A Thought on the Practice of Functional Medicine and Clinical Nutrition as We Enter a New Year and a New Decade: How Does the Massive Escalation of Prescription and Over-the-Counter Drug Use Make Difficult Patients "Difficult" (Hint – It May Not Be What You Think) – Part II


I would now like to continue my exploration of pharmaceutical interactions with common dietary and supplement constituents by finishing my review of the Nomani et al (1) paper, which explores interactions of pharmaceuticals with cola-containing drinks (CCDs). In part I of this series I reviewed the first portion of the paper that highlighted these interactions in more general terms and the mechanisms of these interactions.  The final portion of the paper that I will review now considered interactions with specific drugs.


To begin this discussion, recall from part I of this series that methotrexate is a weak acid.  Because of this, renal excretion of methotrexate is decreased in more acidic urine, which could lead to methotrexate toxicity.  Since, CCDs, due to their content of phosphoric acid, are more acidic in nature, could simultaneous intake of CCDs and methotrexate lead to increased risk of methotrexate toxicity?  Nomani et al (1) answer this question by discussing two patient reports.  The first involves a pediatric population with acute lymphoblastic leukemia:

“Bauter et al. conducted a study on a small pediatric population with acute lymphoblastic leukemia.  The patients received high doses of methotrexate (MTX).  The authors reported a decrease in the rate of MTX renal excretion, since the major route of elimination of MTX

is renal elimination from the body via the kidneys.”

The second is a case report of a 56-year old man with non-Hodgkins lymphoma:

“It has been reported that a 56-year old man with non-Hodgkins lymphoma, who received high doses of methotrexate (MTX), developed acute renal failure.  These observations were attributed to the simultaneous consumption of CCDs and the administration of MTX.”

What was the explanation provided for this unfortunate occurrence?  The authors continue:

“The explanation for these observations is that the primary route of elimination for MTX is renal excretion.  Since MTX is a weak acid, alkaline urine would increase excretion.”

Because of this, it is customary during MTX therapy to make sure the urine is alkaline:

“…the alkalinization of urine is an important part of MTX therapy, especially in the case of high-dose MTX therapy.”

Unfortunately, because CCDs are acidic, they can negate the efforts to alkalize the urine by administrators of MTX to optimize MTX excretion:

“Interestingly, CCDs, with their acidifying properties due to phosphoric acid, can transiently neutralize the effect of standard alkalinization regimens.  In fact, this phenomenon occurred in patients who received MTX in both of the above-mentioned studies.”

How quickly can CCDs lower the pH of the urine?  Nomani et al (1) continue:

“Santucci et al. reported that urine pH unexpectedly decreased from 8.5 to 6.5 when a urine sample has been taken after consuming 330 ml of a CCD.”

(330 ml equals about 11 fluid ounces)

Nomani et al (1) go on to discuss the dangers of MTX therapy with a urine pH that is too low:

“…in the case of treatment with high-dose MTX, a urine pH lower than 7 can result in crystal formation in the urine, which can be nephrotoxic, which can contribute to the development of acute renal failure.”

With the above in mind, the authors recommend:

“…it is recommended that patients who receive high doses of MTX avoid any CCD at least 24 h before, and during the period over which MTX administration occurs until complete elimination of the MTX has taken place.”

It is my hypothesis that your patients who are receiving high doseMTX therapy for various cancers have not been advised to avoid CCDs and other highly acid foods and liquids during the time of MTX administration, thus increasing the risk of MTX-related adverse effects.  Therefore, it is my hope that you will act as missionaries to deliver this vital information to anyone you know who is receiving or is scheduled to receive high dose MTX therapy.


As we all know, lithium has been a commonly prescribed pharmaceutical for bipolar disorder for years.  Nomani et al (1) begin their discussion of lithium by mentioning a case report where a 58-year-old man who had been prescribed lithium for bipolar disorder was presenting with unexplainable low lithium levels.  Why did this happen, according to the authors of the case report?  Nomani et al (1) point out:

“The authors attributed this observation to excessive consumption of CCDs.”

What was the relationship?  In this situation it had to do with the caffeine content of the CCDs, which affects lithium reabsorption in the kidney, which is normally 80%.  Nomani et al (1) state:

“It is well established that lithium is freely filtered and excreted into the urine, but 80% of filtered lithium is reabsorbed.  However, the reabsorption and excretion of lithium is mainly dependent on the sodium concentration of urine.  Excessive consumption of a CCD causes significant accumulation of caffeine. Caffeine, by inhibiting the tubular reabsorption of lithium and increasing the urinary concentration of sodium, can lead to an increase in lithium clearance.”

With the above in mind, the authors issue the following warning:

“Due to the fact that lithium has a narrow therapeutic index, clinicians should inform the lithium patient about the potential interaction between lithium and CCDs and other caffeinated drinks.  It is recommended that the amount of these drinks be limited.”


The interaction of clozapine, a fairly common antipsychotic drug, with CCDs involves the impact of caffeine on the activity of the phase I detoxification enzyme CYP1A2.  Nomani et al (1) begin their discussion on the relationship between clozapine and CCDs by highlighting a case report:

“In one case report, elevation of clozapine serum concentrations, as well as its main active metabolite, norclozapin, has been reported.  The elevation in clozapine serum concentrations was attributed to excessive consumption of CCDs by the patient.”

What is the explanation for this relationship?  As I mentioned above, it had to do with the impact of caffeine on cytochrome P450 detoxification enzymes:

“This observation can be explained by the fact that clozapine is primarily metabolized by CYP P 450, especially CYP1A2.  Additionally, excessive consumption of CCDs results in excessive consumption of caffeine, which is almost exclusively metabolized by CYP1A2.  The caffeine competes with clozapine for binding to CYP1A2 and then may competitively inhibit the CYP1A2-dependent metabolism of clozapine, leading to increased serum concentrations.”


Next, Nomani et al (1) consider the interactions of CCDs with the anticonvulsant carbamazepine.  In contrast to the situation with CCDs and clozapine, CCDs tend have a duel impact on carbamazepine due to the effects of pH and caffeine.  First, consider the impact of the acid pH of CCDs on absorption of the drug:

“…a CCD increased the rate and extent of absorption of carbamazepine.  This increased absorption may be due to enhanced aqueous solubility of this basic compound in the more acidic environment of the stomach, because of the concomitant use of acidic CCDs.”

Why is this a cause for concern?  The authors continue:

“Importantly, an increase in the oral absorption of carabamazepine when ingested with CCDs means an increased bioavailability of carbamazepine, which is accompanied by an increase in hepatotoxicity and the anticonvulsant effects of the drug as reported in experimental animals.”

Caffeine may also have an effect on carbamazepine in the following way:

“…caffeine can significantly reduce the anticonvulsant effects of carbamazepine without affecting the plasma concentrations of the drug.  Therefore, caffeine may have a pharmacodynamic interaction with carbamazepine.”


The interaction of CCDs with warfarin is of particular concern since so many of our older patients are using this drug.  Nomani et al (1) begin their discussion of this interaction with a case report:

“In one case report, a man who drank about 1.5 L of regular cola per day and used warfarin at a dose of 20 mg/day had a consistent INR below 1.5.  When his daily intake of cola was limited to 350 ml, the value of the INR stabilized between 2.5 and 3 with warfarin dose of 7.5 mg/day.  After being within, or above, the target range for several years, during a regular checkup, his INR was found to be 1.1.  He reported a regular coffee intake of more than five espressos per day.  However, when his coffee was limited to one cup of espresso per day, his INR stabilized within the therapeutic range.”

As I hope you agree, this case report is particularly disturbing because of the significant frequency of warfarin use among our patients.  How often do we hear from patients that, because INR measurements appear lower on routine checkups, warfarin doses are regularly increased without even a thought by the prescribing physician as to why the INR was lower?  My experience suggests all too many times. 

What is the nature of the interaction?  With warfarin it appears to be a combination of the impact of the acidic nature of CCDs and the caffeine content.  Nomani et al (1) suggest:

“…warfarin resistance may be due to decreased absorption or increased clearance.  Warfarin is a weak acid.  Theoretically, co-administration with an acidic drink could lead to decreased absorption.”

In the next quote, the authors discuss a highly relevant case report discussing an interaction between warfarin and ascorbic acid:

“A similar decrease in INR and apparent resistance to warfarin has been reported in the case of patients using ascorbic acid, which can decrease gastric pH.  After discontinuing ascorbic acid, the value of the INR increased.”

The next quote by Nomani et al (1) refers to a hypothesis presented by the authors of the above mentioned case report on the patient who ingested excessive CCDs and then, much later, excessive espressos:

“They suggested that warfarin occurs as a pair of enantiomers: R-warfarin is metabolized primarily by cytochromes P450 (CYP) 1A2, 3A4, and by carbonyl reductases, and S-warfarin is metabolized primarily by CYP2C9.  Because S-warfarin has a higher potency than R-warfarin, the efficacy of warfarin is affected primarily when metabolism of S-warfarin is altered.  Exposure to caffeine induces expression of the enzyme CYP1A2, but at a higher caffeine concentration, the contribution of CYP1A2 to caffeine metabolism decreases in favor of CYP2C9, a situation that could affect the metabolism of S-warfarin.”

The final hypothesis mentioned by Nomani et al (1) as to why caffeine can adversely affect warfarin metabolism relates to a subject of great concern for functional medicine practitioners these days, SNPs:

“Additionally, CYP450 polymorphisms could explain the variability of responses to caffeine in warfarin users.”

As I mentioned above, given how many of our patients are taking warfarin, I feel that this issue is greatly under-appreciated and misunderstood by many nutritional practitioners and certainly a high majority of medical physicians is of extreme importance.  While I realize that the ways that caffeine and CCDs in general can interact with warfarin are somewhat complex, what I feel is quite clear is that periodic alterations in bleeding times are not just random, unexplainable occurrences that can only be remedied by increased doses of warfarin and/or addition of other anticoagulant drugs.  Rather, there is a very good chance that these suboptimal changes in bleeding times can be explained by dietary and lifestyle interactions.  Furthermore, as suggested by Nomani et al (1), it would seem that the best course of action would be identification and correction of diet and lifestyle factors that adversely impact warfarin metabolism rather than reflexive, traditional approaches that involve higher doses and/or additional medication.


Nomani et al (1) point out that since the commonly used NSAID, ibuprofen, is a weak acid, using CCDs along with the drug could result in increased ibuprofen absorption:

“…ibuprofen is a weak acid.  Therefore, it is possible that the co-administration of ibuprofen with an acidic CCD increases the proportion of the unionized form of the drug, which, in turn, increases the extent of membrane diffusion and ultimately increases the extent of absorption.”


To end this discussion of the paper by Nomani et al (1) I would like to highlight two important summation quotes from the conclusion. First:

“As it relates to altering the metabolism of various drugs that are taken with CCDs, it should be emphasized that caffeine-related interactions are usually important only in the case of excessive consumption of CCDs, although it should be pointed out that pH- and phosphoric acid-related drug interactions may occur with more typical amounts of CCD ingestion.”

Concerning caffeine, it is my guess that most of our patients are not ingesting high amounts of CCDs.  However, what may not be fully appreciated by patients ingesting even moderate or small amounts of CCDs and many of the practitioners who treat them is, as I suggested in part I of this series, caffeine intake may be excessive when other sources of caffeine such as coffee, sports drinks, and medications are taken into account.  Similarly, the acidifying effect of typical amounts of CCD ingestion, when added to the impact of processed, highly acid foods that are often ingested along with CCDs by many patients, could create a very clinically significant acidifying effect that will not only affect medication use but many other aspects of health in the chronically ill patient.  Therefore, when patients tell us that they ingest only 1-2 bottles of CCDs per week we should not automatically dismiss this level of intake as insignificant in terms of resolution of chief complaints.  Until we have full knowledge of the total intake of medications, caffeine and other highly acid substances, it is impossible to determine the impact of any level of CCD intake on patient health and chief complaints.

With the above in mind, Nomani et al (1) conclude:

“It is absolutely imperative that clinicians be made aware of the reported and potential interactions of medications with consumption of CCDs, especially if those medications have a narrow therapeutic index.”


One important constituent of CCDs that was not addressed to a significant extent in the Nomani et al (1) paper is sugar.  Could there be an interaction between caffeine and sugar in soft drinks that might have an adverse impact on health?  This question was addressed in the paper “Synergistic effects of sugar and caffeine on insulin-mediated metabolomic alterations after an acute consumption of soft drinks” by Gonzalez-Dominguez et al (2).

The first quote I would like to feature from this paper tells us what we already know about sugar but is important to emphasize again because it is the central reason, from a metabolic standpoint, why sugar is so damaging to systemic health:

“…it has been demonstrated that increased sugar intake can provoke insulin resistance, thus inducing a hyperinsulinemic state that may elicit prolonged deleterious effects on the organism.”

As we continue to further appreciate the benefits of cutting edge dietary concepts such as intermittent fasting, a large body of research is making it clear that one of the most important benefits of these types of diets is reducing the amount of time that patients are in a hyperinsulinemic state compared to the usual American diet.  In future newsletters I will be highlighting the vast amount of published literature that makes it clear that chronic hyperinsulinemia, primarily induced by poor diet, is a major, under-appreciated contributor to many if not most of the chronic illnesses presented to us by our patients.

The next quote I would like to feature from the Gonzalez-Dominguez et al (2) paper highlights the fact that soft drinks are a major source of sugar in the Western diet:

“Nowadays, sugar-sweetened soft drinks are one of the major sources of sugars in the Western diet.  The amount of added sugars in these beverages is much higher than that recommended by the World Health Organization, as reported in a recent study.”

What has other research noted about sugar and caffeine co-ingestion?  The authors state:

“Particularly, it should be noted that caffeine co-ingestion significantly alters the glucose homeostasis by reducing insulin sensitivity, possibly as a result of the elevation of blood epinephrine levels, thus sharpening the transitory state of hyperinsulinemia and hyperlipidemia caused by sugar consumption.  Thereby, numerous studies have associated the consumption of soft drinks with increased incidence of obesity, cardiovascular diseases and type-2 diabetes, among other disorders.”

To gain more information on the acute effects of sugar and caffeine co-ingestion on metabolism and health, the authors conducted a study on ten healthy  male volunteers aged 22-26 years. The study occurred over four 1-week periods.  At different times during the study the participants ingested 330 mL of either:

  • Coca-Cola® with varying content of sugar as sucrose and caffeine
  • Normal coke (With sugar and caffeine)
  • Caffeine-free coke (Sugar with no caffeine)
  • Coke Zero (Caffeine with no sugar
  • Caffeine free Coke Zero (No sugar or caffeine)

The different drinks were ingested over a three-hour period.  Blood samples were obtained every 30 minutes during the three-hour period. 

What were the results of the study?  Gonzalez-Dominguez et al (2) comment:

“As would be expected, a marked increase of glucose was observed 30 min after the intervention with sugar-sweetened soft drinks, and then these values tailed off until reaching basal levels.  The acute sugar intake was accompanied by a transitory hyperinsulinemic state, which is known to promote glucose uptake and switch the organism from catabolism to anabolism.” 

What was the proposed reason for this effect?  As suggested above, it has to do with the combined effect of sugar and caffeine on epinephrine and insulin metabolism:

“…it seems clear that combined sugar and caffeine intake elicits a synergic action on insulin homeostasis, where epinephrine appears to play an important role.  Thus, it could be hypothesized that most of the metabolic alterations…could be associated with the broad biological actions elicited by insulin.”

Interestingly, though, the synergistic effect did not end there.  There were also disturbances in lactate production:

“In this sense, we detected a significant increase of serum lactate and creatine levels after ingestion of sugar added beverages, probably as a result of the stimulation of pathways related to energy-metabolism…”

Furthermore, the metabolic impact did not end there.  Lipid and amino acid metabolism were also affected:

“…metabolomic data also showed a decreased content of oleyl- and linoleyl-carnitine at 30 min, which could be indicative of a metabolic shift from fatty acid β-oxidation, the major energy source in fasting conditions, towards the utilization of ingested carbohydrates.  However, this abnormal acyl-carnitine profile was only observed after the intervention with the caffeinated soft drink…, thus demonstrating the great impact that caffeine may have on insulin-mediated metabolic perturbations.”

Before continuing, please note again that the caffeinated soft drink shifted energy metabolism away from fat-burning β-oxidation.  This would suggest that the impact of soft drinks on causing weight gain in the form of increased adipose tissue is more than just an issue of caloric intake, as is continuously and obsessively claimed by the mainstream nutritional community.

What about the impact on amino acid metabolism?  Gonzalez-Dominguez et al (2) state:

“In this context, it should also be noted that only volunteers simultaneously consuming sugar and caffeine presented a disturbed homeostasis of amino acids…”

More specifically:

“Decreased serum levels of various amino acids, including histidine, phenylalanine and tryptophan, could be associated with down-regulated proteolysis in response to suppressed gluconeogenesis…”

Next, the authors comment again on the impact on fatty acid metabolism that was mentioned above:

“Similarly, blocked lipolysis was reflected in reduced content of circulating free fatty acids, together with the accumulation of different classes of glycerolipids.  Hypertriglyceridemia is a traditional hallmark of high sugar consumption as a consequence of the inhibition of lipase activity by insulin.”

How fast did the impact on carbohydrate, fat, and amino acid metabolism occur after ingestion?

“In line with other studies, perturbations in glycolysis (i.e. glucose, lactate), fatty acid β-oxidation (i.e. acyl-carnitines) and lipolysis (i.e. fatty acids) were already detected at 30 min, while altered levels of amino acids were only observed after 180 min, evidencing that suppression of proteolysis is less sensitive to the action of insulin.”

As I hope you can see, ingestion of a sugar/caffeine combination has an adverse effect on the metabolism of all three energy sources – carbohydrate, fat, and protein.  Is it any wonder, then, why patients who regularly consume sugar/caffeine beverages regularly present chief complaints of fatigue and weight gain?  Also, as I mentioned above, please note that this impact is entirely independent of the caloric content of the beverage.  Rather, it tends to be mediated by the adverse impact of a sugar/caffeine combination on insulin and epinephrine metabolism.

If that were not enough, consumption of sugar-sweetened beverages has an adverse impact on liver/gall bladder metabolism:

“Besides these impairments in the energy-related metabolism, the consumption of sugar-sweetened beverages also provoked significant alterations in the homeostasis of bile acids and phospholipids.  It is well known that glucose ingestion elevates post-prandial levels of conjugated bile acids through multiple mechanisms, principally via the stimulation of cholecystokinin release, which is usually accompanied by a simultaneous decrease of various steroid precursors.”

Finally, Gonzalez-Dominguez et al (2) point out that sugar-sweetened drinks can alter phospholipid metabolism, leading to an increased risk of diabetes development:

“…it would be hypothesized that the hyperinsulinemic state provoked by sugar-sweetened drink consumption may elicit deleterious transitory effects on phospholipid turnover, leading to a diabetic-like metabotype.”

Did artificially sweetened soft drinks demonstrate the degree of metabolic alterations discussed above?  Not unexpectedly, no:

“…it should be noted that only a few metabolomic changes were detected after the intervention with artificially-sweetened soft drinks, probably because their consumption do not induce hyperinsulinemia.”

Gonzalez-Dominguez et al (2) conclude their paper with the following summation:

“On the basis of our findings, it could be concluded that sugar is the major player in metabolic perturbations provoked by the consumption of soft drinks, probably as a result of the induction of a hyperinsulinemic state.  Moreover, this study describes for the first time that these alterations are much more pronounced when sugar is administered with caffeine, thus demonstrating a synergic effect of this compound on insulin-mediated perturbations.”

Before leaving this paper, I would like to present a few final thoughts of my own on its important and highly clinically relevant findings.  First, even though this was a three-hour study, the reality of many of our patients is that hyperinsulinemia induction by sugar/caffeine ingestion will often occur during virtually every waking hour.  Why do I say this?  Because, for many patients, the day starts with coffee with added sugar, followed by a cup of tea mid-morning with added sugar, followed by lunch with a sugar-sweetened, caffeinated soft drink, followed by a sugar-sweetened/caffeinated energy drink mid-afternoon, followed by a sugar-sweetened, caffeinated soft drink at dinner, followed by a sugar-sweetened, caffeinated soft drink in the evening with a bowl of popcorn or other snack at the movies or at home in front of the television.

As I suggested above, it is my opinion that a large body of vastly under read and underappreciated research makes it abundantly clear that, among other key metabolic imbalances that contribute to the epidemic of chronic illness in our society today, chronic, diet-induced hyperinsulinemia, is a major player.  In future newsletters I will be exploring the large body of research on hyperinsulinemia and its impact on chronic illness and health in general in great detail.


While the final, recently published study I am about to review does not precisely fit the theme of this series about underappreciated ways that drugs interact with common dietary constituents, I felt that its subject matter and clinical relevance was of such importance that I decided it should be included.  As evidenced by the title of the study “Drug use is associated with lower plasma magnesium levels in geriatric outpatients; possible clinical relevance” by van Orten-Luiten et al (3), it highlights an important reality about a nutrient upon which we generally place great emphasis with our patients, magnesium.  What is this reality?  Contrary to the belief of many nutritional practitioners, low magnesium status is not solely the result of low intake and poor absorption.  Rather, it can relate, particularly in our older patients, to the almost ubiquitous use of pharmaceuticals that I discussed in part I of this series.

The first quote I would like to feature from this paper points out that, even without drugs, low magnesium status due to poor intake is a major, largely ignored issue in our society:

“It is remarkable that although dietary Mg intake of half of the general US population is below the Estimated Average Requirement, Mg inadequacy is not considered a public health problem.”

Furthermore, other health issues can, unbeknownst to many in the health care community, compromise magnesium status:

“Other risk factors for hypomagnesemia are decreased Mg intestinal absorption, intestinal or renal Mg low, alcohol, stress, exercise, diabetes, and drug use.”

As you will see, suboptimal magnesium status due to drug use is the focus of the study.

The database for this study came from 343 Dutch geriatric outpatients.  What were the findings concerning commonly used drugs?  Consider the following:

Magnesium and proton pump inhibitors (PPIs)

“The inverse association between PPIs and Mg as observed in this study is in line with a meta-analysis of nine studies showing a 43% increased risk of hypomagnesemia in PPI users.”

What was the mechanism?

“Mechanistic studies indicate that PPIs decrease Mg absorption by inhibiting pH-sensitive Mg transport via the transient receptor potential melastins (TRPM) 6 and 7 ion channels in the colon.”

What is the clinical impact?

“The increased risk of hypomagnesemia seen in long-term users of PPIs may contribute to the increased risk of myocardial infarction and other cardiovascular problems in users.”

The next quote I would like to feature is a clinical pearl on how to address magnesium status in users of PPIs who are unable to discontinue usage of the drug:

“Interestingly, administration of the prebiotic inulin increased plasma Mg in 11 patients with PPI-induced hypomagnesemia, which might be explained by a decrease of intestinal pH.”

Magnesium and Metformin

As we all know, many of our diabetic patients have been prescribed metformin.  What is the impact of metformin on magnesium status:

“The inverse association observed between the antidiabetic drug metformin and plasma Mg might be caused by metformin-related diarrhea…”

Magnesium and beta blockers

Another very common drug family prescribed to chronically ill patients is beta blockers.  The impact of beta blockers on magnesium is the following:

“Normal-weight users of selective beta blockers also showed an increased probability of having low Mg levels.  Reverse causation may explain this association, as Mg has a mechanistic role in the indications for prescription of these drugs (i.e. hypertension, arrhythmias, angina pectoris, and myocardial infarction).”

Magnesium and anti-coagulant drugs

We all know about the frequency of anti-coagulant drug use among our patients.  What is the impact of these drugs on magnesium status?

“…reverse causation may also explain the increased probability of hypomagnesemia in users of vitamin K antagonists and platelet aggregation inhibitors, their primary indications include thrombosis risk in cardiac arrhythmias and myocardial infarction.”

Van Orten-Luiten et al (3) conclude their paper by stating the following:

“In summary, in hypomagnesemia, risk of malnourishment was increased and Mg supplements were hardly prescribed.  The observed inverse associations between drug use and Mg level could be explained by the drug causing a low Mg level, the underlying disease causing a low Mg level, or by a low Mg level causing the disease (= indication for drug prescription).  The latter might be the case for antithrombotics, beta blockers, antidiabetics, adrenergic inhalants, and bisphosphonates as low Mg level can play an etiological role in pathophysiology of arrhythmias, atherosclerosis, stroke, diabetes, asthma, and osteoporosis.”

Therefore, as I hope you can see, when we are evaluating magnesium status in our patients, we need to be aware of all drug use as well as dietary history to get an accurate picture about the need for increased dietary magnesium or magnesium supplements.

Moss Nutrition Report #290 – 03/01/2020 – PDF Version


  1. Nomani H, Moghadam AT, Emami SA, Mohammadpour AH, Johnston TP, Sahebkar A. Drug interactions of cola-containing drinks. Clin Nutr. 2019;38(6):2545-51.
  2. Gonzalez-Dominguez R et al. Synergistic effects of sugar and caffeine on insulin-mediated metabolomic alterations after an acute consumptin of soft drinks. Electrophoresis. 2017;38:2313-22.
  3. van Orten-Luiten ACB, Janse A, Verspoor E, Brouwer-Brolsma EM, Witkamp RF. Drug use is associated with lower plasma magnesium levels in geriatric outpatients; possible clinical relevance. Clin Nutr. 2019;38(6):2668-76.