Low-energy diets are usually associated with a global improvement in plasma lipid concentrations [13,14]. Bariatric surgery also improves plasma lipid levels [15,16,17].

The study by Kiortsis et al., involved 62 healthy obese patients (21 men aged 32 ± 9.6 years and 41 women aged 37 ± 14.6 years) consuming a low-energy diet for six months. In considering the whole population, the authors observed a body weight loss of 7.5% versus baseline and a statistically significant decrease in plasma levels of total cholesterol, LDL cholesterol, and triglycerides (p < 0.001), while no changes in Lp(a) levels were detected. However, when considering individuals with high Lp(a) values (>20 mg/dL) at baseline, a 17.6% reduction in Lp(a) (p < 0.05) was observed, which was closely related with baseline Lp(a) levels (r = 0.81 p < 0.001), but not with the changes in anthropometric measurements that occurred during weight loss [18]. Therefore, a low-calorie diet that induces weight loss in individuals with obesity may have positive impacts on serum Lp(a) levels, particularly in patients with elevated pretreatment concentrations of Lp(a). However, this effect is influenced by large interindividual variability and depends on the characteristics of the patients.

In the study by Berk et al. [19], after a similar 3–4 month calorie-restricted diet, patients with type 2 diabetes and obesity (n = 131) experienced a significant decrease in body weight (−9.9%), while their Lp(a) plasma levels increased by 14.8 nmol/L. Obese individuals (n = 30) or type 2 diabetics (n = 26) also experienced a decrease in body weight (−7%) associated with an increase in Lp(a) (+12.7 nmol/L), while Lp(a) did not change in obese individuals (n = 26) that underwent bariatric surgery despite considerable weight loss (−14%) [19]. The same research group carried out two further independent long-term clinical trials involving 293 overweight or obese individuals. The first study was designed as a prospective two-arm clinical investigation, including 82 patients undergoing a 7-week low-energy diet followed by Roux-en-Y gastric bypass and a 52-week follow-up (surgery group), and a control cohort of 77 patients undergoing a 59-week low-energy diet and exercise program (lifestyle group). The second study included a third cohort of 134 patients undergoing a 20-week low/very low-energy diet program (lifestyle cohort). In the lifestyle group and in the lifestyle cohort, the Lp(a) plasma level [median (interquartile range)] increased by 36% [14(7–77) vs. 19(7–94) nmol/L, p < 0.001] and 14% [50(14–160) vs. 57(19–208) nmol/L, p < 0.001], respectively. In contrast, in the surgery group, the Lp(a) levels dramatically decreased by 48% after intervention [21 (7–81) vs. 11(7–56) nmol/L, p < 0.001]. Remarkably, arachidonic acid and total n-3 fatty acids (FA) decreased after surgery but increased after lifestyle interventions, while plasma levels of total saturated FA remained unchanged after surgery but decreased after lifestyle interventions. However, the change in Lp(a) seemed to be independent of the weight loss [20].

In 2018, Gomez-Martin et al. [21] demonstrated that vertical sleeve gastrectomy and gastric bypass have a favorable impact on serum lipids at one-year post-surgery in women with high CV risk. This effect includes a reduction in total cholesterol, triglycerides, and oxidized-LDL, although there were no associated changes in Lp(a) levels. On the other hand, in a more recent and very large study (n = 702), Paredes et al. evaluated the 1-year metabolic impact of vertical sleeve gastrectomy. According to the findings of the study by Paredes et al., patients without metabolic syndrome (n = 372) experienced a decrease in Lp(a) levels (14.7 mg/dL vs. 12.3 mg/dL, p = 0.006) after vertical sleeve gastrectomy, while patients with metabolic syndrome did not (13.9 mg/dL vs. 14.6 mg/dL, p = 0.302). The regression model showed that older age and delta HDL-C significantly predicted the change in Lp(a), while the higher the number of metabolic syndrome components and the lower the estimated body fat percentage loss, the lower the odds of Lp(a) reduction after the intervention [22].

Even if the above cited studies are relatively small and designed differently, in particular with regard to diet composition, overall researchers could conclude that a low-energy diet is not sufficient to significantly modify Lp(a) levels in plasma, therefore, a more dramatic intervention is needed. The diet with a metabolic impact most like that which follows bariatric surgery is the very-low carbohydrate/ketogenic diet. In a single case of a 55-year-old triathlete with a BMI of 24.9 kg/m², the use of a very-low carbohydrate/ketogenic diet was associated with a plasma Lp(a) decrease ranging from 26% to 39% in different dietary intervention phases [23]. In a large trial involving 164 overweight or obese patients (BMI: 32.4 ± 4.8 kg/m²) with mixed dyslipidemia, a 20-week treatment with three different weight loss diets (one of which was a low-carbohydrate diet) exerted different effects on Lp(a). These diets were comprised of 20% protein and various contents of carbohydrates and saturated fats (low-carbohydrate diet = 20% carbohydrates, 21% saturated fats; moderate-carbohydrate diet: 40% carbohydrates, 14% saturated fats; high-carbohydrate diet: 60% carbohydrates, 7% saturated fats). At the end of the study, plasma Lp(a) was reduced by nearly 15% (−14.9%; 95% confidence interval (CI): −22.0 to −7.1) only in the group randomized to the low-carbohydrate diet. The low carbohydrate diet was also associated with a significant improvement in insulin-resistance, triglycerides, HDL-cholesterol, and adiponectin plasma levels [24]. It must be acknowledged that these findings were not confirmed in a recent 12-week randomized clinical trial examining the effect of a very low-carbohydrate/high-fat diet associated with a high-intensity interval training program in a cohort of 91 overweight individuals [25].

Even if these clinical trials are relatively small and short-term, their results supported the European Society of Atherosclerosis (EAS) consensus on hyper-Lp(a) management; to follow a low carbohydrate diet to reduce the concentration of Lp(a) by ~15% [26].

Dietary patterns abundant in animal-derived proteins and fats might be linked to increased risks of CV disease and related mortality when compared to diets rich in plant-derived protein [27,28,29]. Once again, the evidence supporting the intake of specific fatty acids and Lp(a) plasma levels is conflicting.

In overweight and obese individuals (n = 31), a 4-week plant-based diet was associated with an improvement in inflammatory and atherogenic biomarkers, among them Lp(a) (−32 nmol/L) [30]. The existing literature indicates that replacing saturated fats with an equivalent amount of unsaturated fats leads to decreased overall mortality [31]. Including nuts in the diet is a practical strategy to boost unsaturated fat intake, as it relates to lower all-cause mortality and mortality specific to CV diseases, in particular [32]. A diet abundant in walnuts, rich in alpha-linolenic acid, polyphenols, plant sterols, and tocopherol, demonstrated an overall enhancement in the blood lipid profile [33]. However, in a randomized, prospective, controlled, crossover, clinical study involving 194 healthy volunteers, an 8-week regimen of 43 g of walnuts daily did not exert any effect on Lp(a) concentration in plasma, even if other lipid fractions (such as non-HDL-C, apoB, total cholesterol, LDL-C, very LDL (VLDL) cholesterol, and triglycerides) were improved [34]. During a 16-week randomized controlled trial, 29 participants classified as overweight or obese (with a BMI of 25–40 kg/m²) were assigned to either consume 42.5 g/day of a mix of nuts (including cashews, almonds, macadamia nuts, Brazil nuts, pecans, pistachios, walnuts, and peanuts) or 69 g/day of isocaloric pretzels. There was no evidence that consumption of mixed nuts had an effect on LDL-C or Lp(a) throughout the intervention [35].

On the contrary, in a small study mainly enrolling Afro-Americans (n = 18/28), with a mean age of 48.3 ± 12.5 years (17 men, 11 women), Lp(a) plasma levels were negatively associated with absolute (grams/day) and relative (percentage of total calories) dietary saturated fatty acid (SFA) intake (R = −0.43, p = 0.02, SFA (% CAL): R = −0.38, p = 0.04), palmitic acid intake (R = −0.38, p = 0.05), and stearic acid intake (R = −0.40, p = 0.03) [36].

Coconut oil might also yield more favorable effects on Lp(a) concentration when compared to unsaturated oils with longer carbon chains. In a controlled crossover study involving young women and comparing two high-fat diets over a 3-week period [37], the consumption of a coconut oil-enriched diet resulted in a 17 mg/L reduction in Lp(a) levels. In contrast, the intake of highly unsaturated long-chain fatty acids led to a 25 mg/L increase in Lp(a) levels [37]. Despite being a saturated fat, it is essential to recognize that coconut oil primarily consists of medium-chain fatty acids (e.g., lauric acid), constituting 50% of its content. Consequently, it may elicit a different response in Lp(a) concentration. However, a more recent randomized, controlled, single-blinded, crossover, clinical trial that enrolled 40 healthy volunteers to investigate the short-term effect of a diet enriched in palm oil, cocoa butter, or extra virgin olive oil, with oleic acid primarily at the sn-2 position (66%, 75%, 87% sn-2 oleic acid, respectively) of the TG molecule, concluded that not one of the tested diets had a significant impact on Lp(a) levels [38].

Despite the slight benefit observed in Lp(a) levels with the consumption of coconut oil, the International guidelines advise that dietary unsaturated fats should constitute less than 10% of total energy intake [39,40]. This recommendation stems from the association of excess unsaturated fat consumption with significantly elevated morbidity and mortality rates related to cancers and CV diseases [27,31].

Trans-fatty acids are associated with adverse CV outcomes. Whether a part of this effect is mediated by an impact of trans-fatty acids on Lp(a) levels is not clear [41].

In a double-blind clinical trial involving 29 men and 29 women, the participants were randomized to eat one of four controlled diets for six weeks each, where fatty acids accounted for 39 to 40% of energy: (A) oleic (16.7% of energy as oleic acid); (B) moderate trans (3.8% of energy as trans monoenes, approximately the trans content of the U.S. diet); (C) high trans (6.6% of energy as trans monoenes); (D) saturated (16.2% of energy as lauric, myristic, and palmitic acids). The saturated diet significantly reduced Lp(a) levels from 8% to 11%. Compared with the oleic diet (A), the trans diet had no adverse effect on Lp(a) levels in the whole cohort. However, the subgroup of individuals with higher Lp(a) levels at baseline (≥30 mg/dL) responded to the high trans diet (C) with a slight, though significant increase (+5%) in Lp(a) levels compared to the oleic (A) and moderate trans (B) diets [42].

In a small, randomized, clinical trial involving 31 young men, the consumption of hydrogenated soybean oil led to a notably higher level of Lp(a) compared to a diet primarily sourced from butter [43]. On the contrary, in a randomized, crossover study with 49 hypercholesterolemic patients following a 6-week diet rich in either butter or margarine, there was no observed change in Lp(a) [44]. Intake of meals high in specific dietary fatty acids can increase postprandial plasma lipids differently [45,46,47,48], including Lp(a) concentration. In a clinical trial enrolling healthy, young men, 16 volunteers were asked to sequentially consume five test fats dominated by (approximately 43% g/kg) stearic, palmitic, oleic, C18:1 trans, or linoleic acid incorporated into meals (1 g fat/kg body weight) after a 12-h fast, in random order on different days, separated by 3-week washout periods. Blood samples were drawn before, and 2, 4, 6, and 8 h after eating. Lp(a) plasma levels were found to increase after each supplementation, except after oleic and C18:1 trans consumption. On the contrary, oleic and C18:1 trans supplementation was associated with less area under the plasma Lp(a) concentration curve compared to those measured after stearic and palmitic acid intake (p < 0.003). So, long-chain stearic and palmitic acids led to significant increases in postprandial Lp(a) levels after an oral fat test in young, healthy men [49].

In conclusion, based on the available data, mainly obtained in small and short-term clinical trials, a mild increase in plant-derived saturated fatty acids could mildly decrease the Lp(a) plasma levels, while trans-fatty acid rich foods should be avoided. The available evidence could not be translated to a suggestion to increase saturated fatty acids in a diet aiming to improve LDL-C and reduce CV risk, however, it suggests that in individuals with high Lp(a) levels, an extreme reduction of saturated fatty acids is not mandatory and probably also negative.

A variable consumption of beverages such as coffee, tea, and alcohol worldwide may be correlated with CV outcomes [50,51,52]. Thus, the potential effects of popular beverages on Lp(a) concentration deserves consideration.

In 15 mildly hypercholesterolemic adults (mean LDL-C = 135 mg/dL) consuming five cups/day of black tea prepared using 180 mL of water for each serving, Lp(a) decreased by 16% as compared with placebo [53]. Of course, the number of individuals was too small to be conclusive with regard to the potential use of black tea as a Lp(a) lowering tool and larger long-term studies should be designed to confirm or disprove this observation. In another study, 53 volunteers with diabetes were randomly assigned to drink either black tea (n = 26) or Hibiscus sabdarrifa tea (n = 27), by using 2 g of tea sachet with 240 mL of boiling water for each serving, twice daily for 1 month. The Lp(a) concentrations remained unchanged from the baseline value of 26 mg/dL in both study groups [54]. Further research is needed, especially with respect to the number of commercially available tea preparations.

A rapidly increasing body of evidence recognizes the potential benefits of coffee in relation to CVD [55,56,57,58], but its effects on Lp(a) plasma levels remain unclear. Moreover, the exact mechanism by which coffee or single coffee components affect Lp(a) levels is yet to be clarified. The type of coffee and method of preparation appear to be important in determining the effect on Lp(a); in fact, coffee diterpenes present in unfiltered coffee brews are among the few dietary constituents that may modulate Lp(a) levels [59]. According to the findings of a systematic review and meta-analysis, the consumption of coffee or coffee diterpenes was associated with either a reduction in Lp(a) of 11 mg/dL (6 trials, 275 individuals), or no effect (2 trials, 56 individuals) [60]. However, it must be recognized that this metanalysis was affected by a large inter-study heterogeneity as regards study design, type of intervention, coffee source, and method of coffee processing [60].

A cross-sectional study with 309 volunteers showed that serum Lp(a) was elevated in chronic boiled coffee drinkers, who had a median Lp(a) of 13.0 mg/dL (range 0–130) compared with filter coffee drinkers who had a median Lp(a) of 7.9 mg/dL (range 0–144). The effect of coffee on Lp(a) is complex and may follow a biphasic time course, that is to say that whilst coffee may have a short-term beneficial effect in reducing Lp(a), in the longer term it may prove to be detrimental [61]. On the other hand, in the large UK Biobank database (n = 447,794 participants aged 37–73 years) no association was observed between coffee or tea intake and Lp(a) plasma levels [62].

In a cross-sectional study involving 300 middle-aged men, the Lp(a) concentrations in subgroups with low (<39 g/week), intermediate (39–132 g/week), and high (>132 g/week) ethanol intake were 137, 109, and 94 mg/L, respectively (P between groups < 0.05). Interestingly, abstainers exhibited a higher Lp(a) concentration (median, 206 mg/L) compared to drinkers [63]. However, in another cross-sectional study of 402 subjects with untreated hypertension, those with light (1–20 g/d), moderate (20–50 g/d), and heavy (>50 g/d) ethanol consumption showed 21%, 26%, and 57% lower median Lp(a) concentrations, respectively, compared to abstainers and occasional drinkers [64]. Notably, red wine consumption appears to have a greater ability to decrease Lp(a) levels than white wine. In a study involving 20 healthy male volunteers, the daily intake of 200 mL of red wine for 10 days resulted in a reduction in Lp(a) levels from 18.6 to 13.2 mg/dL (p < 0.001), whereas a similar effect was not observed with white wine after a 6-week washout period [65]. Of course, the study was too small and short-term to furnish strong evidence that red wine more effectively reduces Lp(a) plasma levels than white wine. In a 4-week randomized crossover study in 67 men with high estimated CV risk, Lp(a) levels were compared after the ingestion of red wine (30 g alcohol/day), the equivalent amount of dealcoholized red wine, and gin (30 g alcohol/day) [66]. The Lp(a) level fell from 54.4 mg/dL (baseline value) to 50.2 mg/dL, only after the intervention with red wine [66]. The conflicting results of the different studies suggest the need of more in depth research on larger cohorts, focusing on the different kinds of alcohol consumed (beer, red wine, white wine, shots of spirits). In any case, the adverse health effects of more than minimal alcohol intake may well outweigh any potential benefit in lowering Lp(a) levels [67,68].