Epidemiologic Evidence for Association between a High Dietary Acid Load and the Breast Cancer Risk

Background and Purpose: Dietary acid load contributes to metabolic acidosis, leading to inflammation and cell transformation, potentially implicated in cancer development. Albeit an increased risk of recurrence among BC survivors was reported for a high acid load, the epidemiologic evidence associating diet-dependent acid load and cancer risk, particularly for breast cancer (BC), is still very limited. Therefore, we have explored in the present study its role in BC risk. Methods: A case-control study was performed on 1461 patients (572 BC cases and 889 age-frequency matched controls), through a multi-topic questionnaire, which included a food frequency questionnaire. Food-derived nutrients were calculated from available databases. The dietary acid load was calculated based on existing measures as potential renal acid load (PRAL) score and net endogenous acid production (NEAP) score. Odds Ratios (ORs) and their 95% confidence intervals were estimated by logistic regression, adjusting for potential confounders. Results: We found direct associations between dietary acid load and BC risk. Both scores were significantly associated (OR=2.46 and OR=1.78 for highest PRAL and NEAP, respectively). A positive BC family history involved higher risks (OR=6.14 and OR=3.38 for highest PRAL and NEAP, respectively). Linear trends were found in all overall and stratified analyses. Conclusions: Results suggest that a low acid load dietary style may reduce BC risk since both PRAL and NEAP scores were directly associated with meat intake and inversely associated with plant-based foods intake. The findings agree with studies focused on food groups and dietary patterns. Further studies are needed to confirm these findings.


Introduction
Breast cancer (BC) is the leading malignancy among Uruguayan women [1], with the highest incidence rate in South America and close to North American pictures [2,3]. An independent effect of modifiable risk factors as dietary patterns, lifestyle factors, macro-and micronutrient intake, physical activity, tobacco smoking, and weight gain on the

Material and Methods
We combined two databases previously used in epidemiologic studies on BC in the Uruguayan population [26][27][28][29]. Those studies were carried out in Uruguay during 1996-2004 in Montevideo's principal public hospitals (Pasteur, Maciel, Clinicas, Oncology Institute) and one private hospital. Formal consents were not required for this type of study at that time. The studies were conducted after being authorized by Directors receiving ethical approval in each participant institution. With a similar structure, both databases enabled us to study a total sample of 1461 participants (572 BC cases, 889 controls). Each one is briefly described as follows and the flowchart in Figure 1 summarizes the methodology of both studies.

Public Hospitals
As part of multi-site epidemiologic research, 480 incident BC cases were eligible for the study period. Nineteen patients rejected an interview, leaving 461 cases to be included (response rate 96.0%). In the same period and hospitals, 685 admitted patients afflicted with diseases unrelated to smoking and drinking were eligible for the study. Eighteen patients rejected the interview, leaving 667 controls (response rate 97.4%). Trained social workers, unaware of the study objectives, worked at the hospitals in two phases: First, they looked routinely for newly diagnosed cancer patients, working with Medical Records personnel's collaboration. Second, they contacted patients who were eligible to be matched by the age-frequencies of the cases, as well as their urban/rural residence, and their regions (Montevideo/Other counties). After obtaining consent for the study, all the participants underwent an in-person interview in the hospitals. No proxy interviews were conducted. According to Uruguayan law, patients admitted to public hospitals were people with low incomes, from the whole country, and free access to medical services. Considering the population's features, they were good representatives of a third world country, different from the community admitted at the private health institution.
A 2:1 ratio for controls/cases was projected, but external economic factors determined budgetary cuts during the early 2000s and further. Consequently, a perfect matching was not possible to be completed.

Private Hospital
An epidemiologic study on BC conducted in 1999-2001 at a pre-paid medical institution in Montevideo (IMPASA) derived 116 incident BC cases and 223 controls women having normal mammography (BI-RADS 1) [30] ≤1 year before the interview. One control and two cases refused the interview, and three cases were excluded for medical reasons, finally achieving 111 cases and 222 controls (response rates: 95.7% and 99.6% respectively). They were age-matched (± five years). All participants, inhabitants of Montevideo (the capital city), did not remain at the hospital during the interviews. Women were <85 years old and belonged to mid-to-high socioeconomic strata. After appointments made by phone, interviews were face-to-face conducted in a hospital's office by a trained nurse blinded by major risk factors.

Interviews and Questionnaire
Participants answered a structured questionnaire which included: socio-demographic variables; occupation; BC history in 1º-2º degree relatives; self-reported height and weight five years before the interview; smoking and alcohol; a history of "mate," tea and coffee drinking; menstrual-reproductive events; and a food frequency questionnaire (FFQ) of 64 items, representative of Uruguayan diet, focused on food consumption five years before the interview. Proxy interviews were not accepted. The FFQ was not validated; however, it was tested for reproducibility [31], allowing individual energy estimation. All dietary questions were open-ended. Local tables of food composition were used for estimating energy and nutrients [32].

Estimation of Dietary Acid Load
We calculated the diet-dependent acid load using previously defined formulas [15,16] and applied in other epidemiologic studies on BC risk [21,25] as well as on recurrence [23,24]: potential renal acid load (PRAL) and net endogenous acid production (NEAP). These measurements were calculated as follows: The PRAL score takes into account the intestinal absorption rates for protein, potassium, calcium, magnesium, and phosphate and has been validated against urine pH in healthy adults [15]. The rate of sulfuric acid production from protein metabolism and the rate of bicarbonate generation from metabolism of intestinally absorbed potassium salts of organic acids are major and highly variable components of the NEAP score, developed by Frassetto et al. [16]. These authors stated that "by considering both, the acidifying effect of protein and the alkalinizing effect of potassium (organic anions), NEAP can be predicted with confidence from the readily available contents of only 2 nutrients in foods." In addition, they found that PRAL and NEAP were highly correlated (r=0.84, p< 0.001). A negative PRAL or NEAP value reflects an alkaline-forming potential, whereas a positive value reflects an acid-forming potential.
An analysis program was compiled to calculate energy, which made the sum of all individual values. Each one was obtained after multiplying the number of servings/year by the ratio calories of the serving/100 g of each, divided by 365 days. Most typical or average servings of solid foods are within the range of 100-150 g. Since iron intake showed a high correlation with energy, we calculated an iron density expressed as daily mg of the mineral/kcal*1000.

Statistical Analysis
The questionnaire variables were initially continuous. When necessary, they were categorized for analysis purposes. Aside from primary descriptive analyses (frequencies, mean values), we calculated Odds Ratios (ORs) and 95% confidence intervals (95% CI) by unconditional logistic regression [33]. Terms for potential confounders were included in the multivariate analyses. The equations included terms for: age (categorical), residence (binary), education (categorical), age at menarche (categorical), menopausal status (binary), age at first live birth (categorical), number of live births (categorical), age at menopause (categorical), family history of BC (binary), BMI (continuous), smoking intensity in pack-years (continuous), alcohol status (categorical) and energy intake (categorical) as independent variables, and cancer (yes/no) as the dependent one. Possible heterogeneities in the stratified analyses were explored through likelihood-ratio tests. The STATA software was used to make all calculations (Release 10, StataCorp LP, College Station, TX, 2007). Table 1 shows the distribution of cases and controls according to selected socio-demographic variables. Although participants were not completely matched, an adequate age distribution was achieved (p = 0.87). There were more rural cases than controls (12.9 % vs. 9.4%, resp.). Most traditional BC risk factors (family history of cancer, reproductive variables) displayed significant or marginal differences between cases and controls. On the other hand, educational level and BMI did not. Some selected lifestyle variables were analyzed and presented in Table 2. Energy, red meat, and alcohol intake were directly and significantly associated with BC risk. On the other hand, fruits, vegetables, and the three infusions (coffee, tea, and 'mate') were inversely and significantly associated with the risk.  Table 3 displays the mean values of PRAL and NEAP scores and their components comparing cases and controls. Both mean scores were higher among BC cases. Regarding the details, differences in protein and phosphorus intakes were significantly higher among cases. However, potassium, magnesium, and calcium intakes were not significantly different between cases and control groups.  Table 4 displays the adjusted ORs for both acid load scores. The highest vs. lowest quartile of PRAL derived a significant estimate (OR=2.46) with a highly significant trend (p trend <.001). The same applies to the NEAP score: both risk and trend estimates were significant (OR=1.78, p trend <.001). The analyses stratified by family history of BC showed higher risks with a positive history: PRAL had an OR=6.14 (p trend <.001), and NEAP had an OR=3.38 (p trend = .001).

Discussion
We found direct associations between dietary acid load and BC risk. The ORs for the highest vs. lowest quartile were significant regarding both scores employed and their linear trends: PRAL (OR=2.54, ptrend <0.001) and NEAP (OR=1.78, ptrend <0.001). Stratified analyses done in a recent prospective study [21] found suggestively stronger associations between PRAL and BC in women who had a sister diagnosed with BC before age 50 years, especially for ER-negative BC. Albeit we had no data on ages of relatives' cancer cases nor hormonal receptors of patients, our stratified analyses by family history also showed a risk increase for PRAL score when this history was present, compared to its absence (OR =6.14 vs. OR=2.12, respectively). Besides, the NEAP score displayed similar associations (OR =3.38 vs. OR=1.58, respectively). Having found significant heterogeneity between the absence/ presence of BC family history, this might imply gene-dietary interactions to be considered for future studies.
The existing epidemiologic evidence related to acid load and cancer risk is limited, restricted to BC [21,23,25], colorectum [34,35], and pancreas [36]. The research works related to BC are not consistent enough [21,[23][24][25]. There were positive associations for acid load with BC risk among American and Puerto Rican women [21,23], but it was not associated among Iranian ones [25]. Some studies reported results about the influence of the acid load on the metabolic condition (hyperinsulinism and/or diabetes) [34,35] and the association with BC recurrence and survival [23,24]. One of these, identified acid load as a novel dietary factor that may lead to inflammation and hyperglycemia [23]. Therefore, a dietary acid load may contribute to inflammation in cancer patients.
It has been claimed that dietary acid load can contribute to metabolic acidosis if the acid-base balance is not correctly adjusted. Metabolic acidosis is a condition characterized by a slight decrease in blood pH [14], and feeding is one of the main factors to produce such a situation. In humans, arterial and venous blood have a pH of 7.35-7.4 and 7.20-7.35, respectively [37]. Besides, there is growing evidence showing that the fluid pH in the interstitial space around metabolic tissues is easily reduced due to weaker pH buffering capacity than that in the cytosol and blood circulation. Whereas the arterial blood pH is strictly regulated due to strong pH buffers such as albumin and hemoglobin, interstitial fluids have only relatively weak pH buffers, bicarbonate, and phosphate, enabling a pH range of 6.60-7.60 [38]. Such interstitial pH reduction might initiate a metabolic dysfunction [39]. In contrast, several nutrients bring benefits in maintaining the interstitial fluid pH within the normal range by improving buffering capacities, suppressing proton production, and activating proton transporters, strengthening the effect of appropriate diet on metabolic health [38].
The excessive consumption of acid precursor foods (such as meat, cheese, and eggs, which are sources of phosphorus and proteins) leads to acid-base balance volubility. If this situation occurs in a prolonged, or chronic way, low-grade metabolic acidosis can become significant and predispose to metabolic imbalances. [13,14,40] Regarding these points, it is well known that metabolic acidosis can cause tissue damage, which can further initiate inflammation [23]. Inflammatory process induces oxidative stress and reduces cellular antioxidant capacity. Overproduction of free radicals reacts with cell membrane fatty acids and proteins, impairing their function permanently. Besides, free radicals can lead to gene mutation and DNA damage that predispose to the development of cancer and age-related disorders [41].
Cancer patients have reduced capabilities for adjusting their acid-base balance [42]. Several metabolic adaptations observed in cancer are recognized as similar to the perturbations observed in diabetic patients [37]. Albeit acidity has more than one source [43], the lactate derived from cancer cells suppresses T cell and NK cell function [44]. Moreover, metformin -an insulin sensitizer-frequently used to reduce hyperglycemia in diabetic patients, has been accepted to reduce cancer incidence [45] and improve survival [46].
An adequate intake of phosphorus for adults in the general population is 550 mg/day. However, European countries' mean intake is higher than 2-fold such value (1000-1767 mg/day) [47]. By the same token, our study population sample showed a mean intake of 766 mg/day (760 mg/day among controls), which is higher than the adequate consumption, based on the quoted reference numbers. Dietary phosphorus has increased over time [48], mainly due to phosphorus-containing additives in food manufacturing and processing [49]. Furthermore, there is evidence that cellular phosphate burden from phosphate toxicity is a pathophysiological determinant of cancer cell growth [50]. According to these authors, a dysregulated phosphate homeostasis can be associated with the genesis of various human tumors.
Besides, a diet that includes phosphorus-enhanced foods can probably add another 600-800 mg to the overall daily intake and non-negligible amounts of sodium chloride (NaCl) [51]. Indeed, its intake is reported to be an independent predictor of plasma bicarbonate concentration. NaCl may exert approximately 50-100% of the acidosis-producing effect of the dietary acidic load and is considered a predictor of diet-induced low-grade metabolic acidosis [52]. Besides, an increasing NaCl dose-dependently decreases blood pH and plasma bicarbonate levels, independent of the partial pressure of CO2, creatinine clearance, and dietary acid load. [13] On the other hand, potassium, magnesium, and calcium are precursors of bases. Therefore, the main foods that release precursors of acids into the bloodstream are mostly of animal sources, and foods that are precursors of bases are mainly those of plant sources [14]. Table 3 shows the components of acid load scores, discriminated by animal/plant source, as it has been previously discussed.
In addition to dietary phosphorus, or to the use of phosphorus-containing additives, the picture turns even more complicated when it was taken into account the previous relationships among acid load, chronic inflammation, and oxidative stress: The interaction of free radicals with polyunsaturated fatty acids of cell membrane causes lipid peroxidation and subsequent cell damage, leading to leakage of intracellular phosphorus into serum [53]. Therefore, Western-like dietary patterns may enhance a vicious cycle involving lipoperoxidation, inflammation, and metabolic acidosis.
Potential diet-gene interactions involving hormonal mechanisms, susceptibility genes, polymorphisms, and epigenetic regulatory mechanisms could be present regarding the dietary acid load and BC. In addition, some links have been uncovered between cancer and circadian disruption, in particular related to BC [54]. Recent research discovered a profound disruption of the circadian clock and diurnal transcriptome when hypoxic cells are permitted to acidify to recapitulate the tumor microenvironment [55]. These authors also found that buffering against acidification or inhibiting lactic acid production fully rescues circadian oscillation.
As for the study's strengths, cases and controls were face-to-face, directly interviewed by the same trained personnel in the same hospital settings, and the studied population included subsets coming from the whole country and belonged to different socio-economic-cultural strata. Besides, times of data collection were coincident.
On the other hand, the present study may share some biases that are common to case-control studies. Withinperson variability over the study period may be a source of information bias. Selection bias was limited by the nearly full participation of the identified cases and controls (rates ~97%), favored by the interview during the hospital stay. Dietary habits were relatively stable in the Uruguayan population, and patients were asked to report any relevant dietary changes occurring during their life. Furthermore, a recall bias related to dietary habits should be negligible in our study population, as the awareness of BC's dietary hypothesis was very limited. Although the FFQ was not validated, it was satisfactorily reproducible [31]. Mineral estimations become one of the limitations of the present study since they were based on average serving sizes rather than actual food sizes. Besides, we could not exclude confounders' role by other dietary factors, such as other constituents of animal foods, the effects of different cooking methods, and the mineral contents in water. For example, highly processed and manufactured foods, which involve the use of phosphorus-containing additives, were not included in the FFQ. Therefore, no estimations of phosphorus and sodium from these potential sources were carried out.

Conclusion
Both calculated dietary acid load scores (PRAL and NEAP) were found directly and significantly associated with BC risk. The risk increase was seen irrespective of menopausal status, and it appeared to be more robust having a positive BC family history. Results suggest that a low acid load dietary style may reduce BC risk since both PRAL and NEAP scores were directly associated with meat intake -which characterize westernized dietary styles-and inversely associated with plant-based foods intakewhich characterize Mediterranean and vegetarian-like styles-. The findings agree with studies focused on food groups and dietary patterns. Potential diet-gene interactions involving hormonal mechanisms, susceptibility genes, polymorphisms, and epigenetic regulatory mechanisms could be present regarding the dietary acid load and BC. Dietary and nutritional optimization, for example, regarding the achievement of a better dietary potassium/sodium ratio, is still not a highly emphasized area of health care. However, it has a great potential to attenuate the impact of prevalent chronic diseases in modern society (including BC) and improve population health [56]. In light of the implications of dietary acid load, although further studies are needed to confirm the present study's findings, we are convinced that from a medical and ethical viewpoint, it is justified to recommend -as we did one decade ago [57]-certain nutritional changes to women, because in the meantime no adverse side effects are expected to occur.

Author Contributions
A.R. was responsible of the conceptualization, investigation, methodology, data analysis, draft writing, visualization, and final review and editing. W.M.L. participated in the visualization, the draft writing, and final review and editing. B.M. and J.M.C. participated in the draft writing, the final review and editing. All authors have read and agreed to the published version of the manuscript.

Funding
The authors received no financial support for the research, authorship, and/or publication of this article.

Data Availability Statement
The data presented in this study are available on request from the corresponding author.

Conflict of Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.