Risk Factors of Prostate Cancer Among Patients Diagnosed at the University College Hospital, Ibadan, Nigeria: A Case-Control Study

Background and aims: Prostate cancer (PCa) is the second most commonly diagnosed cancer in men and its incidence is higher among black folks for reasons yet unclear. Although few risk factors have been linked to the development of PCa among Nigerian men, it remains unclear whether these risk factors modulate the odds of PCa. Methods: The case-control study comprised forty-three cases and one hundred and twenty-nine age-matched controls (± 5 years) without PCa by prostatic specific antigen (PSA) examination. Conditional logistic regression analysis was employed to identify risk factors associated with PCa at P < 0.05 using SPSS 20. Results: Increasing age ( adjusted odds ratio [AOR]: 2.8; 95% CI: 1.7-4.5) was the strongest risk factor for PCa, followed by increasing age at first sex ( AOR: 1.2; 95% CI: 1.1-1.4) and sexual activity ( AOR: 0.2; 95% CI: 0.1-0.9) as a protective risk factor against PCa in this Nigerian Population. Conclusion: The study confirms age as a recognized risk factor and backed evidence for other hypothesized risk factors. The study recommends findings with other confirmatory studies that can help to guide policies for better health care decisions among Nigerian men and interventions centered on routine screening for PSA with an emphasis on the elderly clinic is encouraged.


Introduction
The prostate is an organ that covers the prostatic urethra and is located below the bladder 1 .There are ways to cause medical problems for the prostate and they include; enlarged prostate, prostatitis, and prostate cancer (PCa). 1 PCa is the fourth major cancer globally, accounting for 1.3 million (7.1%) of the overall burden of cancer incidence. 2It is the second most frequently diagnosed form of cancer in men, accounting for 13.5% of the overall incidence of cancer, and also ranks second in both developed and developing countries in incidence. 2The number of new cases ranges by more than 50 times, with the highest rates occurring in North America, Australia, and North/Central Europe; in Southeast and South-Central Asia and Northern Africa, the lowest rates occur. 3tudies have shown that the burden of PCa in developing communities in southwestern Nigeria is prevalent among ages between 46-99 years with a peak incidence in the ≥ 70 years age group and a prevalence of 1.05%. 4,5he risk factors of PCa can be broadly categorized into modifiable and non-modifiable.The non-modifiable factors include increasing age, black race, and family history of PCa. 6,7.Some modifiable risk factors have been identified and they include lifetime alcohol consumption, 8 obesity, 9 smoking, 10 sedentary lifestyle, 11 prostatitis history, 12 and high level of cholesterol. 13Earlier studies have shown that obese men with a body mass index (BMI) greater than 30 kg/m 2 are more likely than men with a BMI lower than 25 kg/m 2 to develop PCa. 9,14,15besity is associated with food eating patterns that tend to play a key role in the development of PCa. 14 Dietary factors such as fish, high vegetable intake, vitamins C and E are protective for PCa. 16A decrease in the risk of PCa has been observed among men who had a high intake of tomatoes and garlic. 17,18Conversely, age is a key risk factor in the development of PCa, and its incidence rate increases with age.The reason behind this could be increased oxidative stress which results in the onset of PCa as age increases. 19,20Cigarette smoking is another potential risk factor for PCa development owing to changes in hormones.Smokers have increased levels of testosterone and androsterone and may be involved in cancer progression. 21ata collected from the GLOBOCAN program of the International Agency for Research on Cancer (IARC) show that PCa is also the leading cancer in sub-Saharan African men, accounting for 14% of all cancer diagnoses and 12% of all cancer-related deaths [22][23][24] Nigerian men suffer the highest burden from PCa in West Africa with estimated yearly age-adjusted incidence and mortality rates of 23.3 and 19.2 per 100 000 respectively.4][25][26] Nigeria being the most populous country in Africa, the rates and percentages above translate to a huge burden in absolute numbers of men affected by PCa.
A systematic review of PCa research in Nigeria showed that the earliest publication on PCa dates to 1973. 27This study reported a low incidence at this time, but in 1980 the incidence rate reported in this Nigerian population was similar to other black men in Washington.However, further reports have shown an increasing prevalence of PCa.Aside from this, there is a global focus on PCa differences among black men, and this demands more efforts from Africa. 27However, few risk factors have been linked to PCa development in this population.These include having a first-degree relative with PCa, decreasing height, decreasing weight, and decreasing waist circumference. 28It remains unclear whether these risk factors modulate the odds of PCa.Thus, our study was set out to assess risk factors of PCa among patients diagnosed at the University College Hospital, Ibadan.

Study Design
A case-control study design was employed for this study.

Study Population
The cases comprised incident PCa patients histologically confirmed either by biopsy, immunochemistry markers or microscopic examination admitted and/or going through treatment at the University College Hospital, Nigeria during the study period from December 2019 to March 2020.The controls were men who did not have a prior history or symptoms related to PCa recruited from orthopedic, ophthalmologic, and hypertensive clinics at the same ascertainment period as the cases and were agematched by a factor of three making the cases and controls similar in age.Study criteria included cases whose diagnosis of PCa was within two years period before study contact and must never have been interviewed as a control for the study.Controls underwent prostatic specific antigen (PSA) screening established by reading of < 3 ng/mL to reduce ascertainment bias, 29 was suitable to be matched to a case by age (± 5 years), 10,30 participants who were of Nigerian descent and were physically and mentally capable of performing the interview.The study excluded cases who were severely ill, unable to give informed consent, controls who had a personal history of cancer, and controls who had a history of radiation therapy or chemotherapy.To control for Berkson's bias, cases were patients whose primary diagnosis for hospitalization was PCa.

Sample Size
A total of 43 cases and 129 controls were estimated using the formula for case-control studies case-control studies with unequal group sizes.

Sampling Method
Consecutive sampling was employed: Every patient that met the inclusion criteria for the study was recruited until the required sample size was achieved

Data Collection Tool
The questionnaire collected information on sociodemographics, cancer history, lifestyle factors, anthropometry and occupational exposure.Medical history of other diseases such as hypertension and diabetes mellitus were collected by reviewing medical records.Face validation of the questionnaire was done by the study supervisors, translated into Yoruba, then pretested at the University College Hospital, Ibadan and preliminary analyses including reliability tests for each section of the questionnaire were carried out.

Data Collection Procedure
Medical records were reviewed weekly to identify eligible cases and controls and their appointment dates.A case was enrolled first during their urological visits, then moved to look for corresponding controls, explained the purpose of the study and interviewed them thereafter.All enrolled participants completed a pre-tested structured questionnaire with an interviewer to obtain the study's exposure variables.Data on dietary intake for food frequency was taken with the aid of a show-card and the physical activity level was assessed by adopting the International Physical Activity Questionnaire (IPAQ).Blood samples were collected under aseptic condition once by pricking from the finger among the controls for rapid PSA examination.

Statistical Analysis
Statistical Package for Social Sciences (SPSS 20) was employed for the statistical analyses.Descriptive statistics including mean, standard deviation, frequency, and proportion were used to summarize the characteristics of the cases and their controls.Chi-square and independentsample t-tests were used to determine the difference between the cases and their controls for categorical and continuous variables as appropriate.Normality of continuous variables was assessed using histogram, Q-Q plots, and Shapiro-Wilk test.Missing data were eliminated using pairwise deletion and regression analysis.Analytically, conditional logistic regression was used to estimate the odds ratio for effect size at 90% and 95% confidence intervals (CIs) for conditional univariate and multivariable logistic regressions respectively.Data analyses were in two stages; the exploratory stage identified potential risk factors defined by variables associated with PCa at P < 0.1 in a univariate logistic regression.Adjusting for confounders, all potential risk factors retaining a significant association with PCa in a multivariable logistic regression model at P < 0.05 were identified as related factors.These related factors were selected using the ENTER method (i.e., all variables were introduced and variables whose coefficients were not significant were eliminated).

Results
The distribution of the socio-demographic characteristics and cancer history of the respondents shows that a proportion of 18(41.9%) of the cases reported being < 70 years while 71(55%) of the controls reported being < 70 years alike.The mean ages for the cases and controls were 70.3 ± 8.0 years and 67.4 ± 8.7 years respectively and there was no significant difference in the mean ages of cases and controls (t test = 1.881 and P value = 0.062) as shown in Table 1.

Lifestyle and Anthropometric Characteristics of the Respondents
The distribution of lifestyle and anthropometric characteristics of cases and controls shows that 14 (32.6%) had ever smoked tobacco among the cases, similarly 46 (35.7%) of the controls had ever smoked tobacco.The mean ± SD of cigarettes smoked per day by the cases was 2.8 ± 2.3 sticks/day and 5.2 ± 6.3 sticks/day by the controls.A proportion of 32 (72.6%) cases revealed they have ever consumed any alcoholic drink while a proportion of 86 (66.7%) of controls reported having ever consumed any alcoholic drink.The mean ± SD of bottles drunk/day for the cases was 3.5 ± 4.0 and 2.7 ± 2.8 for the controls as shown in Table 2.

Dietary, Physical Activity and Occupational Exposure of Respondents
Data on dietary intake reveal that among the controls, 16 (38.1%)reported having consumed fruits for two days/ week.Similarly, in the control, the majority 49 (38%) reported having consumed fruits for two days/week.Among those cases who reported having consumed fruits, the majority 17 (43.6%)reported having consumed two servings per intake, while among the majority of the controls 57 (45.2%) reported having consumed two servings per intake alike.On vegetable intake, the majority 17 (40.5%) of the cases reported having eaten vegetables for ≥ 4 days/week, while the majority 46 (46.1%) of the cases reported having eaten vegetables for three days/week as shown in Table 3.

Risk Factors of Prostate Cancer
Univariate comparisons between cases and controls at P < 0.10 identified potential risk factors for PCa including; age, employment status, family history of cancer, family history of PCa, first degree relative with PCa, intake of a local concoction made with alcohol, age at first sex, sexual inactivity, fruit intake, vegetable intake, serving per vegetable intake, additional salt intake, and aromaticamine related occupations.Age had a significant positive  4 and 5.

Discussion
Cancer is the largest cause of death worldwide and its burden is expected to rise by 70% in the next 14 years. 22A world report on cancer has shown that by 2030 the most cancer population would be in low and middle-income countries to which Nigeria belongs. 22The substantial gaps in our present awareness about the risk factors associated with these cancers need to be resolved.The goal of this study was to assess risk factors associated with PCa among patients diagnosed at the University College Hospital, Ibadan.First-degree relative with PCa was significantly associated with PCa risk by a 3.2-fold at 90% CI.Conversely, in the fully adjusted model at 95% CI, it was insignificantly associated with PCa.This finding is similar to reports in previous studies, 11,12 but a hospital study in Nigeria has reported a significant association at 95% CI without adjusting for confounders. 28The reasons for these inconsistencies in results may be the lack of knowledge or forgetfulness of the diagnosis of PCa in a family member (measurement bias), insufficient sample size, and confounding bias.Age is taken to be the  12 The explanation behind this may be elevated oxidative stress, contributing to the onset of the disease as age increases. 19,20arkedly, the research design of age-matching did not reject the odds, with increasing age remaining a strong risk factor for PCa case status, even with the controls being within ± 5 years of their matched cases.The median age at first sex in six sub-Saharan African countries ranges from 17-20 years. 34The mean age at first sex in our study was 25 years and 22 years for the cases and controls respectively.This shows relative lateness with age at first sex among Nigerian men diagnosed with PCa.After adjusting for age and sexual activity, for every unit increase in the age at first sex there were 17.4% excess odds of developing PCa among the cases.Overage reporting (recall bias) may explain the reason for this association.This finding is inconsistent with previous reports that older age at first sex decreases the risk of PCa. 35,36This is likely due to the difference in study designs employed and as well as relying on self-reports from respondents on sexual activity in our study.Sexual activity as a risk factor has shown to be significantly associated with PCa in the fully adjusted model.Sexual activity lowered the odds of PCa by 79.5%.This result is consistent with a previous study. 13The increase in the frequency of ejaculation may reduce the carcinogens present in the prostatic fluid and stabilize testosterone levels. 37,38There may be some drawbacks to this study: The questionnaire captured selfreported information, largely depending on respondents providing the correct information.Information bias can occur, particularly when reporting on family and sexual background.However, these were minimized by ensuring that the participants were assured of a high degree of confidentiality.Misclassification bias was minimized by using a standard classification tool for all the participants.Ascertainment bias for cases and controls was minimized by using standard classification guidelines for cases and screening for PSA for controls.

Conclusion
Conclusively, increasing age, increasing age at first sex and sexual inactivity increase PCa risk in this population.However, larger-sized population studies are essential to strengthen the available evidence and reduce the inherent flaws.Finally, interventions centered on routine screening for prostatic-specific antigen with an emphasis on the elderly clinic are recommended.

Table 1 .
Frequency Distribution of Socio-demographic Characteristics and Cancer History of the Respondents *Atheist; a Brother, father, son; b Uncle, cousin, nephew; c significant at P < 0.05 in age there were 179.7% excess odds of developing PCa among the cases.After adjusting for age and sexual activity, for every unit increase in the age at first sex there were 17.4% excess odds of developing PCa among the cases and after adjusting for age and age at first sex, sexually

Table 2 .
Frequency Distribution of Lifestyle and Anthropometric Characteristics of the Respondents a Beer, spirit, palm wine, burukutu, etc.,; b Jedi-Jedi, opa-eyin, iba; c Significant at P < 0.05.

Table 3 .
Frequency Distribution of Dietary, Physical Activity and Occupational Exposure of Respondents Significant at p < 0.05 Ф Exposed to Asbestos and Asbestiforms д Exposed to Aromatic amines (e.g.2-naphthyl amine 4-aminobiphenyl), ь Exposed to Cadmium and Cadmium compounds, э Exposed to Benzene and Toluene, ђ Exposed to Poly Aromatic Hydrocarbons │ Exposed to Radiation.

Table 4 .
Associations Between Different Variables and Prostate Cancer in a Univariate Logistic Regression Model at P < 0.10

Table 4 .
[31][32][33]ongest epidemiological risk factor PCa in our study and other studies in India and Pakistan.[31][32][33]Ourstudy has shown that increasing age increases the odds of PCa among Nigerian men by 2.8 folds.It has shown that men that are less than 70 years have lower odds of PCa than older men.This report is similar to a previous study.

Table 5 .
Associations Between Age and Sexual Activity and Prostate Cancer Risk in a Multivariable Logistic Regression Model at P < 0.05