The relationship between the insulinemic potential of diet and lifestyle and risk of breast cancer: a case-control study among iranian adult women

Breast cancer (BC) is the most prevalent cancer, accounting for 25% of all female-related cancers worldwide, and second cancer with the highest mortality rate in women. The disease affects about 1.4 million people annually [1]. The well-established risk factors for BC, including age, genetic mutations, premature menstruation, late pregnancy, late menopause, hormone therapy, oral contraceptives, and cancer family history, are mostly unmodifiable [2]. However, unhealthy lifestyles, including physical inactivity, obesity, and inappropriate diet, are modifiable risk factors that play an important role in cancer pathogenesis [3]. In 2018, The World Cancer Research Center reported that choosing a healthier lifestyle and changing eating habits could prevent 4 million new people from developing cancer [2].

One of the predisposing biological agents for cancer incidence and its development is insulin-related disorders, particularly hyperinsulinemia, that play a crucial role in tumor development through insulin-like growth factor-1 (IGF-1) [4]. Obese and inactive subjects are more prone to imbalanced insulin homeostasis, and unhealthy dietary patterns can lead to hyperinsulinemia and insulin resistance (IR) [5, 6].

Recently, Tabung et al. have proposed dietary and lifestyle indices to predict hyperinsulinemia and IR using serum connecting peptide (C-peptide) and Triglyceride (TGs) to high-density lipoprotein-cholesterol (HDL-C) ratio, respectively. The empirical dietary index for hyperinsulinemia and IR (EDIH and EDIR) includes only food groups related to insulin biomarker responses. In contrast, the empirical lifestyle index for hyperinsulinemia and IR (ELIH and ELIR) is composed of physical activity (PA) and body mass index (BMI), and also food groups correlated with insulin biomarkers.

Previous studies investigated the link between these insulinemic dietary and lifestyle indices and the risk of several cancer types [7‐14] regarding the well-established link between insulin disorders and cancer. A recent study showed that all EDIH, EDIR, ELIH, and ELIR are positively associated with the risk of hepatocellular carcinoma (HCC) [7]. Also, two studies demonstrated that a higher EDIH score increases prostate cancer risk [8, 9]. The Wang et al. study revealed that interventions to reduce the insulinemic potential of diet and lifestyle have protective effects against digestive system cancer [10]. Furthermore, there is a significant relationship between EDIH and the higher incidence of colorectal cancer (CRC) and poorer survival in patients with CRC [11, 12]. However, Lee et al. did not observe any significant relationship between EDIH and EDIR and multiple myeloma (MM) [13].

Accordingly, most previous studies observed a positive relationship between the EDIH and various types of cancer. However, despite the benefits of ELIH and ELIR as lifestyle scores indicating collective effects of diet, PA, and BMI, few studies have examined their association with cancer risk. Also, to our knowledge, there is no study exploring these indices’ relationship with the BC risk. So, we aimed to perform a case-control study to investigate the possible association of dietary and lifestyle indices for hyperinsulinemia (EDIH, ELIH) and insulin resistance (EDIR, ELIR) with BC risk in a sample of Iranian adult women.

The baseline characteristics of participants, including demographic and lifestyle variables, medical history, and dietary intakes, are indicated in Table 1. Participants’ mean ± SD of age and BMI were 47.9 ± 10.3 years and 29.4 ± 5.5 kg/m2, respectively. Individuals in the case group had higher age, first pregnancy age, postmenopausal women percent, and cancer family history, whereas they had lower anti-inflammatory drug consumption and vitamin D supplement intake than the control group (P<0.05). There were no significant differences between cases and controls in insulinemic indices, including EDIH, EDIR, ELIH, EDIH, and other variables.

Table 2 shows the correlation coefficient of insulin indices. There was a significant correlation between EDIH and EDIR(r=0.263), EDIH and ELIH(r=0.215), and EDIR and ELIR(r=0.851).

The association of insulin indices with the risk of BC is presented in Table 3. A significant positive association was observed between higher EDIH score and risk of BC in the highest compared to the lowest tertiles in the age and age first pregnancy model (OR: 2.85; 95% CI: 1.05 – 3.23, P for trend=0.059). After adjusting for confounding factors in the final model, participants with the highest EDIH score had higher odds of BC than those with the lowest EDIH score (OR: 2.24; 95% CI: 1.21 – 4.12, P for trend=0.016). However, based on all logistic regression models, there is no significant association between ELIH, EDIR, and ELIR and the risk of BC.

Table 4 showed the adjusted OR (95% CI) of BC per one SD increment of insulin indices in subgroup analysis based on three variables, including menopausal status, cancer family history, and OCP use. Each SD increase in the EDIH score was associated with an increased risk of BC among postmenopausal women (OR: 1.74, 95% CI: 1.13–2.69) and those who do not use OCP (OR: 1.44;95% CI:1.02 – 2.04). Also, each SD increment of the ELIH score was related to a higher OR of BC based on postmenopausal status (OR; 1.98; 95% CI: 1.35 – 2.89), family history of cancer (OR: 1.94; 95% CI: 1.10 – 3.42), and no use of OCP (OR: 1.46; 95% CI: 1.00 – 2.12). The dietary or lifestyle indices for insulin resistance (EDIR and ELIR) showed no association with BC odds among subgroups.

Current research provides the first evidence about the association between EDIH, EDIR, ELIH, and ELIR and the risk of BC. Based on our findings, EDIH was related to the higher risk of BC based on fully adjusted models. Furthermore, elevated EDIH level was associated with higher BC risk based on postmenopausal and non-using OCP based on subgroup analysis. Moreover, ELIH increment in postmenopausal women, those with a family history of cancer, and not using OCP was associated with an increased risk of BC.

Our findings are consistent with some previous studies investigating the possible association of the insulinemic potential of diet and lifestyle with the risk of various cancers. There is evidence that the higher score of EDIH was associated with a 26% increased colorectal cancer risk in men and women [18]. Similarly, Yang et al. have observed a positive association between higher scores of EDIR and increased risk of hepatocellular carcinoma [19]. Another study has reported that EDIH and ELIH were significantly associated with digestive tract cancers [20]. Furthermore, two studies claimed that participants with hyperinsulinemic diets had a greater risk of advanced and fatal prostate cancer [9, 21]. Although EDIR was associated with increased multiple myeloma risk in another study, EDIH did not show any significant relationship with the risk of multiple myeloma [13]. Although Cheng et al. declared that a higher score of EDIH is potentially related to colon cancer risk, no significant association was observed between a potential insulinemic diet and the risk of colon cancer recurrence, survival, or mortality in patients with late-stage (III) colon cancer [22]. EDIH, EDIR, ELIH, and ELIR are indicators that have been implicated in the pathogenesis of various cancer and complications [23, 24]. Insulin is a key regulator hormone in cell growth and energy metabolism. Up-regulation of insulin secretion and insulin resistance results in elevated IGF-1 production and bioavailability, resulting in cell proliferation and tumor growth [25]. It should be noted that dietary insulin indices are related to insulin resistance [26]. Moreover, some evidence shows that the dietary hyperinsulinemia index potentially affects cancer progression, especially in those with a low level of PA [18].

The subgroup analyses have revealed a positive association between EDIH and ELIH scores and the risk of BC in women who did not use OCP. Previous studies demonstrated that OCP use might increase BC’s risk [27]; however, this increased risk is slight and highly depends on different underlying factors like age, genetics, duration of use, and formulation of pills [28‐31]. Besides, most of those studies included women who used OCP in the 1980s or earlier, when the pills tended to have a higher hormone content than they do now [32]. Furthermore, according to previous studies, OCP consumption increases insulin resistance risk [33, 34]. It is possible that in patients without a history of OCP use, due to fewer metabolic disorders than individuals with this history, the role of hyperinsulinemic diet and lifestyle in the development of metabolic disorders and increased cancer risk for each SD increase in scores is more prominent and therefore the overall relationship is significant.

A higher EDIH and ELIH score also increased the risk of BC in postmenopausal women. Increased endogenous estrogen status [35] is associated with postmenopausal breast cancer risk [36]. Although insufficient evidence is available on the possible effect of pre-and post-menopausal cancer, the risk of uterine, ovarian, and breast cancers increases with aging, especially after age 55 years old [37]. Lifestyle changes, anthropometric changes, including elevated BMI and adiposity, low physical activity, and hormone replacement therapy, increase by getting older and in those experiencing a postmenopausal period, can enhance BC risk [38, 39]. For instance, Toklu and Nogay indicate that unhealthy dietary patterns and eating deep-fried red meat, a sedentary lifestyle, and a high BMI, especially during the postmenopausal period, are risk factors for BC [40].

In addition, according to subgroup analysis results, a significant direct association of ELIH with the risk of BC was observed in subjects with a history of cancer. There is some evidence on the relationship between family history and the risk of BC, which expresses that first-degree family history can promote invasive BC risk [41]. For instance, Reiner et al. reported that having a first-degree relative with BC is associated with higher BC risk [42]. Also, Ahern et al. demonstrated that first- and second-degree relatives can estimate BC occurrence [43]. As well as hereditary has an important role in breast cancer incidence [44], family members may have a similar unhealthy lifestyle and dietary behaviors.

Strengths of the current study include the novelty of the investigation and data gathering from well-documented research centers. Furthermore, we have used valid and reliable questionnaires for dietary intake and physical activity evaluation. In addition, multiple logistic regression models, considering various confounders, were used in our study. However, we acknowledge some limitations.

The main limitation of the current study was the case-control study design which cannot provide a causal relationship. Also, our trained dietitians were not masked about cases and controls; however, during training, focus was given to minimize possible information bias to collect the data without mental background about participants’ cancer status. Since the study is conducted in Tehran, Iran, we cannot generalize the results to the other women population. Also, recall bias is possible due to using FFQ for dietary assessment. Furthermore, conducting stratified analyses based on three variables may increase the possibility of showing chance findings. However, the findings about EDIH, EDIR, and ELIR were repeated. Only ELIH showed a positive association with cancer among postmenopausal women, those with the cancer family history, and those with no OCP consumption. So, these findings should be tested among mentioned subgroups in other studies for a better perception of the ELIH-cancer relationship.

In conclusion, our results showed a possible link between EDIH and higher BC risk. Also, higher EDIH and ELIH scores were associated with a higher risk of BC in women in the postmenopausal period, having a family history of BC, and those who do not use OCP.