All three have both genetic and lifestyle contributions. The aim of this article is to discuss the evidence relating to prevention of these illnesses taking a life-course approach.
Osteoarthritis
OA is the most common form of arthritis and is increasing markedly due to an ageing population. It is characterised by a gradual loss of articular cartilage and changes to other joint structures (i.e., subchondral bone, ligaments, meniscus and synovium) leading, eventually, to total joint replacement. It is the most common joint disorder worldwide and is one of the most frequent causes of pain, loss of function and disability in Western populations . While genetic factors have been implicated in OA, few genes have been independently replicated. Lifestyle factors are also important and this article will discuss obesity (and diet), exercise, injury, vitamin D and smoking.
Obesity and diet
In younger life, knee pain is common in obese children, and is associated with increased weight/body mass index (BMI) . Furthermore, childhood BMI is associated with the presence of knee pain on walking in adults (mean age 36 years) and that being overweight in childhood is associated with higher levels of pain on walking in both normal and overweight adults . Moreover, BMI from age 11 years onwards has been shown to be associated with knee pain at age 45 years .
In adult life, obesity (however assessed) is an independent determinant of pain (knee, back and feet) and this is specifically for fat mass (as lean mass appears protective). It is strongly associated with radiographic change in the knee but less so in the hip and hand. It is also associated with structural changes such increased bone area and subchondral bone mineral density (BMD), bone marrow lesions, cartilage defects and meniscal tear/extrusion as well as cartilage glycosylated aminoglycan (GAG) content. The association with cartilage loss on magnetic resonance imaging (MRI) is less consistent in that it predicts cartilage loss in some studies and the effect on cartilage is at least partially mediated by leptin so is not solely due to loading.
In clinical trials, weight reduction is modestly but significantly effective for pain . Surprisingly, weight gain is more strongly associated with worsening pain than weight loss is with decreasing pain . There are less data for structure but observational studies suggest weight loss will improve cartilage defects and GAG content .
In terms of specific nutrients, there are a range of nutritional risk factors , though there are varying degrees of evidence to support them. These include dietary magnesium intake, energy, carbohydrate and sugar intake , vitamin C intake and vitamin K intake as well as dietary fatty acids. Serum high-density lipoprotein (HDL) cholesterol may be protective for bone marrow lesions in the knee .
Exercise
Longitudinal data in children suggest that children with an above median average sport intensity gained about twice as much tibial cartilage volume as those below median. This was consistent with the cross-sectional data where physical activity was a significant explanatory factor for patella, and medial and lateral tibial cartilage volume ( R 2 7–14% depending on site, all p < 0.05). In younger adult life, cartilage GAG content changes rapidly with exercise .
In later life, aerobic exercise is one of the major conservative management techniques for the treatment of knee OA. Several meta-analyses have definitively concluded there is strong evidence that aerobic exercise improves knee pain and function. As a result, aerobic exercise is currently recommended by all international guidelines for the treatment of knee OA. In terms of specific exercise, a systematic review by Roddy et al. compared the efficacy of aerobic walking and strengthening exercises in patients with knee OA. Out of the 13 randomised controlled trials (RCTs) included, four focused on aerobic walking. They reported a benefit from aerobic walking in reducing pain and self-reported disability.
Only one RCT has been conducted which examined the effect of exercise (specifically lower limb strength training) on disease progression as its primary outcome, assessed as joint space narrowing measured by radiography. Although, the strength training programme did not actually increase strength, the results showed a non-significant trend towards a beneficial effect (joint space narrowing occurred less often in the strength training group than in the ROM group (18% vs. 28%, p = 0.094). Surprisingly, in subgroup analysis, strength training significantly increased the rate of joint space narrowing in those participants with normal radiographs at baseline (34% vs. 19%, p = 0.038) suggesting it may be harmful. Other clinical trials examining either exercise and/or strength training for knee OA have shown no effect on structural progression. MRI has revolutionised the understanding of OA. The few studies which have employed this technology to assess the effect of exercise on knee structure have been observational in nature and show conflicting results. However, the majority have been cross-sectional and prone to bias. In midlife, longitudinal data suggest strenuous exercise may protective against cartilage defect progression . In later life, a recent paper which used pedometers to assess actual physical activity suggested that walking was deleteriously associated with knee structural change (including increases in BMLs, meniscal pathology and cartilage defects) over approximately 2.7 years, especially in those with pre-existing evidence of OA .
These results suggest that physical activity is beneficial for symptoms but may have varying effects on structure depending on age and stage of OA. Indeed, they may be harmful in those with established OA of the knee. Thus, there is a strong need for clinical trials to be performed to confirm or refute this hypothesis.
Is it diet or exercise that works best for prevention?
In a recent large three-arm randomised trial in overweight and obese adults with knee OA , after 18 months, participants in the diet + exercise and diet groups had more weight loss and greater reductions in interleukin-6 levels than those in the exercise group while those in the diet group had greater reductions in knee compressive force than those in the exercise group. This suggests that the combination or diet is more important than exercise alone.
Injury
There is substantial evidence that past knee injury is associated with knee OA and this relationship is likely to be causal. In a meta-analysis , the odds ratio (OR) for knee OA with past knee injury was 4.2 rising to 5.95 for ligament and meniscus injury. In early life, injury appears less harmful with a relative risk (RR) of 2.95 in childhood and adolescence compared to around 5 in adulthood suggesting there is more potential for healing in children . It seems likely, despite the absence of evidence, that injury prevention will decrease the risk of knee OA. There are less data for other sites but digital fracture increases the risk of hand OA .
Vitamin D
A recent review of vitamin D and OA identified two RCTs and 13 observational studies . The RCTs were only reported in abstract form and showed inconsistent results, most likely due to variations in their study design. There was insufficient or limited evidence for associations between 25-(OH)D and hand or hip OA. For knee radiographic OA as assessed by the Kellgren and Lawrence (KL) score, there was moderate evidence showing that low levels of 25-(OH)D were associated with increased progression of radiographic OA. Strong evidence for an association between 25-(OH)D and cartilage loss was apparent when joint space narrowing and changes in cartilage volume were considered collectively as cartilage loss. Since this review was published there has been an observational study showing that 25OHD levels in the moderate deficiency range were associated with the development of knee and hip pain over 5 years , implying the treatment of levels above this will not help pain. There has also been a randomised trial suggesting vitamin D supplementation did not help for symptoms of cartilage loss on MRI . However, this trial had a number of limitations as discussed in subsequent correspondence about this article. A further larger RCT done only in those with 25OHD levels below 50 nmol/l will be completed in mid-2014 and should give a more definitive answer .
Smoking
There is conflicting evidence regarding the role of cigarette smoking in the pathogenesis of OA (reviewed in Ref. ). While investigators in several studies have reported that smoking is not associated with development of radiographic OA, findings of most studies have suggested that smoking has a protective effect against prevalent and incident radiographic knee or hip OA. By contrast, there have been reports linking smoking with a higher prevalence of Heberden’s nodes, more severe spinal osteophytosis and incident knee pain. We reported that there was gene environment interaction for tibial cartilage loss using a prospective design, that is, those with a family history of knee joint replacement had higher cartilage loss if they smoked . By contrast, femoral cartilage loss was greater in smokers in the same study regardless of family history . Amin et al. reported similar findings for knee cartilage focal loss in men .
- 1.
Lifestyle factors are important in OA therapy and prevention but are rarely considered by practitioners
- 2.
Both prevention of weight gain and weight reduction appear important for symptoms
- 3.
Physical activity is important for symptoms but may speed up structural change in those with pre-existing OA
- 4.
Research on other factors is not yet at a point where recommendations can be made.
Osteoarthritis
OA is the most common form of arthritis and is increasing markedly due to an ageing population. It is characterised by a gradual loss of articular cartilage and changes to other joint structures (i.e., subchondral bone, ligaments, meniscus and synovium) leading, eventually, to total joint replacement. It is the most common joint disorder worldwide and is one of the most frequent causes of pain, loss of function and disability in Western populations . While genetic factors have been implicated in OA, few genes have been independently replicated. Lifestyle factors are also important and this article will discuss obesity (and diet), exercise, injury, vitamin D and smoking.
Obesity and diet
In younger life, knee pain is common in obese children, and is associated with increased weight/body mass index (BMI) . Furthermore, childhood BMI is associated with the presence of knee pain on walking in adults (mean age 36 years) and that being overweight in childhood is associated with higher levels of pain on walking in both normal and overweight adults . Moreover, BMI from age 11 years onwards has been shown to be associated with knee pain at age 45 years .
In adult life, obesity (however assessed) is an independent determinant of pain (knee, back and feet) and this is specifically for fat mass (as lean mass appears protective). It is strongly associated with radiographic change in the knee but less so in the hip and hand. It is also associated with structural changes such increased bone area and subchondral bone mineral density (BMD), bone marrow lesions, cartilage defects and meniscal tear/extrusion as well as cartilage glycosylated aminoglycan (GAG) content. The association with cartilage loss on magnetic resonance imaging (MRI) is less consistent in that it predicts cartilage loss in some studies and the effect on cartilage is at least partially mediated by leptin so is not solely due to loading.
In clinical trials, weight reduction is modestly but significantly effective for pain . Surprisingly, weight gain is more strongly associated with worsening pain than weight loss is with decreasing pain . There are less data for structure but observational studies suggest weight loss will improve cartilage defects and GAG content .
In terms of specific nutrients, there are a range of nutritional risk factors , though there are varying degrees of evidence to support them. These include dietary magnesium intake, energy, carbohydrate and sugar intake , vitamin C intake and vitamin K intake as well as dietary fatty acids. Serum high-density lipoprotein (HDL) cholesterol may be protective for bone marrow lesions in the knee .
Exercise
Longitudinal data in children suggest that children with an above median average sport intensity gained about twice as much tibial cartilage volume as those below median. This was consistent with the cross-sectional data where physical activity was a significant explanatory factor for patella, and medial and lateral tibial cartilage volume ( R 2 7–14% depending on site, all p < 0.05). In younger adult life, cartilage GAG content changes rapidly with exercise .
In later life, aerobic exercise is one of the major conservative management techniques for the treatment of knee OA. Several meta-analyses have definitively concluded there is strong evidence that aerobic exercise improves knee pain and function. As a result, aerobic exercise is currently recommended by all international guidelines for the treatment of knee OA. In terms of specific exercise, a systematic review by Roddy et al. compared the efficacy of aerobic walking and strengthening exercises in patients with knee OA. Out of the 13 randomised controlled trials (RCTs) included, four focused on aerobic walking. They reported a benefit from aerobic walking in reducing pain and self-reported disability.
Only one RCT has been conducted which examined the effect of exercise (specifically lower limb strength training) on disease progression as its primary outcome, assessed as joint space narrowing measured by radiography. Although, the strength training programme did not actually increase strength, the results showed a non-significant trend towards a beneficial effect (joint space narrowing occurred less often in the strength training group than in the ROM group (18% vs. 28%, p = 0.094). Surprisingly, in subgroup analysis, strength training significantly increased the rate of joint space narrowing in those participants with normal radiographs at baseline (34% vs. 19%, p = 0.038) suggesting it may be harmful. Other clinical trials examining either exercise and/or strength training for knee OA have shown no effect on structural progression. MRI has revolutionised the understanding of OA. The few studies which have employed this technology to assess the effect of exercise on knee structure have been observational in nature and show conflicting results. However, the majority have been cross-sectional and prone to bias. In midlife, longitudinal data suggest strenuous exercise may protective against cartilage defect progression . In later life, a recent paper which used pedometers to assess actual physical activity suggested that walking was deleteriously associated with knee structural change (including increases in BMLs, meniscal pathology and cartilage defects) over approximately 2.7 years, especially in those with pre-existing evidence of OA .
These results suggest that physical activity is beneficial for symptoms but may have varying effects on structure depending on age and stage of OA. Indeed, they may be harmful in those with established OA of the knee. Thus, there is a strong need for clinical trials to be performed to confirm or refute this hypothesis.
Is it diet or exercise that works best for prevention?
In a recent large three-arm randomised trial in overweight and obese adults with knee OA , after 18 months, participants in the diet + exercise and diet groups had more weight loss and greater reductions in interleukin-6 levels than those in the exercise group while those in the diet group had greater reductions in knee compressive force than those in the exercise group. This suggests that the combination or diet is more important than exercise alone.
Injury
There is substantial evidence that past knee injury is associated with knee OA and this relationship is likely to be causal. In a meta-analysis , the odds ratio (OR) for knee OA with past knee injury was 4.2 rising to 5.95 for ligament and meniscus injury. In early life, injury appears less harmful with a relative risk (RR) of 2.95 in childhood and adolescence compared to around 5 in adulthood suggesting there is more potential for healing in children . It seems likely, despite the absence of evidence, that injury prevention will decrease the risk of knee OA. There are less data for other sites but digital fracture increases the risk of hand OA .
Vitamin D
A recent review of vitamin D and OA identified two RCTs and 13 observational studies . The RCTs were only reported in abstract form and showed inconsistent results, most likely due to variations in their study design. There was insufficient or limited evidence for associations between 25-(OH)D and hand or hip OA. For knee radiographic OA as assessed by the Kellgren and Lawrence (KL) score, there was moderate evidence showing that low levels of 25-(OH)D were associated with increased progression of radiographic OA. Strong evidence for an association between 25-(OH)D and cartilage loss was apparent when joint space narrowing and changes in cartilage volume were considered collectively as cartilage loss. Since this review was published there has been an observational study showing that 25OHD levels in the moderate deficiency range were associated with the development of knee and hip pain over 5 years , implying the treatment of levels above this will not help pain. There has also been a randomised trial suggesting vitamin D supplementation did not help for symptoms of cartilage loss on MRI . However, this trial had a number of limitations as discussed in subsequent correspondence about this article. A further larger RCT done only in those with 25OHD levels below 50 nmol/l will be completed in mid-2014 and should give a more definitive answer .
Smoking
There is conflicting evidence regarding the role of cigarette smoking in the pathogenesis of OA (reviewed in Ref. ). While investigators in several studies have reported that smoking is not associated with development of radiographic OA, findings of most studies have suggested that smoking has a protective effect against prevalent and incident radiographic knee or hip OA. By contrast, there have been reports linking smoking with a higher prevalence of Heberden’s nodes, more severe spinal osteophytosis and incident knee pain. We reported that there was gene environment interaction for tibial cartilage loss using a prospective design, that is, those with a family history of knee joint replacement had higher cartilage loss if they smoked . By contrast, femoral cartilage loss was greater in smokers in the same study regardless of family history . Amin et al. reported similar findings for knee cartilage focal loss in men .
- 1.
Lifestyle factors are important in OA therapy and prevention but are rarely considered by practitioners
- 2.
Both prevention of weight gain and weight reduction appear important for symptoms
- 3.
Physical activity is important for symptoms but may speed up structural change in those with pre-existing OA
- 4.
Research on other factors is not yet at a point where recommendations can be made.
Osteoporosis in younger life
Osteoporosis is a major public health problem as the fragility fractures it causes, both in younger and later life, are common and cause substantial morbidity, mortality and economic costs. BMD is one of the major predictors of osteoporotic fractures . Suboptimal bone growth in childhood and adolescence and bone loss in adult life appear to both play significant roles in the development of osteoporosis. Premenopausal bone mass is at least as important as bone loss in the post-menopausal period for prediction of fracture . Furthermore, bone density is also a risk factor for fracture in children, and in premenopausal women . Furthermore, sustaining a fracture prior to menopause is associated with increased risk of subsequent osteoporotic fracture. In one study, a fracture sustained between age 20 and 50 years increases the risk of risk of fracture after age 50 by 74% .
Even small annual changes in the rate of acquisition (in childhood and early adult life) or rate of loss of bone are potentially important, as cumulative effects could have substantial long-term clinical and public health benefits. For example, modelling suggests if a very small annual decrease in age-related bone loss of 0.03% p.a. (from 0.25% p.a. loss to 0.22% p.a. loss) in the lumbar spine (LS) were to occur from age 30, this could delay the onset of osteoporosis by 2 years . Therefore, maximising bone mass throughout life has important potential benefits for the prevention of fracture throughout the lifespan.
Nutritional interventions for improving peak bone mass
A range of nutritional factors have been postulated to influence children’s bone development and affect peak bone mass, including maternal diet in utero, breast feeding, calcium and dairy intake, vitamin D, fruit and vegetable intake and possible adverse effects from high dietary sodium intake and intake of carbonated beverages. However, the evidence for most of these factors is limited, and in particular, RCTs testing the efficacy of optimizing most of these factors are lacking. This section provides an overview of the current evidence, with the emphasis placed on factors with the strongest evidence base.
While it is widely accepted that an adequate calcium intake in childhood is important for bone development, the evidence from observational and intervention studies are mixed and low calcium/dairy intakes may be related to fracture risk in childhood though again the evidence is not completely consistent . High levels of calcium intake for children are recommended in many developed countries. Current World Health Organisation recommendations based on North American and western European data are from 300 to 400 mg/day for infants, 400–700 mg/day for children and 1300 mg/day for adolescents) but these recommendations may not be applicable to other settings.
However, the usefulness of calcium supplements in children for improving bone outcomes is open to question. A meta-analysis of 19 RCTs in 2859 children reported that calcium supplementation had no effect on BMD at the femoral neck (FN) or LS), two clinically important sites for future osteoporotic fracture. A small effect on total body (TB) bone mineral content (BMC) did not persist after cessation of supplementation. A small persistent effect on upper-limb BMD was equivalent to a 1.7 percentage point greater increase in BMD in the supplemented compared to the control group, which might reduce the absolute risk of fracture at the peak of childhood fracture incidence by at most 0.2% per annum (p.a.) which is of marginal clinical or public health benefit. Furthermore, the data suggested that increasing the duration of supplementation did not lead to accumulation of greater effects, and that the effect size (ES) did not vary with baseline calcium intakes, down to a level of <600 mg/day. A subsequent RCT targeting children (mean age 12 years) with an habitual calcium intake <650 mg/day resulted in greater increases in TB BMC (2.3%) and total hip (TH) and LS BMD (2.5% and 2.2%, respectively) in children supplemented with an average of 555 mg calcium/day after 18 months but as in the meta-analysis, the effects did not persist once supplements ceased . This meta-analysis was restricted to placebo-controlled trials. As a result, some RCTs of dairy products were excluded but qualitatively, the results of those studies were similar, showing no or only small to moderate short-term effects which dissipated after supplementation ceased .
Vitamin D also has a widely accepted role in bone health. The link between vitamin D deficiency and rickets is well understood, though rickets may be caused by both very low calcium intakes as well as vitamin D deficiency, the former being particularly important in developing countries. In developed countries, rickets are most often seen in groups at high risk of moderate to severe vitamin D deficiency . However, subclinical vitamin D deficiency can adversely affect bone mineralisation and potentially could reduce acquisition of bone mass resulting in lower peak bone mass.
In a meta-analysis of six RCTs (343 participants receiving placebo and 541 receiving vitamin D) , overall, vitamin D supplementation had no statistically significant effects on TB BMC, hip BMD or forearm BMD and all ESs were small (standardised mean difference (SMD) ≤0.10). There was a trend to an effect on LS BMD, but again the ES was small (SMD +0.15, (95% confidence interval (CI) −0.01 to +0.31), p = 0.07). In subgroup analysis by baseline mean vitamin D level in each study, there were significant effects on TB BMC (SMD + 0.21, (95% CI 0.01–0.26)) and LS BMD (SMD +0.31 (95% CI 0.00–0.61) in studies where baseline serum vitamin D level was low (mean < 35 nmol/L). These equate roughly to a 2.6% and 1.7% percentage point greater increase from baseline respectively in supplemented groups, but it is not known if these effects will accumulate with ongoing supplementation. Nevertheless, this suggests that at least in vitamin D-deficient children, supplementation could result in clinically useful improvements in bone density. This is particularly the case if future trials can demonstrate that effects accumulate with ongoing supplementation.
Evidence for the impact of other dietary factors on bone development in children is predominantly limited to observational studies. Fruit and vegetable intake is postulated to have beneficial effects on bone through mechanisms including the induction of a mild metabolic alkalosis, vitamin K, vitamin C, antioxidants and phytoestrogens, though a single RCT suggests that phytoestrogens alone have little effect on bone turnover in children . Observational data support a positive relationship between fruit and vegetable intake and bone outcomes in children but this is yet to be tested in intervention studies. It has been suggested that high salt intake may be detrimental due effects on urinary calcium excretion. However, in the few studies assessing bone density, there have been no associations with urinary sodium excretion demonstrated . Urinary sodium has been shown to be associated with high bone turnover in adolescent boys. Initially, more longitudinal studies are needed to determine if sodium intake does in fact have a clinically important effect on bone in children. Carbonated beverage and cola consumption has been linked with decreased BMD in girls but not boys and with increased fracture risk in both sexes . This may be in part due to milk replacement. However, there is also likely to be an independent effect as low milk intake and a higher consumption of carbonated beverages have been shown to be independent fracture risk factors in children with recurrent fractures and two studies have shown that associations between fracture risk, peripheral quantitative computed tomography (pQCT) measures and cola drinks persist after adjustment for milk intake.
The influence of nutritional factors on bone acquisition in children may begin in utero. Again, the factor which has received the most investigation to date is calcium, but RCT evidence is inconsistent . It therefore remains unclear whether improving maternal calcium intake in pregnancy is beneficial for in utero bone development. Zinc supplements have also been tested in a single RCT in pregnant women from a disadvantaged area in a developing country, and these resulted in increased foetal femur diaphysis length. The applicability of this finding to other settings is not known. .
RCTs of other supplements in pregnancy with childhood bone outcomes are lacking to date, though observational evidence suggests possible roles for factors including vitamin D , folate, magnesium, phosphorus, potassium and protein and maternal fat intake. Dietary patterns, rather than individual components of the diet may also be important, for example, a maternal dietary pattern of a high intake of fruit, vegetable and wholemeal bread, pasta and rice and low intake of processed foods was associated with higher TB and lumbar BMC and BMD . Though limited, these data support the need for further research into nutritional interventions in pregnancy.
Human milk-fed infants generally have lower bone accretion compared to formula-fed infants but this does not appear to result in long-term deficits, as shown by an RCT of infant feeding comparing two different formulae and breastfeeding, in which initial differences in BMC accretion did not persist past 12 months of age . Importantly, breast feeding was protective for childhood fractures in a longitudinal study of prepubertal children and in a case control study of children aged 4–15 years, though this was not observed in a longitudinal study of fracture risk from birth to 18 years . There is a risk of rickets in breastfed infants of women who are at high risk of moderate to severe vitamin D deficiency, and at least in developed countries intervention is required either in the form of screening for and correcting significant vitamin D deficiency or by routine supplementation of breastfeeding infants at high clinical risk of deficiency .
Exercise
Weight-bearing activities are an established way to improve bone mineral acquisition in children. A recent meta-analysis of 27 RCTs of weight-bearing activities (defined as force-generating exercises placing higher mechanical stress on the human skeleton than daily living, e.g., jump-training or resistance training programme) in 2985 children (59% female) reported small effects overall for BMC (ES 0.17, 95% CI (0.05–0.29) and areal BMD (ES 0.26 (95% CI (0.02–0.49). More than a third of the observed variance of the studies reporting BMC as an outcome could be explained by differences in habitual daily calcium intake ( β = 0.001 per mg calcium/day, p < 0.001) and baseline pubertal status ( β = −0.157 for intra pubertal/postpubertal vs. prepubertal participants, p < 0.001) suggesting greater benefits in prepubertal children and in those with higher calcium intakes . It is less clear for how long benefits from exercise or higher levels of physical activity in early life persist into later life. Cross-sectional data in current and former elite soccer players suggest that BMD benefits are slowly lost over time (taking more than 30 years for benefits to disappear) , but other studies suggest that benefits can be maintained . Importantly, longitudinal studies suggest that physical fitness measures and being an elite athlete in young adulthood can result in long-term reductions in fracture risk.
Smoking and alcohol
Data on the potential impacts of childhood smoking and alcohol intake on bone health are lacking. A single longitudinal study reported lower rates of LS and TH BMD accrual in adolescents (from ages 13–19) who were higher-frequency smokers but no associations of alcohol intake on any bone outcome and another reported a 43% increased risk of fracture in teenagers who regularly smoked . Evidence on the long-term effects of maternal smoking in pregnancy is conflicting, with one study reporting a negative association between maternal smoking and LS and TH BMD at age 8 but not age 16 but another that both maternal and paternal smoking were associated with increased TB less head BMC and spine BMD in girls not boys at age 10 years, suggesting that the effects were attributable to shared familial characteristics mechanisms . Effects of maternal alcohol consumption on bone density in children are unknown.
Osteoporosis in premenopausal women
Nutrition
The range of potential nutritional interventions to improve peak bone mass and/or slow age-related bone loss in premenopausal women is similar to that in children, but the evidence around these lifestyle modifications in premenopausal women is sparse, not definitive and intervention studies are confined to trials of calcium, vitamin D and one trial of a behavioural intervention. In the absence of such studies, the potential benefits of other nutritional interventions such as improving levels of fruit and vegetable intake, salt intake, intake of animal and vegetable proteins, and the calcium/phosphorus ratio in the diet will not be discussed in this article.
A meta-analysis of four trials of increasing calcium intake in premenopausal women, either by supplementation or by dietary advice reported an ES of 1.3% per year across a combination of sites . The studies were small (total number of participants <200), and results at individual sites were inconsistent , so this evidence cannot be considered definitive but supports the potential role of calcium intake for improving bone density in younger women.
As in children, vitamin D deficiency in premenopausal women is common, but even so, the potential for improved vitamin D levels to improve bone health in these younger women has been inadequately examined. In one RCT, vitamin D 800 IU and 2000 mg calcium were given in combination as a daily dose to female Navy recruits aged 17–35 years during 8 weeks of training, resulting in 20% reduction in stress fractures . Bone density was not measured. The only RCT to measure bone density was undertaken in a vitamin D-deficient immigrant population in Denmark , in which 89 women (age range 18–52 years) were given placebo, 10 or 20 μg of vitamin D 3 daily. There was a high drop-out rate (27 over 12 months). No differences between treatment groups for LS BMD, BMC or bone area were observed, but unexpectedly reduced whole body BMD was reported in the 20-μg supplement group. However, the small sample size and high loss to follow-up means that this result should be interpreted cautiously. The role of vitamin D in bone health in premenopausal women remains unclear.
Exercise
A meta-analysis examining the effects of impact exercise on bone density in premenopausal women identified nine studies with 281 exercise and 240 control participants . Each study used an exercise protocol that included ground reaction force generating impact activity such as running or jumping-type movements and the intervention duration ranged from 6 to 2 months. Overall, the effect of impact exercise on the LS BMD was 0.006 g/cm 2 (95% CI 0.002–0.010) and was 0.012 g/cm 2 (95% CI 0.005–0.020) at the FN, which is modest.
Smoking and alcohol
The effect of smoking and alcohol on bone health in young women is unclear. In a study of women aged 20–29, there was no association between smoking and BMD at either the FN or LS but in women of mean age 33 years, bone density was lower at both these sites . Alcohol intake was not associated with femoral BMD in premenopausal women in one cross-sectional study .
Lifestyle modification programmes
Few lifestyle modification interventions in premenopausal women have measured bone density outcomes . Feedback of fracture risk based on bone density with at least minimal education material can lead to improved FN (0.9% p.a.) but not LS BMD in premenopausal women. The improvement observed in this study was mediated through behaviour changes, specifically increased use of calcium supplements (1.3% p.a.) and increases in self-reported physical activity (0.7% p.a.) .
Men
In men, as in women, early intervention to improve or maintain BMD appears an important approach to prevent osteoporotic fractures in later life. Despite this, in comparison with women, ways to improve peak bone mass or slow age-related bone loss are virtually non-existent in young adult males. A single RCT of brisk walking in exclusively middle-aged men (aged 53–62 years) failed to demonstrate any beneficial effects on bone density at either the LS or proximal femur . There is relatively consistent observational evidence that smoking is detrimental for bone density in younger men. For example, unlike in women in the same study, in men aged 20–29, FN BMD was about 6.8% lower in smokers than non-smokers though there was no association between smoking and LS BMD . In young men, it may be that alcohol consumption is associated with higher BMD, but to date this is based on limited cross-sectional data .
- •
Calcium supplementation in healthy children has no effect on bone density at the hip or LS and only a small effect at the upper limb which is unlikely to result in a clinically important decrease in risk of fracture either in childhood or in later life. Thus, the evidence does not support their routine use in healthy children.
- •
Vitamin D supplements may provide clinically useful improvements in bone density in vitamin D-deficient children if the small observed improvements in bone density observed accumulate with ongoing supplementation, but this remains to be proven. Otherwise, the use of vitamin D supplements in healthy children provides no benefit to bone density.
- •
Breast feeding remains the optimal choice for infant nutrition for bone health, but vitamin D status should be considered in groups at high risk of vitamin D deficiency.
- •
Weight-bearing activities are effective at improving BMC and aBMD, particularly in prepubertal children, and may be more effective with increased calcium intake.
- •
Evidence is insufficient to recommend the use of calcium or vitamin D supplements to improve bone density in young men and women.
- •
Impact exercise can be recommended to improve LS and FN BMD in premenopausal women, though effects are modest.
- •
There is insufficient evidence to provide firm recommendations for lifestyle modifications for young and middle-aged men.
- •
While evidence is limited, it nonetheless would seem prudent to suggest minimising smoking and alcohol intake in children, pregnant women and young men and women for both bone and for general health.
Related posts:
Musculoskeletal health, frailty and functional decline
A time for action: Opportunities for preventing the growing burden and disability from musculoskeletal conditions in low- and middle-income countries
Determinants of MSK health and disability – Social determinants of inequities in MSK health
Determinants of MSK health and disability: Lifestyle determinants of symptomatic osteoarthritis
Corrigendum to “Granulomatous inflammation: the overlap of immune deficiency and inflammation” [Best Pract Res Clin Rheumatol 28 (2014) 191–212]
A public health approach to musculoskeletal health