Evidence guidelines

Guidelines on the evidence required to support indications for listed complementary medicines

Book pagination

24 July 2014

Part B: Further technical guidance

Part B of this guide provides additional technical information on specific subjects and clinical issues relating to listed medicines. Additionally, case studies are provided for greater clarity of particular technical issues, such as indications relating to biomarkers, nutritional supplementation and weight loss.

Equivalency of ingredient, preparation, dosage and dosage form between the evidence and your medicine

The active ingredient should be well characterised in the evidence supporting your indication. Preparations used in the evidence should contain the same ingredient preparation and dosage form as your medicine.

In the case of indications based on vitamins, minerals, nutrients or known therapeutically active components of herbs, this involves careful consideration of the dose, route of administration and dosage regimen employed in the available scientific literature. In order for a piece of evidence to be relevant to your indication, all these factors should closely resemble that intended for the medicine.

Evidence that relates to a herb or herbal substance, the species (and subspecies if applicable), plant part, method of preparation and processing, the equivalent dry weight and the dose of active component used should be consistent with that of the herb or herbal substance in the medicine. If the processing used to prepare a particular herbal product is different to that used in the literature, you will need to hold evidence that the chemical profile of the resulting active ingredient(s) is not substantially different from the active ingredient used in the evidence to support your indication. Unfortunately, many trials inadequately describe or characterise the composition of the herbal treatment. Even when the herbal ingredient is standardised to known active therapeutic components or marker compounds, there can be variation in the concentration of other components in the herbal extract for example that may result in different pharmacological activity in vivo.

Other characteristics of medicines used in clinical trials may also impact on relevance to your proposed indication. For example, modified release dosage forms of a medicine designed for slow or delayed release of an active ingredient may not be relevant to support indications that imply health outcomes that are achieved rapidly (for example: 'for the rapid relief of pain').

Top of page

Scientific studies

Study populations

As indicated in Part A of this guidance, only human studies are considered sufficient as primary evidence to support indications for your listed medicine. Studies used to justify the scientific indications for your medicine should be conducted in populations that are representative of, or can reasonably be extrapolated to the general Australian population.

Relevance of the study population

The health status of the study population should be representative of the target population for your medicine. Many listed medicines are used by healthy individuals. Data collected from study populations with non-serious disorders and in situations where a continuum of health and disease exists, such as individuals in early disease states, can influence the relevance of the evidence for your indication.

In general, data obtained from studies with participants who have serious diseases, conditions or ailments cannot be extrapolated to a healthy population and, as such, are not relevant evidence to support an indication for a listed medicine.

However, in circumstances where a positive modulation of a health benefit is noted in a diseased study population, it may be possible to use these clinical outcomes to provide secondary evidentiary support for your indication. Such an indication should be carefully worded to accurately reflect the evidence base supporting to the indication.

For example:

  • 'May assist with maintaining normal healthy cholesterol levels'.
Justifying the differences between study and target population

When you use clinical studies that employ specific study population groups (for example: subjects with a disease) rather than the target population group for your indication (for example: the general healthy population), you should provide an evidence-based justification.

This process should consider biological factors as well as environmental and behavioural factors including the influence of health practitioner intervention which may differ between healthy and unhealthy populations.

In vitro studies or non-clinical data can provide additional justification to support the outcomes of the primary clinical study using diseased populations. Data generated from diseased study participants could be used to demonstrate that the active ingredients have positive pharmacological effects; however such data may not be appropriate to support an indication that is intended to be used by healthy people to maintain normal biomarker levels. You should also consider the mechanism of action of the medicine or ingredient and whether it is applicable to the general healthy population. The pathophysiological changes in a disease population may result in the ability of a particular ingredient or medicine to be effective. The same result may not be achieved in a general healthy population as it may be dependent on pathophysiological changes associated with the diseased population. Further, any favourable modulation is likely to be dose-dependent; therefore consideration should be given to the effect, if the dose requires modification.

Target sub-populations of the general Australian population

When an indication is directed towards a specific subgroup of the population (for example: elderly or pregnant women), it needs to be supported by evidence derived from the same subgroup of the population. The results from studies of target specific subgroups are not relevant to the general population. For example, clinical studies that use females as the treatment group are not relevant to the general population.

Table 6 provides examples of the characteristics of study populations that are relevant to the target population.

Table 6: Characteristics of study populations that are relevant to the target population
Indication Relevant population
For weight loss in overweight individuals when used in conjunction with a calorie or kilojoule controlled diet and physical activity (or exercise). Male and female participants aged 18-65 years; generally healthy population with BMI 25-30kg/m2 socioculturally similar to the Australian population
Reduces pain. Male and female participants aged 18-65 years; generally healthy population with a range of painful (non-serious) conditions.
Relieves cough in children. Male and female participants aged 2-18 years; generally healthy population with cough associated with a range of (non-serious) conditions.
Calcium helps maintain healthy strong bones. Male and female participants aged 18-65 years; generally healthy population; dietary and lifestyle pattern similar to the Australian population.

Top of page

Study duration

Relevant studies should be of appropriate duration to validate a health benefit included in an indication. Each study should be long enough to clearly demonstrate the health benefit. The appropriate duration of studies depends on the nature of the health benefit. If an indication refers to a short-term benefit such as acute pain relief, trials of several hours duration may be adequate. Conversely, for indications where long-term benefits are implied, studies must be of sufficient duration to establish a sustained response that is likely to be meaningful. This is particularly important for indications relating to maintenance of health or risk reduction, and those that produce favourable modulation of biomarkers and body weight, as the body's homeostatic processes may reduce early gains. Therefore, studies assessing cardiovascular risk factors, weight, or changes in muscle mass or bone strength that are not long enough to establish a sustained clinical benefit are not relevant.

For these reasons, the duration of each study is an important factor and must be considered when assessing the body of evidence relevant to an indication. The minimum relevant study duration should be determined and justified in relation to the relevant indication, and all studies of insufficient duration should be omitted from the primary analysis.

CASE STUDY 1: Study duration for pain relief

A listed herbal medicine containing willow bark (Salix alba) has the scientific indication: 'Helps relieve short-term back pain'.

A clinical trial reports a long-term pain relief effect in subjects suffering osteoporosis and back pain beginning 2 weeks after the initiation of treatment.

The above indication is not supported by the outcomes of this study.

CASE STUDY 2: Study duration for weight loss

A listed herbal medicine has the scientific indication: 'Helps assist weight loss'.

Evidence held to support the indication was a clinical trial of 6 weeks duration.

A reasonable timeframe to achieve a significant degree of weight loss is six months (clinical significance). After about six months, the rate of weight loss usually declines as weight plateaus, and some regain is common [Franz (2007)].

Shorter studies fail to demonstrate the full benefit of a treatment, including the ability to sustain weight loss for a longer period. Therefore, studies will generally be at least six months duration to be considered relevant to indications relating to weight loss must be of at least six months duration.

The above indication is not supported by the outcomes of this study.

Top of page

Study outcomes

Primary and secondary clinical study outcomes

Ideally the health benefit of your medicine will be included in the study as a primary outcome with an adequate sample size. This ensures that the study is sufficiently powered to detect a benefit that is statistically and clinically significant. However, inclusion of the health benefit as a secondary outcome may be acceptable provided that the observed result is shown to be statistically and clinically significant.

Evidence that describes an effect on a biological process generally does not contribute to the evidence base for an indication that refers to a clinical outcome. Such data may, however, be useful in demonstrating biological plausibility of a clinical outcome.

Evidence held to support indications referring to beneficial effects on biological or clinical targets should directly relate to the target described. Evidence relating to a particular clinical outcome, physiological process or health benefit cannot be drawn from data describing different clinical outcomes, physiological processes or health benefits.

CASE STUDY 3: Duration versus symptomatic relief

A clinical study that assesses the effect of an ingredient on the duration of the common cold does not support an indication that describes symptomatic relief of the common cold. Therefore, the indication can only refer to reducing the duration of a cold and not to relief of symptoms.

Assessing the significance of study outcomes
Attrition rates (drop out rates)

Attrition rates are commonly high in studies that evaluate health gains that are modest and require long-term commitment. High attrition can introduce serious bias (attrition bias) into these studies because the reasons for non-completion vary across initially randomised groups. High attrition rates may also diminish the general applicability of the treatment to the Australian population. The resulting data from a high attrition study should be interpreted with caution.

An Intent-To-Treat (ITT) analysis, in which outcomes of the original randomised groups are compared, provides a means of accounting for the effects of dropouts. In an ITT analysis, dropouts from the study are included in the analysis. When an ITT is performed, all efforts should be made to obtain outcome measurements from dropouts at the end of the study. In cases where this is not possible, baseline measurements of study parameters should be carried forward (for example, for a study outcome related to weight loss, body weight recorded at the beginning of treatment would be the same at the end of the study). A treatment effect demonstrated in an ITT analysis underestimates the efficacy of the treatment but may be a good reflection of effectiveness under real world conditions [Koepsell & Weiss (2003)]. When dropouts are not accounted for in the analysis of results, attrition bias (exclusion bias) may result.

Number of participants (power calculations)

It is important that studies enrol sufficient numbers of participants to detect a significant and reliable treatment effect. The number of participants required to be reasonably certain of a reliable result needs to account for the degree of health benefit, the variability of individual results and the number of participants dropping out of the study (attrition rate). As a consequence, studies may need to include larger numbers of participants to account for a high attrition rate.

How many patients are needed for a clinical trial?

An essential part of critically appraising a clinical study is to determine if a sufficient number of study participants were included in the trial in order to reliably detect and measure effects of the intervention.

This is described as the studies 'power'. The larger the number of participants, the greater the potential statistical power.

The number of study participants required for a study to demonstrate clinical significance depends on several factors: the study aim and design, the type and sensitivity of the primary end-point, how the data will be analysed, the significance level and allocation ratio of treatment to control as well as the anticipated standard deviation or the anticipated results in the control group.

There are many published resources available in which these factors are explained including good and bad practices for sample size calculation.

Often high quality clinical studies have been designed to provide meaningful statistical calculations. Study clinicians have considered the factors that are important to achieve the desired outcome and have designed the study accordingly. If you choose to use a clinical study to support a scientific indication, you are not expected to perform power calculations, but to consider any limitations of the statistical calculations that the study authors have reported, including the number of drop outs and the impact this may have on the reported study outcomes.

Appropriate statistical methods must be used to compare the effects of treatment between groups, and to compare the number of individuals achieving a clinically significant result in each group. The analysis should also account for any potential confounders. An Intent-to-Treat (ITT) analysis should also be performed, particularly when attrition rates are high. Previously unplanned analyses undertaken after the completion of a trial (post-hoc analyses) are to be avoided as they are unlikely to have been considered in power calculations and study design.

An indication can only be justified when the available evidence supports the described health outcome. The balance of evidence should support an outcome that is:

  • statistically significant; and
  • clinically significant (or meaningful to the consumer).
Statistical significance (p-value)

When considering a study result, it must be unlikely (probability of less than 5%) that the observed health benefit could have been a chance occurrence. The 'p' value indicates the probability that an effect is due to chance, assuming there is no real difference between intervention and control groups. Therefore, a 'p' value of less than 0.05 indicates with acceptable certainty that an observed effect or health benefit is unlikely to be due to chance.

Confidence intervals provide an alternative measure of statistical certainty. Confidence intervals of 95% are commonly employed to show the range within which the true outcome value could be expected to occur with 95% certainty. When 95% confidence intervals are generated for primary study outcome measures, the 95% confidence intervals of the intervention and exposed groups must not overlap.

However, statistical significance does not provide information about the degree of benefit produced or whether it is likely to be meaningful.

Clinical significance

Not all statistically significant differences are clinically significant (Berry (1986), Sackett et. al. (1985), Levitt (1981)). A statistically significant outcome indicates only that there is likely to be a relationship between intervention and outcome. Clinical significance is more difficult to define but is commonly considered to represent a degree of benefit that is worthwhile in real life to justify intervention, and may consider factors such as cost, side effects and inconvenience.

A number of general principles can provide guidance about clinical significance. For listed medicines, it might be regarded as a degree of benefit that is meaningful to the consumer. The number of participants required to detect a clinically significant difference between treatment and control groups depends on the type and level of health benefit, the standard deviation of the health effect, the significance level (p-value) and statistical power of the study and the type of hypothesis being tested.

In general terms, most research studies contain 0.8 sample power, meaning that there is an 80% probability of finding a significant difference with a given sample size, if a real difference truly exists and having excluded the role of chance. High quality studies will recruit many more subjects than required in order to maintain adequate numbers in the trial even when there are drop outs recorded throughout the study. The meaningfulness of a predetermined 'significant clinical benefit' may then vary between patients depending on a number of factors such as state of disease, comorbidities, personal circumstances, and alternative options for treatment.

Judgements about clinical significance are often made by experienced clinicians within a context of ongoing monitoring and supervised care. Listed medicines, however, are freely available to consumers and may not involve practitioner intervention or supervision. Determining the clinical significance of health outcomes associated with listed medicines is particularly difficult for the following reasons:

  • Listed medicines are self-selected by consumers from a wide variety of backgrounds, with varied expectations and variable educational and financial resources.
  • The health outcomes provided by listed medicines may be modest, not readily apparent, and/or achieved over long periods of time.
  • Healthy consumers may be satisfied with smaller gains in health than individuals with a pre-existing condition.

Notwithstanding these factors, consideration should be given to the likely significance (meaningfulness) of an observed health outcome to the intended target population. Table 7 provides a useful approach to the assessment of clinical significance for listed medicines.

Table 7: Approach to the assessment of clinical significance for listed medicines
Excellent Good Satisfactory Poor
Clinical impact Meaningful health benefit very likely to be achieved by consumers Meaningful health benefit likely to be achieved by consumers Impact on target population uncertain-health benefit possible. Unlikely to be meaningful
Ethnic, cultural and social factors

The characteristics of study participants must also reflect the characteristics and lifestyle of the target population for the medicine. Consideration of genetic, ethnic and socio-cultural factors is important when assessing the relevance of scientific evidence used to substantiate indications as differences in any of these may result in discrepancies between results reported in study data and expected results in an Australian population. The Australian population is culturally and ethnically diverse. Scientific data obtained from studies conducted in homogenous ethnic populations may be limited in their relevance to the general Australian population. Factors such as diet, lifestyle, support networks and religious beliefs may all impact on the study findings. Your evidence summary should include a justification to illustrate that the results derived from clinical studies that employ homogenous ethnic study populations are relevant to the general Australian target population, and therefore adequate supportive evidence for your medicine.

Top of page

Relevance of context for your evidence

It is important to recognise that the body of evidence relevant to your indication is generally derived under conditions that are more restrictive than those experienced by consumers of listed medicines in 'real-world' situations. In research studies, tight control of experimental conditions and intensive monitoring are important in controlling for confounding across treatment and placebo groups. Studies conducted in this way are ideal for estimating potential efficacy but may overestimate effectiveness within its target population.

Studies that are less prescriptive may provide useful adjunctive information about 'real-world' medicine effectiveness. However, such studies may not accurately predict potential medicine efficacy, as the results of such studies may be subject to bias due to differences in environmental conditions, participant characteristics and compliance.

Provided that measures are taken to ensure that the characteristics of your medicine, its indications, and its target population are consistent with the supportive evidence base, well controlled efficacy studies are considered the 'gold standard' for supporting health benefits associated with your listed medicine. However, in situations where real-life effectiveness is likely to be significantly less than that observed in trials, the expected result in the general population should still be clinically meaningful.

Contextual qualifiers can be included in an indication so that it accurately reflects the evidence base held to support the indication.

For example:

  • 'May assist with weight loss when used with calorie controlled diet and exercise'.
  • 'May assist with weight management when used with calorie controlled diet and exercise'.

For indications that imply an ability of a listed medicine to assist with weight loss, the same level of evidence (that is, relevance, high quality, statistical & clinical significance and balance of the evidence) is required as for any other indication for a listed medicine. Specifically:

  • the proposed indications should be appropriate for listed medicines (that is the indications cannot refer to obesity or a serious condition)
  • the proposed indications should be consistent with the evidence held to support such indications
  • the design and quality of the study should allow accurate conclusions of study outcomes to be drawn
  • the study aims and the limitations identified by the study authors should be taken into account when developing your indication
  • the clinical trial duration needs to be sufficiently long to support indications that refer to long term benefits, for example: sustainable weight loss is greater than 6 months; and
  • the study should be carried out in a study group representative of the population group for which the indication is made. Any extrapolation of results obtained from subjects outside the target population group must be appropriately justified.

The use of qualifiers relating to the biological or clinical target of an indication restricts the applicability of the indication to a specific type of a condition or process (such as mild pain rather than pain more broadly) and narrows the relevant evidence base:

For example:

  • 'may assist with reducing pain'; is less limiting than: 'may assist with reducing mild pain'.
CASE STUDY 4: The use of qualifiers to limit the health benefit

A listed herbal medicine containing Actaea racemosa (Black cohosh) has the scientific specific indication:

  • 'May help to relieve menopausal symptoms'.

This indication may imply the relief of all symptoms associated with menopause (that is: hot flushes, insomnia, irritability, anxiety, vaginal dryness and so on).

The evidence base to support the indication using the same preparation and dose of Actaea racemosa reports a significant reduction in the frequency and intensity of hot flushes only. The effect on other menopause symptoms and signs were not examined. A more appropriate limiting indication that is supported by the evidence is:

  • 'Helps reduce the frequency and severity of menopause hot flushes'.

Qualification may provide additional clarity to a particular indication, such as 'assists', 'helps', 'when used in conjunction with x or y' to convey that the strength of the evidence held to support a particular health benefit may not be as strong as other instances whereby a qualifier was not used.

Study design and methods

Studies should clearly document aims and methods. Study design (including the presence or absence of randomisation and blinding), measurement techniques and statistical methods must be clearly outlined. Inclusion and exclusion criteria and the baseline characteristics of study participants should be described. The baseline distribution of potential confounders must be shown and any potential confounding factors must be considered and accounted for during the analysis. In addition, any limitations and ability to apply the results to the general population should be discussed.

Intervention and control groups (study trial arms)

All participants enrolled in a clinical trial are considered to be derived from a common population and may be allocated to control (placebo) or intervention (treatment) groups. Randomisation of participants to intervention and control arms of the trial helps reduce innate inter-group differences and potential bias. The method of randomisation must be clearly described so as to enable the reviewer to assess the possibility of unblinding. Baseline characteristics of treatment and control groups should always be documented to establish equivalence in key areas such as age, weight, diet and other factors that may contribute to non-treatment differences in health benefit between groups.

Ideally, trials should be conducted under conditions where the only difference between groups is that one is exposed to the intervention (treatment), while the other is not. This is often achieved in controlled trials, but is less likely to occur in cohort studies and case-control studies. In these methodologies, the presence of potential confounders and study biases may impact on study results and must be considered and accounted for in the analysis of the study. When confounders exist within a study, they lessen the study's quality and the degree of confidence in the reported study outcomes. Under these circumstances, care should be taken in describing the indication.

Top of page

Reference to secondary methods of assessment

In some clinical studies, the study authors may refer to secondary methods to assess study outcomes (for example, visual analogue scales as a method to subjectively assess study outcomes (pain, hunger etc)). While this reference to secondary methods is appropriate, it is important that these methods are accurately validated, to ensure the results can be reproduced. The characteristics of the secondary methods are often reported in other studies or publications, and it is this original research that validates the scales.

If your scientific evidence relies on secondary methods of analysing study outcomes, then these original studies will also need to be cited and provided to the TGA on request.

CASE STUDY 5: Reference to secondary methods of assessment

A clinical study was provided to support the indication

  • 'assists with the reduction of appetite'.

The supporting clinical study used a visual analogue scale (VAS) secondary assessment method [reference: Silverstone et. al. (1981)] to determine the changes in four parameters associated with the assessment of appetite ('hunger', 'thoughts of food', 'urge to eat', and 'fullness of stomach').

However, the method of determining appetite in the clinical study that administered VAS at thirty day intervals, was inconsistent with the method reported by Silverstone et. al. (1981), which administered the VAS at hourly intervals.

As the changes to this secondary method of assessment (VAS) were not appropriately validated in the clinical study, the study does not support the indication.

Top of page

Indications relating to biomarkers

Your indication may refer to the favourable modulation of measurable biomarkers of disease (such as Body Mass Index (BMI), blood pressure, blood glucose and cholesterol) in healthy individuals. This is difficult to substantiate when the primary evidence is derived from a diseased population. The extrapolation of study findings from a diseased study population to the healthy Australian population can be problematic and potentially misleading. A small change in a given biological surrogate may be associated with negligible biological dysfunction and minimal increase in risk of serious forms of disease, whereas larger changes are more likely to be associated with pathophysiological processes and an increased risk of overt illness which requires health practitioner involvement.

Evidence may also need to be sourced to illustrate that favourable modulation of biomarkers in a diseased population does not result in serious conditions in a healthy population, for example: a study reporting the significant reduction of blood glucose in hyperglycaemic patients in the trial should not result in hypoglycaemia in healthy individuals. In these cases, further justification is required to demonstrate that your medicine is safe for its intended purpose.

If your evidence uses study populations with baseline biomarker levels that lie outside normal healthy levels it is unlikely to be considered relevant to support indications relating to biomarkers levels in the healthy Australian population.

For example:

  • 'May assist with maintaining normal cholesterol levels in healthy individuals'.

In addition, indications should only target healthy individuals with biomarker levels that lie within the normal healthy range (as outlined in the Table 1 above). Because of the continuum between health and disease, all biomarker and risk reduction indications should include a disclaimer that recommends consumers consult a healthcare practitioner if they are concerned about their health status.

Indications that refer to the modulation of biomarker levels cannot be supported by evidence of traditional use.

Table 8 summarises the characteristics of values of biomarker ranges for healthy study populations.

Table 8: Summary of common biomarker ranges for healthy study populations
Biological Parameter Value
Body weight (BMI) 20-24.9 kg/m2
Blood sugar Fasting 4.0-5.5mmol/L
Blood pressure 100/60-140/90mmHg
Blood cholesterol (total) <5.5mmol/L
Blood cholesterol (LDL) 2.0-3.4mmol/L

Top of page

Indications relating to weight loss

Indications relating to weight loss or management require supporting scientific evidence that demonstrates that the initial weight loss is meaningful to the consumer (that is, clinically significant) and can be maintained after the initial weight loss period.

In Australia, registered medicines targeting obese populations are required to demonstrate an absolute reduction in weight loss of at least 10% over one year (EMA (2007)). This degree of weight loss may not be desirable or appropriate for mildly overweight individuals. It is commonly accepted that a loss of 5% of initial body weight over six months is likely to represent a clinically significant degree of weight loss and is considered a minimum degree of weight loss required for listed medicines that are indicated for weight loss. Lesser degrees of weight loss are unlikely to be clinically significant and therefore generally considered inadequate to support indications associated with weight loss. It is possible for lifestyle modification alone to give similar weight loss results (Franz et. al. (2007), Wu (2009), Sacks (2009), Rose and Day (1990)).

In weight loss trials the control group commonly achieved some degree of weight loss due to changes in lifestyle, such as dietary intake and exercise. Evidence supporting weight loss indications claimed for listed medicines should demonstrate that the degree of weight loss is meaningful and unlikely to be attained through diet and exercise alone.

Rose and Day (1990) postulated that a mean reduction in BMI of approximately 1 kg/m2 (one BMI unit) across a population could make significant impacts on the prevalence of obesity and overweight individuals within the population. A mean body weight loss of 3% is likely to be equivalent to a mean loss of one BMI unit in the population enrolled in a clinical trial. However, the clinical study population will often include obese individuals which are different to a healthy target population.

Obese people expend more energy for a given activity because of their larger body mass. Therefore, for the same level of dietary energy and physical activity, the reduction in body weight will be different for obese (BMI >30 kg/m2) and overweight individuals (BMI 25-30 kg/m2). This difference may be negligible for small increments in BMI but is likely to become increasingly significant as BMI increases.

Thus the degree of weight loss is likely to be different when comparing overweight and obese individuals when given the same treatment protocol. As such, studies that include obese participants with a BMI >30 kg/m2 cannot be generalised to otherwise healthy overweight individuals.

It follows therefore, that in non-randomised controlled studies, the treatment group (BMI 25-29.9 kg/m2) should show at least a 5% greater weight loss than the placebo group to counter for potential confounding. There must be a reasonable chance that meaningful weight loss will be achieved in consumers investing in the medicine. Single mean values may be misleading and it is important that the effect of an ingredient or medicine represents a consistent effect across the whole target population. At least 50% of participants in the treatment group must achieve a loss of at least 5% of initial body weight, making it 'more likely than not' that consumers will achieve a clinically significant benefit from appropriate use of the medicine.

For each clinical study used to support weight loss indications, the meaningfulness of the observed effect to the general Australian population should also be assessed. Study outcomes that report statistical significant changes in weight loss parameters must also demonstrate clinical significance, or provide a meaningful health benefit to the consumer of your medicine.

Weight loss indications and scientific evidence

In general, for indications relating to weight loss in overweight individuals (BMI 25-30 kg/m2), clinical significance (that is a health benefit that is meaningful to the consumer) is only achieved if supporting scientific evidence demonstrates:

  • a mean overall loss of at least 5% initial body weight in the treatment group, which is at least 3% greater (for RCT) OR 5% greater (for non-RCT) than that of the placebo group. In both cases the difference must be statistically significant (p<0.05); and
  • at least 50% of participants in the treatment group must have achieved a loss of at least 3% or 5% of initial body weight, respectively; and
  • the study duration is a minimum of 6 months.

CASE STUDY 6: Primary and secondary clinical outcomes

A clinical study reporting the ability of Caralluma fimbriata to reduce hunger (primary outcome), does not support a weight loss indication, as only studies that directly assess weight loss can be considered relevant to the evidence base for a weight loss indication.

Changes in fat metabolism, thermogenesis, metabolic rate or reduction in hunger do not necessarily translate into weight loss and evidence supporting these indications cannot be extrapolated to support an indication for weight loss.

Table 9 provides examples of terms that are often related to, or used to convey weight loss that should not be substituted for the term weight loss in an indication. The evidence should support your indication, thus if your evidence refers to the reduction of hunger, then your indication could refer to reducing hunger, without extending this to weight loss or management.

Table 9: Terms that should not be substituted for the term 'weight loss' in an indication
Metabolism Body shape and composition Weight-related Appetite

Increased metabolic rate

Enhanced metabolism

Enhanced fat metabolism


Increased calorie burning

Fat loss

Increased muscle mass



Weight maintenance

Weight control

Weight management

Appetite suppression

Enhanced satiety


Top of page

Indications relating to nutrient supplementation

If a listed medicine states that it is intended to supplement a named nutrient, it must provide at least 25% of the Recommended Dietary Intake (RDI), Adequate Intake (AI) or nutrient reference value for that nutrient and the nutrient should be in a form that is available for absorption by the body.

Supplementation non-specific indications

Non-specific supplementation indications are those commonly linked to medicines that only contain vitamins, minerals or nutritional substances as ingredients.

Statements relating to supplementation with vitamins, minerals or other essential nutrients (for example: a source of calcium) that imply a general health benefit (such as the maintenance of good health) are often supported by high-quality and credible scientific literature, such as internationally recognised pharmacopoeias or monographs, descriptive studies, case series or reports of relevant expert committees.

Providing the listed medicine provides the required amount of the nutrient, vitamin or mineral; reference texts, such as pharmacopoeias or monographs, or other evidence-based reference texts, are sufficient to support non-specific claims.

Supplementation specific indications

When the supplementation claim is linked to a specific therapeutic benefit, then additional scientific evidence is required to support the claim.

The salt of the nutrient/mineral/vitamins should be in a form that is readily absorbed by the body. The dosage directions for the medicine should ideally optimise the effect of the medicine (for example, take with food). For each nutrient, availability from non-food sources depends on the dosage, transport mechanism, age, gender, deficiency status and whether the supplements are taken with a meal (for example: Calcium citrate is 2.5 times more bioavailable (rate and extent of availability in plasma) than calcium carbonate when taken with a meal (Heller et. al. (2001)). Where vitamins, minerals or other nutrients are the subject of other indications, the dose must be consistent with the evidence to support the indication.

All indications for nutrient-containing medicines, whether implicit or explicit, must be appropriate for listed medicines. In general the indication must not refer to serious forms of a disease, condition, ailment or defect. Often the term 'dietary' is used as a qualifier to limit potential references to serious diseases and nutrient deficiencies.

For example:

  • 'Helps to prevent dietary vitamin B12 deficiency'.
  • 'Assists in the prevention of dietary iron deficiency'.
  • 'Prevention of dietary vitamin [XX] deficiency'.
  • 'Prevention of dietary mineral [YY] deficiency'.
  • 'For the management of medically confirmed dietary vitamin [XX] deficiency'.
  • 'For the management of medically confirmed dietary mineral [YY] deficiency (This indication is not to be used for the treatment of iron deficiency conditions)'.

Book pagination