There have been substantial changes in the nutritional management of many diseases in the last 20 years, which have been accompanied by a growing recognition of its importance. Many of the changes in clinical nutrition have been associated with the introduction of standards, clinical audit and the implementation of evidence-based practice, which has led to a re-evaluation of some established dietary interventions using a hierarchy-of-evidence approach. Although there are few randomised controlled trials on which to base such work, the examination of other, often less-robust, evidence has led to some traditional dietary interventions being modified. Examples in gastroenterology include the use of low-fat diets in gall bladder disease and the restriction of protein in hepatic encephalopathy, where the current evidence suggests that neither should be used routinely in clinical practice. Where therapeutic dietary restrictions are required, as with low-Na diets in ascites, there is very little information on how these restrictions influence total nutrient intake and, if intake is impaired, how the detrimental effects of an inadequate intake should be balanced with the therapeutic effects of restriction. Studies are required to ensure that nutritional interventions are not only effective but also free from undesirable side effects. The mode and timing of the delivery of nutritional support has also been re-evaluated and the benefits of early enteral feeding have been recognised. The delivery of dietary advice is a new area that is being considered, with practitioners in clinical nutrition using behaviour-change skills to facilitate optimum nutrition rather than simply providing patients with advice. For such developments to continue in clinical nutrition it is essential that all practice should be systematically evaluated and, where necessary, modified in the light of sound current research findings, and that gaps in our present knowledge base are identified and addressed.

As a clinically-effective nutritionist faced with needing to find the answer to a clinical question quickly it is necessary to search the best evidence efficiently. The requirements are to: (1) avoid having lots of papers to read; (2) be able to access this information at one's place of work; (3) restrict reading to trials and systematic reviews of trials. If a decision has to be made while on the ward, the best resource is probably an easy-to-use book, Clinical Evidence. This resource is also available on-line, but there may not be access to the Internet on the ward. If a little more time is available, and access to the Internet, the following plan is suggested: (1) work out what the question is and highlight the search terms; (2) using the best search engine available, search MEDLINE from 1990 for titles and abstracts of papers containing the search terms, this is the search strategy (limit the search by publication type by requesting randomised controlled trials only, English language only and human only); (3) if more or less than approximately ten hits are obtained, alter the limits of the search (not the search strategy); (4) read the abstract, or full paper where available, of the relevant hits and appraise this evidence.

Common to both acute and chronic disease are disturbances in energy homeostasis, which are evidenced by quantitative and qualitative changes in dietary intake and increased energy expenditure. Negative energy balance results in loss of fat and lean tissue. The management of patients with metabolically-active disease appears to be simple; it would involve the provision of sufficient energy to promote tissue accretion. However, two fundamental issues serve to prevent nutritional demands in disease being met. The determination of appropriate energy requirements relies on predictive formulae. While equations have been developed for critically-ill populations, accurate energy prescribing in the acute setting is uncommon. Only 25–32% of the patients have energy intakes within 10% of their requirements. Clearly, the variation in energy expenditure has led to difficulties in accurately defining the energy needs of the individual. Second, the acute inflammatory response initiated by the host can have profound effects on ingestive behaviour, but this area is poorly understood by practising clinicians. For example, nutritional targets have been set for specific disease states, i.e. pancreatitis 105–147 KJ (25–35 kcal)/kg; chronic liver disease 147–168 kj (35–40 kcl)/kg, but given the alterations in gut physiology that accompany the acute-phase response, targets are unlikely to be met. In cancer cachexia attenuation of the inflammatory response using eicosapentaenoic acid results in improved nutritional intake and status. This strategy poses an attractive proposition in the quest to define nutritional support as a clinically-effective treatment modality in other disorders.

Weight loss is a frequent complication in patients with chronic obstructive pulmonary disease (COPD) and is a determining factor for functional capacity, health status and mortality. Weight loss in COPD is a consequence of an inbalance between increased energy requirements and dietary intake. Both metabolic and mechanical inefficiency may contribute to elevated energy expenditure during physical activity, while systemic inflammation has been associated with hypermetabolism at rest. Disease-specific symptoms and systemic inflammation may impair appetite and dietary intake. Altered intermediary metabolism may cause disproportionate wasting of fat-free mass in some patients. A combination of nutritional support and exercise as an anabolic stimulus appears to be the best approach to obtaining marked functional improvement. Patients responding to this treatment even demonstrated a decreased mortality. The effectiveness of anti-catabolic modulation requires further investigation.

Cystic fibrosis (CF) is a genetically-inherited disorder that results in energy imbalance. Undernutrition is common in children with CF and associated with poor health outcomes. To ensure optimal growth and nutrition, children with CF are recommended to consume 120–150% of the recommended daily allowance (RDA) for energy, but most studies show they typically are only able to achieve 100% of the RDA. While biological factors clearly contribute to poor dietary adherence, recent studies have documented behavioural and environmental barriers to adherence that includes parent-child interaction at mealtimes. While not ‘abnormal’, parent behaviours such as paying increased attention to the child in the form of coaxing, commanding and feeding when the child is engaged in behaviours incompatible with eating (food refusal, talking, leaving the meal) may serve to reinforce these child non-eating behaviours. Thus, dietary counselling alone, albeit necessary, is typically insufficient because of failure to specifically address these behavioural and environmental barriers to dietary treatment. Behavioural intervention that targets both nutrition education and behavioural management has been found to be effective in achieving an average increased energy intake of 4200 kJ (1000)kcal/d and weight gain of 1·48 kg over 9 weeks in children with CF. This intervention utilizes self-monitoring, goal setting and shaping to structure the delivery of treatment. It also teaches parents to utilize child behaviour-management techniques to motivate children to increase their energy intake. These behavioural strategies include differential attention (praising and ignoring), contingency management and behavioural contracting. The potential application of these techniques to dietary counselling is suggested.

Growth failure and poor nutritional status are features of children with chronic disease. Oral protein–energy supplements are one of a number of interventions provided with the aim of improving nutritional status in these children. The present paper describes a Cochrane systematic review assessing the efficacy of these products in children with chronic disease. The objective was to examine the evidence that in children with chronic disease oral protein–energy supplements alter nutrient intake, nutritional indices, survival and quality of life. All randomised controlled trials of the use of oral protein–energy supplements in children with chronic disease were identified through searching electronic databases and hand searching the abstract books of nutrition conferences. Studies identified were independently assessed for eligibility and methodological quality, and data on outcomes of interest were combined in a meta-analysis where possible. Two trials were eligible for inclusion in the review, both of which were undertaken with children with cystic fibrosis. No statistical differences could be found between treatment and control groups when data from both studies were combined. Oral protein–energy supplements are widely used to improve the nutritional status of children with chronic disease. No conclusions can be drawn on the efficacy of these products based on the limited data available. Further randomised controlled trials are required to investigate the use of these products in children with chronic disease. Until further data are available, these products should be used with caution.

Malnutrition has long been recognised as a risk factor for post-operative morbidity and mortality. Traditional metabolic and nutritional care of patients undergoing major elective surgery has emphasised pre-operative fasting and re-introduction of oral nutrition 3–5 d after surgery. Attempts to attenuate the consequent nutritional deficit and to influence post-operative morbidity and mortality have included parenteral, enteral and oral sip feeding. Recent studies have emphasised that an enhanced rate of recovery can be achieved by a multi-modal approach focused on modulating the metabolic status of the patient before (e.g. carbohydrate and fluid loading), during (e.g. epidural anaesthesia) and after (e.g. early oral feeding) surgery. Using such an approach preliminary results on patients undergoing elective colo-rectal surgery indicate a significant reduction in hospital stay (traditional care, n 48, median stay 10 d v. enhanced recovery programme, n 33, median stay 7d; P<0·01) can be achieved. Such findings emphasise the potential role of multi-modal care programmes in the promotion of early recovery from major surgical trauma.

Vitamin D is produced endogenously when the skin is exposed to sunlight and can be obtained exogenously from a few natural food sources, from food fortification and from supplements. Generally, vitamin D intake is low ≤2–3 μg/d in Europe. Casual exposure to sunlight is thought to provide most of the vitamin D requirement of the human population. However, skin synthesis of vitamin D may not compensate for the low nutritional intake in Europe, even in countries with high supplies from food fortification and supplements. For assessment of vitamin D nutritional status the concentration of 25-hydroxyvitamin D (25(OH)D) in serum is considered to be an accurate integrative measure reflecting an individual's dietary intake and cutaneous production. A substantial percentage of the elderly and adolescents in Europe have a low concentration of 25(OH)D; in the elderly this percentage ranges from approximately 10 in the Nordic countries to approximately 40 in France. Low vitamin D status seems to be aggravated by disease and immobility, and by a low frequency of supplement use.

Prolonged vitamin D deficiency resulting in rickets is seen mainly during rapid growth. A distinct age distribution has been observed in the Copenhagen area where all registered hospital cases of rickets were either infants and toddlers or adolescents from immigrant families. Growth retardation was only present in the infant and toddler group. A state of deficiency occurs months before rickets is obvious on physical examination. Growth failure, lethargy and irritability may be early signs of vitamin D deficiency. Mothers with low vitamin D status give birth to children with low vitamin D status and increased risk of rickets. Reports showing increasing rates of rickets due to insufficient sunlight exposure and inadequate vitamin D intake are cause for serious concern. Many countries (including the USA from 2003) recommend vitamin D supplementation during infancy to avoid rickets resulting from the low vitamin D content of human milk. Without fortification only certain foods such as fatty fish contain more than low amounts of vitamin D, and many children will depend entirely on sun exposure to obtain sufficient vitamin D. The skin has a high capacity to synthesize vitamin D, but if sun exposure is low vitamin D production is insufficient, especially in dark-skinned infants. The use of serum 25-hydroxyvitamin D to evaluate vitamin D status before development of rickets would be helpful; however, there is no agreement on cut-off levels for deficiency and insufficiency. Furthermore, it is not known how marginal vitamin D insufficiency affects children's bones in the long term.

Physical activity (PA) is a popular therapy for the prevention and treatment of bone loss and osteoporosis because it has no adverse side effects, it is low cost, and it confers additional benefits such as postural stability and fall prevention. Bone mass is regulated by mechanical loading, and is limited but not controlled by diet. The mechanism by which strain thresholds turn bone remodelling ‘on’ and ‘off ’ is known as the mechanostat theory. Research in animals has shown that optimal strains are dynamic, with a high change rate, an unusual distribution and a high magnitude of strain, but the results of randomized controlled trials in human subjects have been somewhat equivocal. In the absence of weight-bearing activity nutritional or endocrine interventions cannot maintain bone mass. Biochemical markers of bone turnover predict bone mass changes, and findings from our research group and others have shown that both acute and chronic exercise can reduce bone resorption. Similarly, Ca intervention studies have shown that supplementation can reduce bone resorption. Several recent meta-analytical reviews concur that changes in bone mass with exercise are typically 2–3%. Some of these studies suggest that Ca intake may influence the impact of PA on bone, with greater effects in Ca-replete subjects. Comparative studies between Asian (high PA, low Ca intake) and US populations (low PA, high Ca intake) suggest that PA may permit an adaptation to low Ca intakes. Whether Ca and PA interact synergistically is one of the most important questions unanswered in the area of lifestyle-related bone health research.

Vitamin K, originally recognised as a factor required for normal blood coagulation, is now receiving more attention in relation to its role in bone metabolism. Vitamin K is a coenzyme for glutamate carboxylase, which mediates the conversion of glutamate to γ-carboxyglutamate (Gla). Gla residues attract Ca2+ and incorporate these ions into the hydroxyapatite crystals. There are at least three Gla proteins associated with bone tissue, of which osteocalcin is the most abundant and best known. Osteocalcin is the major non-collagenous protein incorporated in bone matrix during bone formation. However, approximately 30% of the newly-produced osteocalcin stays in the circulation where it may be used as an indicator of bone formation. Vitamin K deficiency results in an increase in undercarboxylated osteocalcin, a protein with low biological activity. Several studies have demonstrated that low dietary vitamin K intake is associated with low bone mineral density or increased fractures. Additionally, vitamin K supplementation has been shown to reduce undercarboxylated osteocalcin and improve the bone turnover profile. Some studies have indicated that high levels of undercarboxylated osteocalcin (as a result of low vitamin K intake?) are associated with low bone mineral density and increased hip fracture. The current dietary recommendation for vitamin K is 1 μ/kg body weight per d, based on saturation of the coagulation system. The daily dietary vitamin K intake is estimated to be in the range 124–375 μg/d in a European population. Thus, a deficiency based on the hepatic coagulation system would be unusual, but recent data suggest that the requirement in relation to bone health might be higher.

Severe vitamin A toxicity is known to have adverse effects on skeletal health. Studies involving animal models and case reports have documented that hypervitaminosis A is associated with bone resorption, hypercalcaemia and bone abnormalities. More recently, some epidemiological studies have suggested that high habitual intake of vitamin A could contribute to low bone mineral content and fracture risk. The evidence relating to the possible deleterious role of vitamin A in bone health is of variable quality and is potentially confounded by collinearity of nutrient intake and difficulties in assessing vitamin A exposure. Furthermore, because intake of vitamin A varies between studies it is not possible to define an intake threshold associated with harm.

Approximately 99% of body Ca is found in bone, where it serves a key structural role as a component of hydroxyapatite. Dietary requirements for Ca are determined by the needs for bone development and maintenance, which vary throughout the life stage, with greater needs during the periods of rapid growth in childhood and adolescence, during pregnancy and lactation, and in later life. There is considerable disagreement between expert groups on the daily Ca intake levels that should be recommended, reflecting the uncertainty in the data for establishing Ca requirements. Inadequate dietary Ca in early life impairs bone development, and Ca supplementation of the usual diet for periods of ≤3 years has been shown to enhance bone mineral status in children and adolescents. However, it is unclear whether this benefit is long term, leading to the optimisation of peak bone mass in early adulthood. In later years inadequate dietary Ca accelerates bone loss and may contribute to osteoporosis. Ca supplementation of the usual diet in post-menopausal women and older men has been shown to reduce the rate of loss of bone mineral density at a number of sites over periods of 1–2 years. However, the extent to which this outcome reduces fracture risk needs to be determined. Even allowing for disagreements on recommended intakes, evidence indicates that dietary Ca intake is inadequate for maintenance of bone health in a substantial proportion of some population groups, particularly adolescent girls and older women.

Na-induced calciuria has been well documented and provides a physiological basis for the proposed role of dietary Na (or salt) as a risk factor for osteoporosis. However, the evidence is based primarily on acute salt-loading studies, and there are insufficient data on the effects of high salt intake on net Ca retention to predict long-term effects on bone health. Results of investigations on salt and bone turnover, as assessed by bone biomarkers, are inconsistent, but the large variations in inter-individual response to acute and chronic Na loading may be related to salt sensitivity. Results of cross-sectional and prospective investigations on high salt intake and long-term bone health are inconclusive, probably reflecting the difficulty of conducting such studies in free-living populations. However, the mean urinary Ca loss of 1 mmol/100 mmol Na suggests that chronic changes in salt intake may have large effects on Ca and bone balance, especially in individuals with a reduced capacity to compensate for Na-induced Ca loss. Investigating the relationship between salt intake and bone health requires a greater focus on whole diets (including components such as K, Mg, P and protein), reliable measures of salt intake, appropriate bone health outcome measures, and improved subject characterisation (e.g. salt sensitivity). The reasons for inter-individual variability should be explored using post-genomic techniques.

The effects of dietary protein on bone health are paradoxical and need to be considered in context of the age, health status and usual diet of the population. Over the last 80 years numerous studies have demonstrated that a high protein intake increases urinary Ca excretion and that on average 1 mg Ca is lost in urine for every 1 g rise in dietary protein. This relationship is primarily attributable to metabolism of S amino acids present in animal and some vegetable proteins, resulting in a greater acid load and buffering response by the skeleton. However, many of these early studies that demonstrated the calciuric effects of protein were limited by low subject numbers, methodological errors and the use of high doses of purified forms of protein. Furthermore, the cross-cultural and population studies that showed a positive association between animal-protein intake and hip fracture risk did not consider other lifestyle or dietary factors that may protect or increase the risk of fracture. The effects of protein on bone appear to be biphasic and may also depend on intake of Ca- and alkali-rich foods, such as fruit and vegetables. At low protein intakes insulin-like growth factor production is reduced, which in turn has a negative effect on Ca and phosphate metabolism, bone formation and muscle cell synthesis. Although growth and skeletal development is impaired at very low protein intakes, it is not known whether variations in protein quality affect the achievement of optimal peak bone mass in adolescents and young adults. Prospective studies in the elderly in the USA have shown that the greatest bone losses occur in elderly men and women with an average protein intake of 16–50 g/d. Although a low protein intake may be indicative of a generally poorer diet and state of health, there is a need to evaluate whether there is a lower threshold for protein intake in the elderly in Europe that may result in increased bone loss and risk of osteoporotic fracture.

The use of dietary phyto-oestrogens as a possible option for the prevention of osteoporosis has raised considerable interest because of the increased concern about the risks associated with the use of hormone-replacement therapy. However, the evidence in support of a bone-sparing effect in post-menopausal women is still not sufficiently convincing. Most studies have been performed on soyabean isoflavones (genistein and daidzein), either in the purified form or as a soyabean-based product or extract. In vitro studies using primary cell cultures or stabilised cell lines indicate that treatment with genistein may lead to a reduction in bone resorption, but effects on bone formation have also been shown. Investigations using animal models have provided convincing evidence of major improvements in bone mass or bone turnover following soyabean feeding. Cross-sectional observations in South-East Asian populations with moderately high intakes of soyabean isoflavones (50 mg/d) have shown that women in the high quartile of intake have higher bone mineral density (BMD) and reduced bone turnover, an effect that has not been shown in populations with low average intakes. Human trials have given an indication of a possible effect on lumbar spine BMD, although they have been either short term (<6 months) or methodologically weak. Unresolved issues are: the optimal dose compatible with safety; the individual differences in response that can be related to diet and genotypes; the duration of exposure. Furthermore, there needs to be an evaluation of the relative biological effects of phyto-oestrogens other than isoflavones (lignans, resorcylic acid lactones, flavanols, coumestans) that are also present in European diets.

These famous words by Mencken in the early 20th century about the meaning of life and death, may also apply to the struggle of the healthy skeleton against the deleterious effects of retained acid!’ ( Kraut & Coburn, 1994). The health-related benefit of a high consumption of fruit and vegetables and the influence of this food group on a variety of diseases has been gaining increasing prominence in the literature over a number of years. Of considerable interest to the osteoporosis field is the role that bone plays in acid–base balance. Natural, pathological and experimental states of acid loading and acidosis have been associated with hypercalciuria and negative Ca balance, and more recently the detrimental effects of ‘acid’ from the diet on bone mineral have been demonstrated. Suprisingly, consideration of the skeleton as a source of ‘buffer’ contributing to both the preservation of the body's pH and defence of the system against acid–base disorders has been ongoing for over three decades. However, it is only more recently that the possibility of a positive link between a high consumption of fruit and vegetables and indices of bone health has been more fully explored. A number of population-based studies published in the last decade have demonstrated a beneficial effect of fruit and vegetable and K intake on axial and peripheral bone mass and bone metabolism in men and women across the age-ranges. Further support for a positive link between fruit and vegetable intake and bone health can be found in the results of the Dietary Approaches to Stopping Hypertension (DASH) and DASH-Sodium intervention trials. There is now an urgent requirement for the implementation of: (1) fruit and vegetable and alkali administration–bone health intervention trials, including fracture risk as an end point; (2) reanalysis of existing dietary–bone mass and metabolism datasets to look specifically at the impact of dietary ‘acidity’ on the skeleton.

There is compelling evidence to suggest that both the development of bone to peak bone mass at maturity and subsequent loss depend on the interaction between genetic, hormonal, environmental and nutritional factors. The major part (≤80%) of the age-specific variation in bone turnover and bone density is genetically determined. However, the notion of genetic determinant is of little value unless the specific genes that are involved can be identified. Most work in this area of osteoporosis research has focused on the candidate gene approach, which has identified several candidate genes for osteoporosis, including genes encoding the vitamin D receptor (VDR), oestrogen receptors (α and β), apolipoprotein E, collagen type I α 1 and methylenetetrahydrofolate reductase, amongst many others. However, in general, findings from numerous studies of the association between such genes and various bone variables have been inconsistent. In addition to possible gene—gene interactions it is likely that there are interactions between these genes and certain environmental factors, especially nutrition, that may mediate expression of bone-related phenotypes. While these potential interactions add a level of complexity to our understanding of these apparent genetic effects on bone, identification of a role for genetic factors without knowledge of their interaction with nutrients can do little to advance prevention and treatment of osteoporosis. This information is especially important because, unlike genotype, diet and nutrition can be modified. The aim of the present review is to critically evaluate current knowledge relating to candidate genes for osteoporosis, with particular emphasis on their interaction with nutrients and dietary factors in determining bone health.