Load the foods you're looking for into a temp table, then join with LIKE :

CREATE TABLE #findfood (food_to_find nvarchar(50));

INSERT INTO #findfood (food_to_find)
VALUES ('apple'),('carrot'),('banana');

CREATE TABLE #food (food_name nvarchar(100));

INSERT INTO #food (food_name)
VALUES ('apple cider'),('cake apple'),('carrot cake'),('banana pudding');

SELECT food_to_find, COUNT(*) AS count_food_name
FROM #findfood
    INNER JOIN #food ON #food.food_name LIKE '%' + #findfood.food_to_find + '%'
GROUP BY food_to_find;

Why don't you do :

SELECT CASE WHEN food LIKE '%apple%' THEN 'apple'
            WHEN food LIKE '%carrot%' THEN 'carrot'
            WHEN food LIKE '%banana%' THEN 'banana' END AS placeholder, COUNT(*) 
FROM   Atable 
WHERE  dates > '2021-01-01' ---
GROUP  BY CASE WHEN food LIKE '%apple%' THEN 'apple'
            WHEN food LIKE '%carrot%' THEN 'carrot'
            WHEN food LIKE '%banana%' THEN 'banana' END

--> example :

CREATE TABLE Atable (food VARCHAR(32), dates date);

INSERT INTO Atable VALUES
('qsqszcbananasqwdcqdesc', '2022-01-01'), 
('qsdvdszcbananasqwdcqdeqfeqsdc', '2022-01-01'), 
('qsqszccarrotsqwdcqdesc', '2022-01-01');

--> results :

placeholder (Aucun nom de colonne)
banana      2
carrot      1